Constructed Variables CHIS 2017-2018 Adult...

CHIS 2017-2018 – Constructed Variables – Adult PUF File

1

Constructed Variables CHIS 2017-2018 Adult Survey UCLA PUF Version 1.0, 2020 Contact: California Health Interview Survey UCLA Center for Health Policy Research 10960 Wilshire Blvd., Suite 1550 Los Angeles, CA 90024 Email: [email protected]

Copyright © 2020 by the Regents of the University of California


2

TABLE OF CONTENTS INTRODUCTION .............................................................................. 8

BASIC DEMOGRAPHIC INFORMATION .......................................... 8SRAGE_P Self-Reported Age (PUF Recode) .......................................................... 8SRAGE_P1 Self-Reported Age (PUF 1-Year Recode) ............................................ 8SRSEX Gender ................................................................................................... 9SRAS Self-Reported Asian .............................................................................. 9SRCH Self-Reported Chinese ........................................................................... 9SRKR Self-Reported Korean ............................................................................ 9SRJP Self-Reported Japanese ......................................................................... 9SRPH Self-Reported Filipino ......................................................................... 10SRVT Self-Reported Vietnamese ................................................................... 10SRH Self-Reported Latino/Hispanic ........................................................... 10SRO Self-Reported Other Race ................................................................... 11SRW Self-Reported White ............................................................................ 11SRAA Self-Reported African American ......................................................... 11SRAI Self-Reported American Indian ........................................................... 11LATIN2TP Latino/Hispanic Subtypes – 2 Levels .................................................. 11LATIN9TP Latino/Hispanic Subtypes – 9 Levels .................................................. 13OMBSRR_P1 OMB Self-Reported Race Ethnicity (PUF 1-Year Recode) .......... 14OMBSRREO OMB Self-Reported Race Ethnicity .............................................. 15RACECEN Race – Census 2000 Definition ........................................................... 16RACEDOF Former DOF Race – Ethnicity ............................................................. 17RACECN_P1 Race – Census 2000 Definition (PUF 1-Year Recode) ................. 18RACEDF_P1 Race – Former Department of Finance Definition (PUF 1-Year Recode) ..... 18CATRIBE California Tribal Heritage ................................................................... 19ASIAN8 Asian Subtypes - 8 (PUF Recode) ....................................................... 19MARIT Marital Status – 3 Categories .............................................................. 20MARIT2 Marital Status – 4 Categories .............................................................. 21MARIT_45 Marital Status – Age 45 Years and Older ............................................ 21

GENERAL HEALTH ....................................................................... 21ASTCUR Current Asthma ................................................................................... 21ASTS Asthma Symptoms Past 12 Mos Population Diagnosed w/ Asthma ...... 22ASTYR Asthma Symptoms Past 12 Months ..................................................... 22AB106_P Visited ER/URGT Care for Asthma Because Unable to See Own Doctor22AB108_P1Confidence to Control and Manage Asthma (PUF 1-Year Recode) ...... 22AB42_P1 Workdays Missed Due to Asthma in Past 12 Months (PUF 1-Year Recode) ....... 22AB51_P1 Type I or Type II Diabetes (PUF 1-Year Recode) .............................. 23AB114_P1 Confidence to Control and Manage Diabetes (PUF 1-Year Recode) ... 23AB120_P1 Confidence to Control and Manage Heart Disease (PUF 1 Year Recode)23AB23_P1 Age First Told Have Diabetes (PUF 1-Year Recode) ......................... 23


3

AB27_P1 Number of Times Doctor Checked for Hemoglobin A1C Last Year (PUF 1-Year Recode) 23AB28_P1 Number of Times Doctor Checked for Feet for Sores Last ................... 23Year (PUF 1-Year Recode) ...................................................................................... 23DIABETES Doctor Ever Told Have Diabetes (Non-Gestational) .......................... 23DIAMED Taking Insulin or Pills ......................................................................... 24HGHTI_P Height – Inches (PUF Recode) ............................................................ 24HGHTM_P Height – Meters (PUF Recode) ........................................................... 24HEIGHM_P Height – Meters (UCLA) (PUF Recode) ....................................... 24WEIGHK_P Weight – Kilograms (UCLA) (PUF Recode) ................................ 24WGHTK_P Weight – Kilograms (PUF Recode) .................................................... 24WGHTP_P Weight – Pounds (PUF Recode) .......................................................... 25BMI_P Body Mass Index for Adults (PUF Recode) ........................................ 25RBMI BMI Descriptive .................................................................................. 25WHOBMI Body Mass Index: WHO Definition .................................................... 25OVRWT Overweight or Obese ........................................................................... 25

HEALTH BEHAVIORS ................................................................... 26AE_FRIES Number of times eating fries per week ................................................. 26AE_FRUIT Number of times eating fruit per week ................................................. 26AE_VEGI Number of times eating vegetables per week ........................................ 26AE_SODA Number of times drank soda per week ................................................ 26AESODA_P1 Number of times drank soda per week (PUF 1-Year Recode) ...... 26SMKCUR Current Smoker .................................................................................... 27SMOKING Current Smoking Habits ...................................................................... 27AE16_P1 Number of Cigarettes Per Day in Past 30 Days (PUF 1-Year Recode)27AD32_P1 Number of Cigarettes Per Day (PUF 1-Year Recode) ........................ 27NUMCIG Number of Cigarettes Per Day ............................................................ 27AC31_P1 Number of Times Ate Fast Food in the Past Week (PUF 1-Year Recode)AC34_P1 Number of Times Male Had 5+ Drinks in Past 12 Months (PUF 1-Year Recode) ………………………………………………………………………….28AC35_P1 Number of Times Female Had 4+ Drinks in Past 12 Months (PUF 1-Year Recode) ………………………………………………………………………….28AC42_P How Often Find Fresh Fruit/ Veg. in Neighborhood (PUF Recode) .. 28AC47_P1 Number of Classes of Water Drank Yesterday (PUF 1-Year Recode)28 AC48_P1 Number of Classes of Low-Fat or Non-Fat Milk Drank Yesterday (PUF 1-Year Recode) ………………………………………………………………………….28AJ143_P1 Place Received Main Birth Control Method or Prescription (PUF 1-Year Recode) ………………………………………………………………………….29AJ146_P1 Place Received Main Birth Control Method (PUF 1-Year Recode) .... 29AC53NUM_P1 How Long Since Smoked on Daily Basis (PUF 1-Year Recode) . 29AC54NUM_P1How Soon After Awaking Smokes 1st Cigarette (PUF 1-Year Recode)…30AC55_P1 Where Usually Buy Cigarettes (PUF 1-Year Recode) ........................ 30AC80_P1 Use Hookah Pipe Every Day (PUF 1-Year Recode) ........................... 30AC82_P1 Days Used E-Cigs in Past 30 Days (PUF 1-Year Recode) ................. 31AC52_P1 Age Began Smoking Cigarettes Regularly (PUF 1-Year Recode) . 31


4

AJ142_P1 Main Type of Birth Control Received From Doctor in Past Year (PUF 1-Year Recode) …………………………………………………………………………………31PREDIAB Doctor Ever Told Have Pre- or Borderline Diabetes (Non Gestational)…….31

MENTAL HEALTH ........................................................................ 32DISTRESS Serious Psychological Distress Past Month ........................................ 32DSTRS_P1 Serious Psychological Distress Past Month (PUF 1-Year Recode) .... 32DSTRS12 Likely Has Psychological Distress Past Year ....................................... 32DSTRS30 Likely Has Psychological Distress Past Month .................................... 33DSTRSYR Serious Psychological Distress for Worst Month in Past Year ........... 33CHORES2 Chore Impairment Past 12 Months ....................................................... 33SOCIAL2 Social Life Impairment Past 12 Months .............................................. 33WORK2 Work Impairment Past 12 Months ...................................................... 34FAMILY2 Family Life Impairment Past 12 Months ............................................. 34

ADDITIONAL DEMOGRAPHIC INFORMATION ............................. 34AH33NEW Born in U.S. ......................................................................................... 34AH34NEW Mother Born In U.S. ............................................................................ 34AH35NEW Father Born In U.S. ............................................................................. 35LANGHOME Types of Languages Spoken at Home ........................................... 35LNGHM_P1 Types of Languages Spoken at Home (PUF 1-Year Recode) ...... 36SPK_ENG English Use and Proficiency ................................................................ 38CITIZEN2 Citizenship Status for Adults ............................................................... 38YRUS_P1 Years Lived in the U.S. (PUF 1-Year Recode) .................................... 38PCTLF_P Percent Life in U.S. (PUF Recode) ..................................................... 39WRKST Working Status .................................................................................... 39WRKST_P1 Working Status (PUF 1-Year Recode) .......................................... 40WRKST_S Spouse Working Status ........................................................................ 41AHEDC_P1 Educational Attainment (PUF 1-Year Recode) ............................. 41AHEDUC Educational Attainment ....................................................................... 41AKWKLNG Time at Main Job ........................................................................... 42SERVED Length of Time Served in Active Duty ............................................... 42FAMTYP_P Family Type (PUF Recode) ........................................................... 42FAMT4 Family Type – 4 Levels ......................................................................... 43FAMSIZE2_P1 Family Size Supported by Household Income (PUF 1-Year recode) ………………………………………………………………………….43AK33_P1 Number of Persons in U.S. Supported by Household Income ............. 43Not Currently Residing in Household (PUF 1-Year Recode) .................................. 43

HEALTH INSURANCE .................................................................... 43COVRDCA Private/Employer Based Coverage from Covered California ............. 43AI45_P1 Main Reason Spouse Not in Employer Health Plan (PUF 1-Year Recode) ......... 44AH45A_P1 Main Reason Spouse Ineligible for Employer Health Plan ................. 44(PUF 1-Year Recode) ............................................................................................... 44AH52_P1 Sign for Medicare HMO Directly or Otherwise (PUF 1-Year Recode) 44


5

HMO HMO Status ......................................................................................... 45AI22A_P Name of Health Plan (PUF Recode) ................................................... 45AI25NEW RX Coverage Edited for Medi-Cal/HF ................................................ 45IHS Covered by Indian Health Services .................................................... 45INS Currently Insured ................................................................................ 46INS12M Number of Months Covered by Health Insurance in Past 12 Month ..46INS64 Type of Current Health Coverage Source – Under 65 ........................ 47INS64_P Current Health Coverage Under 65 Yrs Old (PUF Recode) ............... 47INS65 Type of Current Health Coverage Source for the Elderly ................... 48INSANY Any Health Insurance in the Last 12 Months ...................................... 49INSEM Covered by Employer-Based Plans ..................................................... 49INSMC Covered by Medicare .......................................................................... 50INSMD Covered by Medi-Cal .......................................................................... 50INSOG Covered by Other Government Plans .................................................. 50INSOT Covered by other non-government insurance plan ................................ 51INSOT_S Spouse covered by other insurance (non-government plans) ................ 51INSPR Covered by Plans Purchased On Own ................................................. 51INSPS Covered by Employer-Based Plans as Primary Coverage .................. 52INSTYPE Insurance Type .................................................................................... 52INST_12 Health Ins Coverage in Last 12 Mos, Incl Current Status: 8 Lvls ....... 53INSLT_P Health Ins Coverage in Last 12 Mos, Incl Current Status: 8 Lvls (PUF Recode) ………………………………………………………………………….53INS9TP Type of Current Health Coverage Source for All Ages – 9 Levels ........ 53INSLMT12M Reached Health Insurance Payment Limit in Past 12 Months ........ 54HSAAMT_P1 Amount In Health Savings Accounts (PUF 1-Year Recode) .......... 54PINSTYPEType of Health Insurance Adult Had Before Current Plan ................... 54TCURPLAN Time Adult Has Been Enrolled in Current Plan: Months ............... 55UNINSANY Uninsured in Past 12 Months ......................................................... 55OFFTK Offer, Eligibility, Acceptance of Employer-Based Insurance (EBI) ... 55MC_TYPE Type of Medicare Coverage ................................................................ 56AH101_P Person who helped Find Health Plan ................................................... 56AH95_P # Times Visited ER in Past 12 Months (PUF Recode) ....................... 56AH95_P1 # Times Visited ER in Past 12 Months (PUF 1-Year Recode) ........... 57AH71_P1 Health Plan Deductible More than $1,000 (PUF 1-Year Recode) ...... 57AH72_P1 Health Plan Deductible More than $2,000 (PUF 1-Year Recode) ...... 57AH98_P1 Difficulty Finding Plan with Needed Coverage (PUF 1-Year Recode) . 57AH99_P1 Difficulty Finding Plan that is Affordable (PUF 1-Year Recode) ......... 57ELGMAGI3 Medi-Cal Eligibility for Uninsured: ACA MAGI (2 Levels) ........ 58

HEALTH CARE UTILIZATION AND ACCESS ................................ 58ACMDNUM Number of Doctor Visits in the Past Year ..................................... 58DOCT_YR Visited a Doctor During the Past 12 Months ...................................... 59AJ50_P Language Doctor Speaks To Respondent (PUF Recode) .................... 59USOC Usual Source of Care Other Than ER ................................................. 59USUAL Have Usual Place to Go When Sick or Needing Health Advice ......... 59AH3_P1 Kind of Place for Usual Source of Health Care (PUF 1-Year Recode)60


6

USUAL_TP Usual Source of Care – 7 Levels .................................................... 60USUAL5TP Usual Source of Care – 5 Levels .................................................... 60AJ131_P1 Main Reason for Delaying Needed Care (PUF 1-Year Recode) .......... 61FORGO Had to Forgo Necessary Care .............................................................. 61RN_FORGO Reasons Forgone Necessary Care .................................................. 61ER Emergency Room Visit in the Past Year ............................................ 62CARE_PV Had a Preventive Care Visit in the Past Year ....................................... 62TIMAPPT Able to Get an Appointment in a Timely Way .................................... 62

EMPLOYMENT, INCOME, POVERTY, & FOOD SECURITY .......... 62AG9_P1 Type of Employer on Spouse’s Main Job (PUF 1-Year Recode) ....... 62AK10_P Earnings Last Month Before Taxes and Deductions (PUF Recode) ... 63AK10A_P Spouse’s/Partner’s Earnings Last Month (PUF Recode) ...................... 63AK2_P1 Main Reason for Not Working Last Week (PUF 1-Year Recode) ........ 63AK7_P1 Length of Time Working at Main Job (PUF 1-Year Recode) ............... 63AK22_P Household’s Total Annual Income (PUF Recode) .............................. 63FPG Federal Poverty Guideline .................................................................. 63POVGWD_P Family Poverty Threshold Level (PUF Recode) ........................... 63POVGWD2 Family Poverty Threshold Level ........................................................ 64POVLL Poverty Level ....................................................................................... 65POVLL_ACA Poverty Level as Times of 100% FPL (PUF Recode) ................... 66ELIGPRG3 Uninsured Medi-Cal/Healthy Families Eligible (3 levels) .................. 66FSLEV Food Security Status Level .................................................................. 67FSLEVCB Food Security Status (2 Levels) ........................................................... 67ELDER_IDX Elderly Single/Couple Income Below County Cost of Living Thresholds .... 67

PUBLIC PROGRAM PARTICIPATION ............................................ 68AL18_P Total Alimony/Child Support Paid Last Month (PUF Recode) .......... 68

GEOGRAPHICAL INFORMATION .................................................. 68UR_CLRT2 Rural and Urban – Claritas (By Zip Code) (2 levels) ......................... 68UR_IHS Rural and Urban – Indian Health Service ............................................ 69UR_OMB Rural and Urban – OMB ..................................................................... 69UR_RHP Rural and Urban – Office of Rural Health Policy ............................... 69UR_BG6 Rural and Urban – Claritas (By Block Group) (6 Levels) ..................... 70UR_CLRT6 Rural and Urban – Claritas (By Zip Code) (6 Levels) ......................... 71UR_TRACT6 Rural and Urban – Claritas (By Census Tract) (6 Levels) .............. 72

OTHER CONSTRUCTED VARIABLES ............................................ 73AM38_P1 Main Reason for Last Move (PUF 1-Year Recode) ............................ 73HHSIZE_P Household Size (PUF Recode) ............................................................ 73HHSIZE_P1 Household Size (PUF 1-Year Recode) .......................................... 73PROXY Proxy Interview ................................................................................... 73TIMEAD_P1 Length of Time Lived at Current Address (Months) (PUF 1-Year Recode) 73


7

SRTENR Self-Reported Household Tenure (HH) ............................................... 74INTVLANG Language of Interview ................................................................... 74INTVLNGC_P1 Language of Interview (PUF 1-Year Recode) .............................. 74AHCHLDC Amount Per Week Paid For Child Care ............................................. 75ACHLDC_P1 Amount Per Week Paid for Child Care (PUF 1-Year Recode) ..... 75PC_NEWP Primary care: have difficulty finding a provider that accepts new patient ......... 76PC_INS Primary care: have difficulty finding a provider that accepts their insurance .... 76SC_NEWP Specialty care: have difficulty finding a provider that accepts new patient ....... 76SC_INS Specialty care: have difficulty finding a provider that accepts their insurance .. 76VOTE_REG Currently Registered to Vote ......................................................... 76VOTE_RSNNOT Main Reason Not Registered to Vote ......................................... 76VOTE_PRES16 Voted in 2016 Presidential Election ............................................. 77VOTE_ENG1 Voter Engagement ......................................................................... 77


8

INTRODUCTION This document provides details on how constructed variables that appear in 2017-2018 Public Use Files (PUFs) are created. The variables in this document may appear in the 2017 PUF, the 2018 PUF, and/or the 2017/2018 2-Year PUF. The information provided in this document is meant to serve as a guide for users of CHIS data to understand how these variables are constructed to provide needed insight for further research and analysis, as well as the potential for user re-construction. Many CHIS variables have response categories that you’ll find consistently in the data across variables that are not found described in this dictionary. These categories and values often represent responses that are either inapplicable, missing, refused, or some other reason why a respondent does not have a response for the given variable. These values are consistent across variables and are defined as outlined below:

Value: Label: -1 Inapplicable -2 Proxy Skipped

-5 Insured (for some uninsured eligibility variables)

-7 Refused -8 Don’t Know -9 Not Ascertained

Please note that while this document includes the conditions that describe how each variable is constructed, not all values may appear in the corresponding year's source data dictionary if responses were not recorded for them. This may be the case if certain response categories within a variable are not applicable to the respondent population due to age group. In other words, if there were no respondents that selected a certain response category for a question in a given year, that value will still be defined in this document but will not be included in that year's source data dictionary. If there are any questions regarding this document or any of the information contained within it, please contact the CHIS Data Access Center (DAC) at [email protected].

BASIC DEMOGRAPHIC INFORMATION

SRAGE_P Self-Reported Age (PUF Recode) The SRAGE_P is a top coded version of SRAGE for adults. SRAGE is a data collection vendor generated variable which assigns current age to the respondent. For missing values, SRAGE is imputed. SRAGE_P has been created specifically for CHIS 2-year Public Use File data files to ensure respondent confidentiality and also provide granular-level detail. Note: Top code is 85

SRAGE_P1 Self-Reported Age (PUF 1-Year Recode) The SRAGE_P1 is a recoded version of SRAGE, which groups respondent age to create a 14-level categorical variable. SRAGE is a data collection vendor generated variable that assigns current age to the respondent. For missing values, SRAGE is imputed.


9

SRAGE_P1 has been created specifically for CHIS one-year data files to ensure respondent confidentiality and also provide granular-level detail.

SRSEX Gender SRSEX is a variable created by the CHIS data collection vendor. It is a dichotomous variable indicating the self-reported gender of the adult respondent.

SRAS Self-Reported Asian SRAS is dichotomous indicator of whether or not a respondent self-reports as being Asian. SRAS was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where race is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRCH Self-Reported Chinese SRCH is a dichotomous indicator of whether or not a respondent self-reports as being Chinese. SRCH was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRKR Self-Reported Korean SRKR is a dichotomous indicator of whether or not a respondent self-reports as being Korean as defined by the Office of Management and Budget. SRKR was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRJP Self-Reported Japanese SRJP is a dichotomous indicator of whether or not a respondent self-reports as being Japanese as defined by the Office of Management and Budget. SRJP was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.


10

SRPH Self-Reported Filipino SRPH a dichotomous indicator of whether or not a respondent self-reports as being Filipino as defined by the Office of Management and Budget. SRPH was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRVT Self-Reported Vietnamese SRVT is a dichotomous indicator of whether or not a respondent self-reports as being Vietnamese. SRVT was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRASO OMB Self-Reported Other Asian Group SRASO is a dichotomous indicator of whether or not a respondent self-reports as belonging to Other Asian ethnicity as defined by the Office of Management and Budget. SRASO was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRASO2 OMB Self-Reported Other Asian Group (Excludes Japanese) SRASO2 is a dichotomous indicator of whether or not a respondent self-reports as belonging to Other Asian ethnicity as defined by the Office of Management and Budget. SRASO was constructed by the CHIS data collection vendor, and is based on AA5A and AA5E. It differs from SRASO because it no longer includes respondents who self-report being Japanese. The data collection vendor imputes cases through the hot deck approach where Asian ethnicity is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRH Self-Reported Latino/Hispanic SRH is dichotomous indicator of whether or not a respondent self-reports a Latino or Hispanic ethnicity. SRH was constructed by the CHIS data collection vendor, and is primarily based on AA4. The data collection vendor imputes cases through the hot deck approach where race or ethnic information is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.


11

SRO Self-Reported Other Race SRO is a dichotomous indicator of whether or not a respondent self-reports as being of a race other than White, Black/African American, Asian, American Indian/Alaska Native/Native American, Other Pacific Islander, or Native Hawaiian. SRO was constructed by the CHIS data collection vendor, and is primarily based on AA5A. The data collection vendor imputes cases through the hot deck approach where race or ethnic information is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRW Self-Reported White SRW is dichotomous indicator of whether or not a respondent self-reports as being white. SRW was constructed by the CHIS data collection vendor, and is primarily based on AA5A. The data collection vendor imputes cases through the hot deck approach where race or ethnic information is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRAA Self-Reported African American SRAA is a dichotomous indicator of whether or not a respondent self-reports as being African American. SRAA was constructed by the CHIS data collection vendor, and is primarily based on AA5A. The data collection vendor imputes cases through the hot deck approach where race or ethnic information is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

SRAI Self-Reported American Indian SRAI is a dichotomous indicator of whether or not a respondent self-reports as being American Indian. SRAI was constructed by the CHIS data collection vendor, and is primarily based on AA5A. The data collection vendor imputes cases through the hot deck approach where race or ethnic information is missing. For more information about the imputation process, refer to CHIS Methodology Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

LATIN2TP Latino/Hispanic Subtypes – 2 Levels The purpose of this variable is to provide a 2-level measurement of which group(s) the respondents identify with of those who report that they are of Latino/Hispanic origin (if SRH=1). This variable is derived from items, SRH and AA5_1 through AA5_21 (14-21 are upcoded categories). A. First, the numbers of Latino/Hispanic ancestries reported for each case are counted using items AA5_1 through AA_21. B. The cases with respondents who report only one Latino/Hispanic ancestry are assigned values for the temporary variable LATINTEMP.


12

1. The respondents who report that they belong to a single Latino/Hispanic ancestry are assigned

the following LATINTEMP values:

Condition: LATINTEMP Value: LATINTEMP Label:

If AA5_1=1 (Mexican/Mexicano) or AA5_2=1 (Mexican American) or AA5_3=1 (Chicano) or

1 Mexican

If AA5_4=1 (Salvadoran)

2 Salvadoran

If AA5_5=1 (Guatemalan)

3 Guatemalan

If AA5_6=1 (Costa Rican) or AA5_7=1 (Honduran) or AA5_8=1 (Nicaraguan) or AA5_9=1 (Panamanian) or

4 Central American

If AA5_10=1 (Puerto Rican)

5 Puerto Rican

If AA5_11=1 (Cuban)

6 Cuban

If AA5_12=1 (Spanish American) or AA5_17=1 (Portuguese) or AA5_20=1 (Other European Origin)

7 Latino European

If AA5_14=1 (Colombian) or AA5_15=1 (Argentinean) AA5_16=1 (Peruvian) AA5_19=1 (Other South American Origin)

8 South American

If AA5_18=1 (Other Caribbean Origin) or AA5_21=1 (Other Latino/Hispanic)

9 Other Latino

C. The cases with more than one ancestry reported are assigned LATINTEMP values.

1. The cases with more than one ancestry reported are all assigned to the “two or more Latino types” category:


More than one Latino/Hispanic Ancestry

10 Two or more Latino types

D. The respondents who report that they are not of Latino or Hispanic origin (if SRH=2) are assigned a skip value (-1) for this variable. E. Finally, all of the cases are assigned values for the LATIN2TP variable using the categories generated with LATINTEMP. Each case is tested through the following conditions until a LATIN2TP is assigned:

Condition: LATIN2TP Value: LATIN2TP Label: If LATINTEMP=1 (Mexican) 1 Mexican If LATINTEMP=2 (Salvadoran), 3 2 Other


13

(Guatemalan), 4 (Central American), 5 (Puerto Rican), 6 (Cuban), 7 (Latino European), 8 (South American), 9 (Other Latino), 10 (Two or more Latino Types) If LATINTEMP=-1 -1 Non-Latino

F. If LATIN2TP cannot be determined, and AA5 is missing, then country of birth is used (AH33) to assign Latino ethnic group in LATIN2TP. LATIN9TP Latino/Hispanic Subtypes – 9 Levels The purpose of this variable is to provide a 9-level measurement of which group(s) the respondents identify with of those who report that they are of Latino/Hispanic origin (if SRH=1). This variable is derived from items, SRH and AA5_1 through AA5_21 (14-21 are upcoded categories). A. First, the numbers of Latino/Hispanic ancestries reported for each case are counted using items AA5_1 through AA_21. B. The cases with respondents who report only one Latino/Hispanic ancestry are assigned values for the temporary variable LATINTEMP.

2. The respondents who report that they belong to a single Latino/Hispanic ancestry are assigned the following LATINTEMP values:


If AA5_1=1 (Mexican/Mexicano) or AA5_2=1 (Mexican American) or AA5_3=1 (Chicano) or

1 Mexican

If AA5_4=1 (Salvadoran)

2 Salvadoran

If AA5_5=1 (Guatemalan)

3 Guatemalan

If AA5_6=1 (Costa Rican) or AA5_7=1 (Honduran) or AA5_8=1 (Nicaraguan) or AA5_9=1 (Panamanian) or

4 Other Central American

If AA5_10=1 (Puerto Rican)

5 Puerto Rican

If AA5_11=1 (Cuban)

6 Cuban

If AA5_12=1 (Spanish American) or AA5_17=1 (Portuguese) or AA5_20=1 (Other European Origin)

7 Latino European

If AA5_14=1 (Colombian) or AA5_15=1 (Argentinean) AA5_16=1 (Peruvian) AA5_19=1 (Other South American Origin)

8 South American

If AA5_18=1 (Other Caribbean Origin) or AA5_21=1 (Other Latino/Hispanic)

9 Other Latino


14

C. The cases with more than one ancestry reported are assigned LATINTEMP values.

2. The cases with more than one ancestry reported are all assigned to the “Two or more Latino types” category:


More than one Latino/Hispanic Ancestry

10 Two or more Latino types

D. The respondents who report that they are not of Latino or Hispanic origin (if SRH=2) are assigned a skip value (-1) for this variable. E. Finally, all of the cases are assigned values for the LATIN9TP variable using the categories generated with LATINTEMP. Each case is tested through the following conditions until a LATIN2TP is assigned:

Condition: LATIN9TP Value: LATIN9TP Label:

If LATINTEMP=1 1 Mexican If LATINTEMP=2 2 Salvadoran If LATINTEMP=3 3 Guatemalan If LATINTEMP=4 4 Central American If LATINTEMP=5 5 Puerto Rican If LATINTEMP=7 6 Latino European If LATINTEMP=8 7 South American If LATINTEMP=6 or 9 8 Other Latino If LATINTEMP=10 9 2+ Latino Types If LATINTEMP=-9 -9 Latino – Subtype not

ascertained F. If LATIN9TP cannot be determined, and AA5 is missing, then country of birth (AH33) or mother’s/father’s countries of birth (AH34 and AH35) are used to assign Latino ethnic group in LATIN9TP, if possible. If LATIN9TP=-9 (Subtype not ascertained), then:

Condition: LATIN9TP Value: LATIN9TP Label:

If AH33=5 2 Salvadoran If AH33=10 3 Guatemalan If AH33=18 or AH34=18 or AH35=18 1 Mexican If AH33=6, 8, 11, 14, 21, 30 or AH34=6, 8, 11, 14, 21, 30 or AH35=6, 8, 11, 14, 21, 30

6 Latino European

If AH33=28 7 South American If AH33 = all others 8 Other Latino

OMBSRR_P1 OMB Self-Reported Race Ethnicity (PUF 1-Year Recode) OMBSRR_P1 is a recoded version of OMBSRREO. The OMBSRREO variable follows the Office of Management and Budget-revised guidelines (1997) and the Census modification of the OMB guidelines that combines ethnicity (Hispanic/Latino identification) with OMB race categories. OMBSRREO is


15

constructed by the CHIS data collection vendor, and separates race and ethnicity with the following categories:

OMBSRREO Value:

OMBSRREO Label:

1 Hispanic 2 White, Non-Hispanic (NH) 3 African American Only, NH 4 American Indian/Alaskan Native Only,

NH 5 Asian Only, NH 6 Native Hawaiian/Pacific Islander, NH 7 Two or More Races, NH

If the person self-reported as Latino (SRH=1), the variable OMBSRREO was set to 1. This assignment is independent of the values of the race variables. If the person self-reported as non-Latino (SRH=2) and reported OMB race alone or in combination with one or more OMB races (e.g., White alone, White and Black or African American, White and Black or African American and American Indian and Alaska Native) then OMBSRREO was given the value 2, 3, 4, 5, 6 or 7 depending on the values of SRW, SRAA, SRAI, SRAS, and SRPI. In order to use the California Department of Finance population estimates as control totals for survey weighting, any sampled person who reported Some Other Race had to be recoded into the OMB categories. If the person self-reported as non-Latino (SRH=2) and reported both an OMB race and “Some Other Race” (SRO=1), then OMBSRREO was assigned using only the specified OMB race(s) and considered as single race respondents based on the OMB definition. For example, non-Latino White and some other race became non-Latino White alone. Persons who self-reported as non-Latino and “Some Other Race” only (SRH=2 and SRO=1) had their values imputed into one of the OMB race categories. For more information, please refer to Chapter 8 Methodology Series Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf. The OMBSSREO values were further collapsed as follows for the OMBSRR_P1 one-year recode.

Condition: OMBSRR_P1 Value:

OMBSRR_P1 Label:

If OMBSRREO = 1 1 Hispanic If OMBSRREO = 2 2 White, Non-Hispanic (NH) If OMBSRREO = 3 3 African American Only, NH If OMBSRREO = 4 4 American Indian/Alaskan Native

Only, NH If OMBSRREO = 5 5 Asian Only, NH If OMBSRREO in (6,7) 6 Other/Two or More Races, NH

OMBSRREO OMB Self-Reported Race Ethnicity The OMBSRREO variable follows the Office of Management and Budget-revised guidelines (1997) and the Census modification of the OMB guidelines that combines ethnicity (Hispanic/Latino identification) with OMB race categories. OMBSRREO is constructed by the CHIS data collection vendor, and separates race and ethnicity with the following categories:


16

OMBSRREO Value:

OMBSRREO Label:

1 Hispanic 2 White, Non-Hispanic (NH) 3 African American Only, NH 4 American Indian/Alaskan Native Only,

NH 5 Asian Only, NH 6 Native Hawaiian/Pacific Islander, NH 7 Two or More Races, NH

If the person self-reported as Latino (SRH=1), the variable OMBSRREO was set to 1. This assignment is independent of the values of the race variables. If the person self-reported as non-Latino (SRH=2) and reported OMB race alone or in combination with one or more OMB races (e.g., White alone, White and Black or African American, White and Black or African American and American Indian and Alaska Native) then OMBSRREO was given the value 2, 3, 4, 5, 6 or 7 depending on the values of SRW, SRAA, SRAI, SRAS, and SRPI. In order to use the California Department of Finance population estimates as control totals for survey weighting, any sampled person who reported Some Other Race had to be recoded into the OMB categories. If the person self-reported as non-Latino (SRH=2) and reported both an OMB race and “Some Other Race” (SRO=1), then OMBSRREO was assigned using only the specified OMB race(s) and considered as single race respondents based on the OMB definition. For example, non-Latino White and some other race became non-Latino White alone. Persons who self-reported as non-Latino and “Some Other Race” only (SRH=2 and SRO=1) had their values imputed into one of the OMB race categories. For more information, please refer to Chapter 8 Methodology Series Report 5: http://healthpolicy.ucla.edu/chis/design/Documents/CHIS_2017-2018_MethodologyReport5_WeightingAndVarianceEstimation.pdf.

RACECEN Race – Census 2000 Definition The RACECEN variable uses the Census SF1 definition/tabulation of race. RACECEN is derived from the imputed data collection vendor variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. Cases are assigned either to one of several single-race categories or to a multiple-race category. 1. The number of races reported for each case is counted using the race variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. 2. The cases with a single race reported are assigned RACECEN values:

Condition:

RACECEN Value: RACECEN Label:

If only SRPI=1 1 Pacific Islander If only SRAI=1 2 American Indian/Alaska

Native If only SRAS=1 3 Asian If only SRAA=1 4 African American If only SRW=1 5 White If only SRO=1 6 Other Single Race

3. The cases with more than one race reported are assigned to the multiple-race category:


17

Condition: RACECEN Value: RACECEN Label:

If more than one of the following (SRPI, SRAI, SRAS, SRAA, SRW, SRO)=1

7 More Than One Race

RACEDOF Former DOF Race – Ethnicity The RACEDOF variable uses the former definition of race classification from the California Department of Finance’s race categories. This variable is derived from the imputed data collection vendor variables SRH, SRPI, SRAI, SRAS, SRAA, SRW and SRO. Latino is considered to be a race category for this variable and is given priority. RACEDOF values are assigned in the following hierarchical manner: 1. The number of races reported for each case is counted using the race variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. 2. All cases that are reported to be Latino (if SRH=1) are assigned to the Latino category:

Condition: RACEDOF Value: RACEDOF Label:

If SRH=1 1 Latino 3. The remaining cases with a single race reported are assigned to one of several non-Latino categories:

Condition:

RACEDOF Value: RACEDOF Label:

If only SRPI=1 2 Non-Latino Pacific Islander

If only SRAI=1 3 Non-Latino American Indian/Alaska Native

If only SRAS=1 4 Non-Latino Asian If only SRAA=1 5 Non-Latino Afr. Amer. If only SRW=1 6 Non-Latino White If only SRO=1 7 Non-Latino Other, One

Race 4. The remaining cases with more than one race reported are assigned to the non-Latino multiple-race category:

Condition: RACEDOF Value: RACEDOF Label:

If more than one of the following (SRPI, SRAI, SRAS, SRAA, SRW, SRO) = 1

8 Non-Latino, Two+ Races

Note: The non-Latino single race category (RACEDOF=7) is not included in the original population projection by the Department of Finance (DOF). Corrections for this category assignment will be made in the construction of CHIS RACEDOF variables (2005 and beyond).


18

RACECN_P1 Race – Census 2000 Definition (PUF 1-Year Recode) The RACECN_P1 variable uses the Census SF1 definition/tabulation of race. RACECN_P1 is derived from the imputed data collection vendor variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. Cases are assigned either to one of several single-race categories or to a multiple-race category. 1. The number of races reported for each case is counted using the race variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. 2. The cases with a single race reported are assigned RACECN_P1 values:

Condition:

RACECN_P1 Value: RACECN_P1 Label:

If only SRPI=1 or only SRO=1 1 Other single race If only SRAI=1 2 American Indian/Alaska

Native If only SRAS=1 3 Asian If only SRAA=1 4 African American If only SRW=1 5 White

3. The cases with more than one race reported are assigned to the multiple-race category:

Condition: RACECN_P1 Value: RACECN_P1 Label:

If more than one of the following (SRPI, SRAI, SRAS, SRAA, SRW, SRO)=1

7 More than one race

RACEDF_P1 Race – Former Department of Finance Definition (PUF 1-

Year Recode) The RACEDF_P1 variable uses the former definition of race classification from the California Department of Finance’s race categories. This variable is derived from the imputed data collection vendor variables SRH, SRPI, SRAI, SRAS, SRAA, SRW and SRO. Latino is considered to be a race category for this variable and is given priority. RACEDF_P1 values are assigned in the following hierarchical manner: 1. The number of races reported for each case is counted using the race variables SRPI, SRAI, SRAS, SRAA, SRW, and SRO. 2. All cases that are reported to be Latino (if SRH=1) are assigned to the Latino category:

Condition: RACEDF_P1 Value: RACEDF_P1 Label:

If SRH=1 1 Latino 3. The remaining cases with a single race reported are assigned to one of several non-Latino categories:

Condition:

RACEDF_P1 Value: RACEDF_P1 Label:

If only SRPI=1 or only 2 Non-Latino Other, One


19

SRO=1 Race If only SRAI=1 3 Non-Latino American

Indian/Alaska Native If only SRAS=1 4 Non-Latino Asian If only SRAA=1 5 Non-Latino Afr. Amer. If only SRW=1 6 Non-Latino White

4. The remaining cases with more than one race reported are assigned to the non-Latino multiple-race category:

Condition: RACEDF_P1 Value: RACEDF_P1 Label:

If more than one of the following (SRPI, SRAI, SRAS, SRAA, SRW, SRO) =1

8 Non-Latino, Two + Races

Note: The non-Latino single race category (RACEDF_P1=7) is not included in the original population projection by the Department of Finance (DOF). Corrections for this category assignment will be made in the construction of future CHIS RACEDOF variables (2005 and beyond).

CATRIBE California Tribal Heritage The California Tribal Heritage variable indicates whether or not the respondents who report themselves as being American Indian/Alaska Native (if SRAI=1) identify themselves with a California or a non-California tribal heritage. This variable is constructed using questionnaire items AA5B_1 – AA5B_11, SRAI, AA5BOS, and AA5DOS. Since the questionnaire response categories in AA5B include only non-California tribes, it was important to construct CATRIBE in order to capture verbatim responses in AA5BOS or AA5DOS that may indicate California tribal heritage among the population of American Indians responding to the questionnaire. Therefore, any American Indian/Alaska Native respondents (if SRAI=1) who report a California Tribe in AA5BOS or AA5DOS are considered to have California Tribal Heritage (CATRIBE=1). All remaining respondents who identify themselves with at least one of the non-California tribes in AA5B_1-11, or indicate a non-California tribe as a verbatim answer in AA5BOS, are considered to be of non-California Tribal Heritage (CATRIBE=2). Respondents who do not indicate being of American Indian/Alaska Native heritage are assigned a skip value, CATRIBE= (-1). All other respondents whose tribe cannot be ascertained are assigned a value (-9) for this variable. Note: This variable indicates reported tribal heritage. The cases included in this variable all reported themselves as American Indian/Alaska Natives (if SRAI=1), but may or may not be enrolled members of a federal or state recognized tribe (please see the AA5C variable for this information).

ASIAN8 Asian Subtypes - 8 (PUF Recode) The ASIAN8 variable is derived from ASIAN10 variable, which utilizes questionnaire items AA5A_4 and AA5E_1 through AA5E_23 (19-23 are re-classified categories). The purpose of this variable is to provide a re-categorization representing the specific Asian subgroups for those who report that they are Asian (if AA5A_4=1).


20

A. First, the number of Asian ethnic groups reported for each case is counted using ASIAN10_1 through ASIAN10_10

B. The cases with adults who report only one Asian ethnic group are assigned ASIAN8 values.

1. The adults who report only one ethnic group in ASIAN10 are assigned the following values for the ASIAN8 variable:

Condition: ASIAN8 Value:

ASIAN8 Label:

If ASIAN10=1

1 Chinese

If ASIAN10=2

2 Japanese

If ASIAN10=3

3 Korean

If ASIAN10=4

4 Filipino

If ASIAN10=5

5 South Asian

If ASIAN10=6

6 Vietnamese

If ASIAN10=7 or 8

7 Southeast Asian

If ASIAN10= 9 or 10

8 Other Asian/2+ Asian Types

2. The adults who only report one Asian ethnic group, and are not yet assigned an ASIAN8

value, are imputed to assign an ASIAN8 value.

3. Adults that are identified as not belonging to an ASIAN ethnic or race group are assigned the value of ASIAN8=(-1).

Note: This variable is available in the CHIS 2017-2018 Two-Year file only.

MARIT Marital Status – 3 Categories MARIT is a three-level variable constructed from the variable AH43, which displays the marital status of the respondent.

Condition:

MARIT Value: MARIT Label:

If AH43=1 1 Married If AH43=2, 3, 4, 5 2 Other

(Widowed/Separated/Divorced/Living with Partner)

If AH43=6 3 Never Married If AH43=-1 -1 Inapplicable If AH43<0 -9 Not Ascertained


21

MARIT2 Marital Status – 4 Categories MARIT2 is a four-level variable constructed from the variable AH43, which displays the marital status of the respondent.

Condition:

MARIT2 Value: MARIT2 Label:

If AH43=1 1 Married If AH43=2 2 Living with Partner If AH43=3, 4, 5 3 Widowed/Separated/Divorce

d If AH43=6 4 Never Married If AH43=-1 -1 Inapplicable If AH43=-9 -9 Not Ascertained

MARIT_45 Marital Status – Age 45 Years and Older MARIT_45 is a recategorized variable replicating AH43 (Marital Status) for only those aged 45 years and older. If SRAGE < 45 then MARIT_45 = -1.

Condition:

MARIT_45 Value: MARIT_45 Label:

If AH43=4, 5 4 Divorced/Separated If AH43=6 5 Never Married If AH43=1 1 Married If AH43=2 2 Living with Partner If AH43=3 3 Widowed

GENERAL HEALTH

ASTCUR Current Asthma The ASTCUR variable is derived from questionnaire items AB40 and AB41 and is based on the diagnosis of asthma by a doctor (AB17). ASTCUR is a dichotomous variable that indicates whether or not the adult respondent currently has asthma. Of those who were diagnosed with asthma by a doctor (AB17=1), those who still have asthma (AB40=1) or have had an asthma attack or episode in the past 12 months (AB41=1) are considered to currently have asthma (ASTCUR=1). Of those who were diagnosed with asthma by a doctor (AB17=1), but do not still have asthma (AB40=2) or have not suffered from an asthma attack in the past 12 months (AB41=2) are not considered to currently have asthma (ASTCUR=2). Finally, those who were never diagnosed with asthma by a doctor (AB17=2) are not considered to currently have asthma (ASTCUR=2). Cases in which asthma status cannot be ascertained are imputed to assign a value of ASTCUR.


22

ASTS Asthma Symptoms Past 12 Mos Population Diagnosed w/ Asthma

The ASTS variable is derived from questionnaire items AB19 and AB66. This variable provides a measure of the presence of asthma symptoms in the past year for those who have been diagnosed with asthma. Cases are assigned values based on the following conditions:

Condition: ASTS Value: ASTS Label:

If AB19 = 2, 3, 4 or 5 or AB66 in 2, 3, 4 or 5

1 Symptoms

If AB19=1 or AB66=1 2 No symptoms Respondents are assigned a skip value (ASTS=-1) if they were never diagnosed with asthma (AB19=-1 and AB66=-1)

ASTYR Asthma Symptoms Past 12 Months This variable is derived from questionnaire item AB19, which measures the frequency of asthma symptoms in the past 12 months. ASTYR is a dichotomous variable that determines whether or not the adult respondent has had any asthma symptoms in the past 12 months. Adults who have not had any asthma symptoms in the past 12 months (AB19=1) are assigned a value of ASTYR=2. Adults who report having asthma symptoms daily (AB19=5), weekly (AB19=4), monthly (AB19=3) or less than monthly (AB19=2) are assigned a value of ASTYR=1. Those who have never been diagnosed with asthma (AB17=2) are assigned a value of ASTYR= (-1). Cases in which frequency of asthma symptoms cannot be determined are imputed to assign a value

AB106_P Visited ER/URGT Care for Asthma Because Unable to See Own Doctor (PUF Recode)

AB106_P is a recoded version of AB106, which identifies adult respondents with current asthma who visited the ER due to their condition and because they were unable to see their primary care doctor in the past 12 months. The variable also identifies those who do not have a personal health care provider. The construction of AB106_P from AB106 follows the same rules as AB110_P and AB116_P, whereby responses indicating, “Doesn’t Have a Doctor” are categorized under AB106 =1 for AB106_P.

AB108_P1 Confidence to Control and Manage Asthma (PUF 1-Year Recode)

AB108_P1 is a recoded version of AB108, a Likert scale measure which asks adult respondents about their confidence to control and manage their asthma. The construct combines AB108=3 and AB108=4 to create a three-level variable with the following categories: Very Confident (AB108_P1=1), Somewhat Confident (AB108_P1=2), and Not too Confident/Not at all Confident (AB108_P1=3).

AB42_P1 Workdays Missed Due to Asthma in Past 12 Months (PUF 1-Year Recode)

AB42_P1 is constructed from AB42, which is a continuous variable. The variable categorizes the number of workdays missed due to the asthma in the past 12 months into the following levels: AB42_P1 = 0 (0 Days), 1 (1-2 Days), 3 (3-5 Days), and 5 (5+ Days).


23

AB51_P1 Type I or Type II Diabetes (PUF 1-Year Recode) AB51_P1 is constructed from AB51 and recodes those respondents reporting AB51 = 3 (Another Type) as AB51_P1 = -1, due to the potentially identifiable nature of this category. The construct presents the percentage of adults with Type I or II diabetes only.

AB114_P1 Confidence to Control and Manage Diabetes (PUF 1-Year Recode)

AB114_P1 is a recoded version of AB114, a Likert scale measure which asks adult respondents about their confidence to control and manage their diabetes. The construct combines AB114=3 and AB114=4 to create a three-level variable with the following categories: Very Confident (AB114_P1=1), Somewhat Confident (AB114_P1=2), and Not too Confident/Not at all Confident (AB114_P1=3).

AB120_P1 Confidence to Control and Manage Heart Disease (PUF 1- Year Recode)

AB120_P1 is a recoded version of AB120, a Likert scale measure which asks adult respondents about their confidence to control and manage their heart disease. The construct combines AB120=3 and AB120=4 to create a three-level variable with the following categories: Very Confident (AB120_P1=1), Somewhat Confident (AB120_P1=2), and Not too Confident/Not at all Confident (AB120_P1=3).

AB23_P1 Age First Told Have Diabetes (PUF 1-Year Recode) AB23_P1 is constructed from AB23, which is a continuous variable. The construct categorizes the ages that respondents were first diagnosed with diabetes into the following ranges: AB23_P1 =1 (Ages 18 and Under), 2 (Ages 19 - 29), 3 (Ages 30 - 39), 4 (Ages 40 - 49), 5 (Ages 50 - 59), 6 (Ages 60 - 69), 7 (Ages 70 - 79), and 8 (Ages 80 and Over).

AB27_P1 Number of Times Doctor Checked for Hemoglobin A1C Last Year (PUF 1-Year Recode)

AB27_P1 is constructed from AB27, which is a continuous variable. The construct categorizes the number of times that respondents had hemoglobin A1C levels checked by a doctor in the last 12 months: AB27_P1 = 0 (0 times), 1 (1 time), 2 (2 times), 3 (3 times), 4 (4 times), and 5 (5 or more times). A separate category exists for respondents who have never heard of hemoglobin A1C/A1C testing (AB27 = 995).

AB28_P1 Number of Times Doctor Checked for Feet for Sores Last Year (PUF 1-Year Recode)

AB28_P1 is constructed from AB28, which is a continuous variable. The construct categorizes the number of times that respondents had a doctor check their feet for sores or irritations in the last 12 months: AB28_P1 = 0 (0 times), 1 (1 time), 2 (2 times), 3 (3 times), 4 (4 times), and 5 (5 or more times).

DIABETES Doctor Ever Told Have Diabetes (Non-Gestational) The DIABETES variable is constructed from questionnaire items AB22 and AB51. The purpose of this constructed variable is to determine if a respondent has ever been told by a doctor that they have diabetes. If a respondent answers 1 (Yes) to AB22 and does not answer 5 (Borderline or Pre-Diabetes) to


24

AB51, then DIABETES = 1 (Yes). Otherwise, if the respondent answers 2 (No) or 3 (Borderline or Pre-Diabetes) to AB22 or if the respondent answers 1 (Yes) to AB22 and 5 (Borderline or Pre-Diabetes) to AB51, then DIABETES = 2 (No).

DIAMED Taking Insulin or Pills The DIAMED variable is derived from questionnaire items AB24 and AB25. This variable categorizes the use of insulin and medication by adults who have been told they have diabetes (AB22=1). Values are assigned based on the following criteria:

Condition:

DIAMED Value: DIAMED Label:

If AB24=1 and AB25=1 1 Taking insulin and pills If AB24=1 and AB25=2 2 Taking insulin only If AB24=2 and AB25=1 3 Taking pills only If AB24=2 and AB25=2 4 Not taking insulin or pills

Those who have never been told they have diabetes by a doctor (AB22=2) were assigned a value of DIAMED= (-1). Finally, cases in which the use of insulin could not be ascertained are imputed to assign a value of DIAMED.

HGHTI_P Height – Inches (PUF Recode) The HGHTI_P variable is a recoded version of the HGHTI variable. Top code: 77 inches. Bottom Code: 42.

HGHTM_P Height – Meters (PUF Recode) The HGHTM_P variable is a recoded version of the HGHTM variable. Top code: 2 meters. Bottom code: 1 meter.

HEIGHM_P Height – Meters (UCLA) (PUF Recode) The HEIGHM_P variable is a recoded version of the HEIGHTM variable using the UCLA definition of conversion. It uses the top and bottom code of HEIGHTM. Top code: 2 meters. Bottom code: 1 meter.

WEIGHK_P Weight – Kilograms (UCLA) (PUF Recode) The WEIGHK_P variable is a recoded version of the WEIGHTKG variable using the UCLA definition of conversion. It uses the top code of WEIGHTKG. Top code: 150 kilos.

WGHTK_P Weight – Kilograms (PUF Recode) The WGHTK_P variable is a recoded version of the WGHTK variable.


25

Top code: 150 kilos. Bottom code: 41 kilos.

WGHTP_P Weight – Pounds (PUF Recode) The WGHTP_P variable is a recoded version of the WGHTP variable. Top code: 330 lbs. Bottom code: 90 lbs.

BMI_P Body Mass Index for Adults (PUF Recode) The BMI_P variable is continuous and is a recoded version of BMI. BMI_P is calculated with the PUF variables that represent weight (WEIGHK_P - top coded) and height (HEIGHM_P - bottom coded). If no top or bottom code is assigned for weight or height, BMI=BMI_P.

RBMI BMI Descriptive The RBMI variable is constructed based on BMI values. This variable categorizes BMI into 4 weight range groups. Values are assigned based on the following criteria:

Condition: RBMI value:

RBMI label:

If 0<=BMI<=18.49 1 Underweight If BMI>18.49 and BMI<=24.99 2 Normal If BMI>24.99 and BMI<=29.99 3 Overweight If BMI>29.99 4 Obese

Cases in which BMI cannot be determined are imputed to assign an RBMI value.

WHOBMI Body Mass Index: WHO Definition WHOBMI is a four-level categorized variable replicating the continuous CHIS variable, BMI (Body Mass Index), using definitions of BMI levels provided by WHO.

Condition:

WHOBMI Value: WHOBMI Label:

If 0<=BMI<=18.49 1 0 - 18.49 (Underweight) If 18.5<=BMI<=22.99 2 18.5 – 22.99 (Acceptable

Risk) If 23.0<=BMI<=27.49 3 23.0 – 27.49 (Increased

Risk) If 27.5<=BMI 4 27.5 or Higher (Higher

High Risk) Note: For more information on the WHO’s BMI definition levels, see http://www.who.int/nutrition/publications/bmi_asia_strategies.pdf.

OVRWT Overweight or Obese The OVRWT variable is constructed based on RBMI criteria. This variable is dichotomous and determines whether the adult respondent is considered to be physically overweight or obese. A value is assigned for


26

the final constructed variable, OVRWT, based on RBMI that determines whether the adult is overweight or obese.

Condition: OVRWT Value:

OVRWT Label:

If RBMI=3 or 4 1 Yes If RBMI=1 or 2 2 No

Cases in which RBMI cannot be determined are imputed to assign an OVRWT value.

HEALTH BEHAVIORS AE_FRIES Number of times eating fries per week The variable AE_FRIES is constructed from the variable AE3. AE3 asks about total number of times the respondent ate french fries, home fries, or hash browns in the past month. Responses of total consumption in the past month were divided by 4.33 to obtain average weekly fries consumption. The numbers were rounded up or down to the nearest whole number for number of times per week. AE_FRUIT Number of times eating fruit per week The variable AE_FRUIT is constructed from the variable AE2. AE2 asks about total number of times the respondent ate fruit in the past month. Responses of total consumption in the past month were divided by 4.33 to obtain average weekly fruit consumption. The numbers were rounded up or down to the nearest whole number for number of times per week. AE_VEGI Number of times eating vegetables per week The variable AE_VEGI is constructed from the variable AE7. AE7 asks about total number of times the respondent ate vegetables in the past month. Responses of total consumption in the past month were divided by 4.33 to obtain average weekly vegetables consumption. The numbers were rounded up or down to the nearest whole number for number of times per week. AE_SODA Number of times drank soda per week The variable AE_SODA is constructed with variable AC11 and AC11UNT. AC11 asks about total soda consumption in the past month, and the questionnaire allows for respondents to report their soda consumption per week or per day. For respondents who report daily consumption, CHIS multiplied their responses by 7 to obtain weekly consumption. For those reporting total consumption in the past month, their responses were divided by 4.33 to obtain average weekly consumption. The numbers were rounded up or down to the nearest whole number. Individuals who drank soda only one or two times in a month were rounded down to an average of 0 times of soda consumed per week.

AESODA_P1 Number of times drank soda per week (PUF 1-Year Recode)


27

The variable AESODA_P1 categorizes AE_SODA, which is a continuous variable, into the following ranges: AESODA_P1=0 (0 times), 1 (1 time), 2 (2-3 times), 3 (4-6 times), 4 (7-13 times), 5 (14-20 times), and 6 (21 or more times).

SMKCUR Current Smoker The SMKCUR variable was derived from questionnaire items AE15 and AE15A. If the adult indicated smoking every day (AE15A=1) or some of the days (AE15A=2) then the respondent was considered to be a current smoker (SMKCUR=1). If the respondent indicated never smoking more than 100 cigarettes in one’s lifetime (AE15=2) or not smoking cigarettes daily (AE15A=3) then he or she was considered to be a non-smoker (SMKCUR=2).

SMOKING Current Smoking Habits The SMOKING variable is constructed from questionnaire items AE15 and AE15A. This variable categorizes the smoking habits of the adult respondent. If an adult indicates smoking 100 or more cigarettes in his/her lifetime (AE15=1) and smokes every day (AE15A=1) or some of the days (AE15A=2) then SMOKING=1. If the adult respondent indicates smoking 100 or more cigarettes in his/her lifetime and currently does not smoke at all then SMOKING=2. If an adult respondent has never smoked 100 or more cigarettes in his/her lifetime (AE15=2), then SMOKING=3. Cases in which current smoking status cannot be determined are imputed to assign a value to SMOKING.

AE16_P1 Number of Cigarettes Per Day in Past 30 Days (PUF 1-Year Recode)

AE16_P1 is constructed from AE16, which is a continuous variable. The construct categorizes the number of cigarettes smoked per day in the past 30 days among adults who smoke every day (AE15A=1) and adults who smoke some days (AE15A=2).

AD32_P1 Number of Cigarettes Per Day (PUF 1-Year Recode) AB32_P1 is constructed from AD32, which is a continuous variable. The construct categorizes the average number of cigarettes smoked per day among adults who smoke every day (AE15A=1) and adults who smoke some days (AE15A=2).

NUMCIG Number of Cigarettes Per Day The constructed NUMCIG variable is derived from questionnaire items AE15, AE15A, AE16 and AD32. This variable categorizes the number of cigarettes smoked per day by adults who smoke every day (AE15A=1) and adults who smoke some days (AE15A=2). Values are assigned according to the following criteria.

Condition:

NUMCIG Value: NUMCIG Label:

If AE15=2 or AE15A=3 1 None (If AD32=0 or 1) or (AE16=0 or 1) 2 <=1 cigarette (If AD32>=2 and AD32<=5) or (AE16>=2 and AE16<=5)

3 2–5 cigarettes

(If AD32>=6 and AD32<=10) or (AE16>=6 and AE16<=10)

4 6–10 cigarettes

(If AD32>=11 and AD32<=19) or (AE16>=11 and AE16<=19)

5 11–19 cigarettes


28

If AD32>=20 or AE16>=20 6 20 + cigarettes Cases in which number of cigarettes cannot be determined are imputed to assign a value to NUMCIG.

AC31_P1 Number of Times Ate Fast Food in the Past Week (PUF 1- Year Recode)

AC31_P1 is a top coded version of questionnaire item AC31. Note: Top code is 9

AC34_P1 Number of Times Male Had 5+ Drinks in Past 12 Months (PUF 1-Year Recode)

AC34_P1 is constructed from AC34, which is a continuous variable. The construct categorizes the number of times that male respondents had 5 or more alcoholic drinks in a single day over the last 12 months into the following levels: AB34_P1 = 0 (0 days), 1 (1 day), 2 (2 days), 3 (3 days), 4 (4 days), 5 (5 days), 6 (6 days), 7 (7 days), 8 (8-10 days), 11 (11-14 days), 15 (15-19 days), 20 (20-24 days), 25 (25-29 days), and 30 (30 or more days).

AC35_P1 Number of Times Female Had 4+ Drinks in Past 12 Months (PUF 1-Year Recode)

AC35_P1 is constructed from AC35, which is a continuous variable. The construct categorizes the number of times that female respondents had 4 or more alcoholic drinks in a single day over the last 12 months into the following levels: AB35_P1 = 0 (0 days), 1 (1 day), 2 (2 days), 3 (3 days), 4 (4 days), 5 (5 days), 6 (6 days), 7 (7 days), 8 (8-10 days), 11 (11-14 days), 15 (15-19 days), 20 (20-24 days), 25 (25-29 days), and 30 (30 or more days).

AC42_P How Often Find Fresh Fruit/ Veg. in Neighborhood (PUF Recode)

The AC42_P variable is a recoded version of AC42. This variable provides a Likert measure of how often the respondent is able to find fresh fruits and vegetables in his/her neighborhood (always, usually, sometimes, never). Cases were collapsed into one single category if no fruits/vegetables are consumed, the respondent does not shop for fresh fruits/ vegetables or the respondent does not shop for fresh fruits/vegetables in his/her neighborhood (AC42_P=5).

AC47_P1 Number of Classes of Water Drank Yesterday (PUF 1-Year Recode)

AC47_P1 is a top coded version of questionnaire item AC47. Note: Top code is 20. AC47_P1= 99 refers to less than 1 glass of water.

AC48_P1 Number of Classes of Low-Fat or Non-Fat Milk Drank Yesterday (PUF 1-Year Recode)

AC48_P1 is a top coded version of questionnaire item AC48. Note: Top code is 20. AC48_P1= 99 refers to less than 1 glass of milk.


29

AJ143_P1 Place Received Main Birth Control Method or Prescription (PUF 1-Year Recode)

The AJ143_P1 variable is a recoded version of questionnaire item AJ143.

Condition:

AJ143_P1 Value: AJ143_P1 Label:

If AJ143=1 or 2 1 Private Doctor’s Office or HMO Facility

If AJ143=3 3 Hospital or Hospital Clinic If AJ143=4 or 5 4 Planned Parenthood or County

Health Department If AJ143=6,7,8,9, or 91 6 Some Other Place

AJ146_P1 Place Received Main Birth Control Method (PUF 1-Year Recode) The AJ146_P1 variable is a recoded version of the questionnaire item AJ146.

Condition: AJ146_P1 Value: AJ146_P1 Label:

AJ146= 1 or 2 1 Private doctor’s office or HMO facility AJ146= 3 3 Hospital or hospital clinic AJ146= 4 or 5 4 Planned Parenthood or county health

department AJ146= 6, 9, or 91 6 Other

AC53NUM_P1 How Long Since Smoked on Daily Basis (PUF 1-Year Recode) AC53NUM_P1 is a ten-level, categorical variable constructed from questionnaire item AC53. AC53 is a continuous variable that displays how long it has been since respondent smoked on a daily basis if respondent stated they had smoked regularly with at least 100 cigarettes smoked in their lifetime.

AC53NUM_P1 Value:

AC53NUM_P1 Label:

1 1 MONTH OR LESS 2 MORE THAN 1 MONTH TO 1 YEAR 3 2-3 YEARS 4 4-5 YEARS 5 6-10 YEARS 6 11-20 YEARS 7 21-30 YEARS 8 31-40 YEARS 9 41-50 YEARS 10 MORE THAN 50 YEARS Note: Category value of “999” is for respondents in the universe of AC53 who responded that they never smoked on a regular basis.


30

AC54NUM_P1 How Soon After Awaking Smokes 1st Cigarette (PUF 1-Year Recode)

AC54NUM_P1 is a nine-level, categorical variable constructed from questionnaire item AC54. AC54 is a continuous variable that displays how soon after respondent awakes that they usually smoke their first cigarette, if respondent stated they smoke daily or some days per week, with at least 100 cigarettes smoked in their lifetime.

AC54NUM_P1 Value:

AC54NUM_P1 Label:

1 5 MINUTES OR LESS 2 6 TO 10 MINUTES 3 11 TO 20 MINUTES 4 21 TO 30 MINUTES 5 31 MINUTES TO 1 HOUR 6 MORE THAN 1 HOUR TO 2 HOURS 7 MORE THAN 2 HOURS TO 4 HOURS 8 MORE THAN 4 HOURS TO 10 HOURS 9 MORE THAN 10 HOURS Note: Category value of “999” is for respondents in the universe of AC54 who responded that they do not smoke after waking up.

AC55_P1 Where Usually Buy Cigarettes (PUF 1-Year Recode) AC55_P1 is a six-level variable constructed from questionnaire item AC55:

Condition:

AC55_P1 Value: AC55_P1 Label:

If AC55=1 1 Convenience Stores or Gas Stations

If AC55=2 2 Super Markets If AC55=3 3 Liquor Stores or Drug Stores If AC55=4 4 Tobacco Discount Stores If AC55=5 5 Other Discount or

Warehouse Stores If AC55=6, 7, 8, or 9 6 Somewhere Else or No

Usual Place Note: Category value of “99” is for respondents in the universe of AC55 who responded that they do not buy cigarettes.

AC80_P1 Use Hookah Pipe Every Day (PUF 1-Year Recode) AC80_P1 is a two-level variable constructed from questionnaire item AC80:

Condition:

AC80_P1 Value: AC80_P1 Label:

If AC80=1 or 2 1 Every Day/Some Days If AC80=3 2 Not At All


31

AC82_P1 Days Used E-Cigs in Past 30 Days (PUF 1-Year Recode) AC82_P1 is a five-level, categorical variable constructed from questionnaire item AC82. AC82 is a continuous variable that displays during the past 30 days how many days did respondent use electronic cigarettes, if respondent stated that they have ever smoked electronic cigarettes.

AC82_P1 Value:

AC82_P1 Label:

1 0 DAYS 2 1 TO 2 DAYS 3 3 TO 5 DAYS 4 6 TO 29 DAYS 5 EVERY DAY

AC52_P1 Age Began Smoking Cigarettes Regularly (PUF 1-Year Recode) AC52_P1 is a bottom coded and top coded version of questionnaire item AC52. The question is asked of respondents who respond that they smoke cigarettes daily or on some days per week and who have smoked at least 100 cigarettes in their lifetime. Note: Bottom code is 7; top code is 40

AJ142_P1 Main Type of Birth Control Received From Doctor in Past Year (PUF 1-Year Recode)

AJ142_P1 is a four-level variable constructed from questionnaire item AJ142:

Condition:


If AJ142=1, 2, 4, 7, 8, or 91 91 Other Birth Control and Hysterectomy

If AJ142=3 3 IUD (Mirena Paragard) If AJ142=5 5 Birth Control Pills If AJ142=6 6 Other Hormonal Methods

(Injection/Depo-Provera)

PREDIAB Doctor Ever Told Have Pre- or Borderline Diabetes (Non Gestational)

PREDIAB is a two-level variable constructed from the variables AB22 (Doctor Ever Told Have Diabetes), AB51 (Type 1 or Type 2 Diabetes), and AB99 (Doctor Ever Told Have Pre- or Borderline Diabetes). If AB22=3 (Borderline or Pre-Diabetes) or AB99=1 (Yes) or AB51=5 (Borderline or Pre-Diabetes), then PREDIAB=1 (Yes). If AB99=2 (No) then PREDIAB=2 (No). Otherwise, PREDIAB=-9 (Not Ascertained).


32

MENTAL HEALTH DISTRESS Serious Psychological Distress Past Month The DISTRESS variable is a continuous measure of generalized psychological distress during the past month using the Kessler 6-Item Psychological Distress Scale (K6). It is created with questionnaire items AJ29, AJ30, AJ31, AJ32, AJ33, and AJ34. Items are reverse coded so that cases with a greater frequency of symptoms receive higher scores. Scores are assigned based on the following criteria:

The DISTRESS value is the total of the assigned scores for items AJ29 to AJ34. The maximum value is 24 and the minimum is 0.

DSTRS_P1 Serious Psychological Distress Past Month (PUF 1-Year Recode) The DISTRESS variable is a continuous measure of generalized psychological distress during the past month using the Kessler 6-Item Psychological Distress Scale (K6). It is created with questionnaire items AJ29, AJ30, AJ31, AJ32, AJ33, and AJ34. Items are reverse coded so that cases with a greater frequency of symptoms receive higher scores. Scores are assigned based on the following criteria:

The DISTRESS value is the total of the assigned scores for items AJ29 to AJ34. While the maximum value is 24 and the minimum is 0, the PUF 1-Year recode version of this variable top codes the maximum value at 21 for data confidentiality reasons Note: Top code is 21.

DSTRS12 Likely Has Psychological Distress Past Year The DSTRS12 variable indicates if the respondent likely has psychological distress during the past year, based on the value of DSTRSYR. The assignment of value follows this manner:

Condition: DSTRS12 Value: DSTRS12 Label: If DSTRSYR>=13 1 Yes If 0<=DSTRSYR<13 2 No

Value (AJ29- AJ34)

Assigned score

1 (all of the time) 4 2 (most of the time) 3 3 (some of the time) 2 4 (a little of the time) 1 5 (not at all) 0

Value (AJ29- AJ34)

Assigned score

1 (all of the time) 4 2 (most of the time) 3 3 (some of the time) 2 4 (a little of the time) 1 5 (not at all) 0


33

DSTRS30 Likely Has Psychological Distress Past Month The DSTRS30 variable indicates if the respondent likely has psychological distress during the past month, based on the value of DISTRESS. The assignment of value follows this manner:

Condition: DSTRS30 Value: DSTRS30 Label: If DISTRESS>=13 1 Yes If 0<=DISTRESS<13 2 No

DSTRSYR Serious Psychological Distress for Worst Month in Past Year The DSTRSYR variable measures psychological distress for worst month in past year using responses from two sets of questionnaire items, AJ29-AJ34, and AF63-AF68. A temporary reverse value of each questionnaire item is first assigned in the following manner:

Condition: (same applies to AJ30-AJ34, AF63-AF68)

AJ29temp Value:

If AJ29=5 0 If AJ29=4 1 If AJ29=3 2 If AJ29=2 3 If AJ29=1 4

The sum of the temporary variables AJ29temp-AJ34temp is compared with that of the temporary variables AF63temp-AF68temp. The higher value will be assigned as DSTRSYR.

CHORES2 Chore Impairment Past 12 Months CHORES2 is similar to CHORES from previous CHIS cycles, but is not trendable with CHORES due to changes in the input variable, AF70B. Similar to CHORES, CHORES2 identifies whether a respondent has had any impairment completing chores in the past year due to his/her emotions. The variable assigns reverse scores using AF70B so that higher scores reflect greater impairment. Scores are assigned based on the following criteria:

Condition: CHORES2 Value:

CHORES2 Label:

AF70=3 0 None AF70=2 1 Moderate AF70=1 2 Severe

SOCIAL2 Social Life Impairment Past 12 Months SOCIAL2 is similar to SOCIAL from previous CHIS cycles, but it is not trendable with SOCIAL due to changes in the input variable, AF71B. Similar to SOCIAL, SOCIAL2 identifies whether a respondent has had any social impairment in the past year due to his/her emotions. The variable assigns reverse scores using AF71B so that higher scores reflect greater impairment. Scores are assigned based on the following criteria:

Condition: SOCIAL2Value: SOCIAL2 Label:


34

AF71=3 0 None AF71=2 1 Moderate AF71=1 2 Severe

WORK2 Work Impairment Past 12 Months WORK2 is similar to WORK from previous CHIS cycles, but is not trendable with WORK due to changes in the input variable, AF69B. Similar to WORK, WORK2 identifies whether a respondent has had any work impairment in the past year due to his/her emotions. The variable assigns reverse codes scores using AF69B so that higher scores reflect greater impairment. Scores are assigned based on the following criteria:

Condition: WORK2 Value: WORK2 Label: AF69=3 0 None AF69=2 1 Moderate AF69=1 2 Severe AF69=4 (does not work) -1 Not applicable

Respondents who are older than 70 years of age were skipped and are assigned the value, WORK= (-1).

FAMILY2 Family Life Impairment Past 12 Months FAMILY2 is similar to FAMILY from previous CHIS cycles, but is not trendable with FAMILY due to changes in the input variable, AF72B. Similar to FAMILY, FAMILY2 identifies whether a respondent has had any impairment in family life within the past year due to his/her emotions. The variable assigns reverse scores based on AF72B, so that higher scores reflect greater impairment. Scores are assigned based on the following criteria:

Condition: FAMILY2 Value: FAMILY2 Label: AF72B=3 0 None AF72B=2 1 Moderate AF72B=1 2 Severe

ADDITIONAL DEMOGRAPHIC INFORMATION

AH33NEW Born in U.S. The AH33NEW variable is derived from source variable AH33. It dichotomizes AH33 responses into two categories indicating whether or not the respondent was born in the United States or one of its territories. Respondents who were born in U.S. or one of its territories (i.e. American Samoa, Guam, Puerto Rico, and the Virgin Islands) are coded as having been born in the U.S. (AH33NEW=1). All other responses are coded as having been born outside the U.S. (AH33NEW=2). AH34NEW Mother Born In U.S. The AH34NEW variable is derived from source variable AH34. It dichotomizes AH34 responses into two categories indicating whether or not the respondent’s mother was born in the United States or one of its territories. Respondents whose mother was born in U.S. or one of its territories (i.e. American Samoa, Guam, Puerto Rico, and the Virgin Islands) are coded as having been born in the U.S. (AH34NEW=1). All other responses are coded as having been born outside the U.S. (AH34NEW=2).


35

AH35NEW Father Born In U.S. The AH35NEW variable is derived from source variable AH35. It dichotomizes AH35 categories into two categories indicating whether or not the respondent’s father was born in the United States or one of its territories. Respondents whose father was born in U.S. or one of its territories (i.e. American Samoa, Guam, Puerto Rico, and the Virgin Islands) are coded as having been born in the U.S. (AH35NEW=1). All other responses are coded as having been born outside the U.S. (AH35NEW=2).

LANGHOME Types of Languages Spoken at Home The LANGHOME variable indicates the languages spoken in the homes of the respondents. This variable is derived from questionnaire items AH36_1 - AH36_22. The variable considers households in which multiple languages are spoken. The LANGHOME variable is created with the categories generated with the LANGTEMP construct variable.

1. First, values are assigned to a LANGTEMP variable based on criteria from questionnaire items AH36_1 – AH36_22.

Condition: LANGTEMP value:

LANGTEMP label:

If AH36_1=1 1 English If AH36_2=1 2 Spanish If AH36_3=1 or AH36_6=1 3 Chinese If AH36_4=1 4 Vietnamese If AH36_5=1 5 Tagalog If AH36_7=1 6 Korean If AH36_8=1 7 Asian Indian Languages If AH36_9=1 8 Russian If AH36_12=1 9 Japanese If AH36_13=1 10 Other Asian Language If AH36_14=1 or AH36_15=1 or AH36_16=1

11 European

If AH36_18=1 12 Farsi If AH36_19=1 13 Armenian If AH36_20=1 14 Arabic If AH36_21=1 15 African/Afro-Asiatic If AH36_22=1 16 Other language

2. For cases in which respondents speak two or more languages, values for LANGTEMP were

assigned based on the following criteria:

If AH36_1=1 and AH36_2=1 17 English and Spanish If AH36_1=1 and (AH36_3=1 or AH36_6=1)

18 English and Chinese

If AH36_1=1 and (AH36_9=1 or AH36_14=1 OR AH36_15=1 OR AH36_16=1 or AH36_17=1)

19 English and European language

If AH36_1=1 and (AH36_4=1 OR AH36_5=1 or AH36_7=1 OR AH36_8=1 or AH36_13=1)

20 English and another Asian Language


36

If sum of AH36_1_22=2 and AH36_1=1

21 English and one other language

If sumAH36_1-22>=2 22 Other two or more languages/ three or more languages

3. Next, each case is tested through the following steps until a LANGHOME value can be assigned:

Condition: LANGHOME

Value:

LANGHOME Label:

If LANGTEMP=1 1 English If LANGTEMP=2 2 Spanish If LANGTEMP=3 3 Chinese If LANGTEMP=4 4 Vietnamese If LANGTEMP=6 5 Korean If LANGTEMP= Tagalog (5) Asian Indian Languages (7) Japanese (9) Other Asian language (10)

6 Other one Asian language

If LANGTEMP= Russian (8) European (11) Farsi (12) Armenian (13) Arabic (14) African/Afro-Asiatic (15) Other language (16)

7 Other one language only

If LANGTEMP=17 8 English and Spanish If LANGTEMP=18 9 English and Chinese If LANGTEMP=19 10 English and a European

language If LANGTEMP=20 11 English and another Asian

language If LANGTEMP=21 12 English and one other language If LANGTEMP=22 13 Other two or more languages

LNGHM_P1 Types of Languages Spoken at Home (PUF 1-Year Recode) The LNGHM_P1 variable indicates the languages spoken in the homes of the respondents. This variable is derived from questionnaire items AH36_1 - AH36_22 and construct LANGHOME. The variable considers households in which multiple languages are spoken. The LNGHM_P1 variable is created with the categories generated with the LANGTEMP construct variable.

4. First, values are assigned to a LANGTEMP variable based on criteria from questionnaire items AH36_1 – AH36_22.

Condition: LANGTEMP

value:

LANGTEMP label:

If AH36_1=1 1 English If AH36_2=1 2 Spanish If AH36_3=1 or AH36_6=1 3 Chinese If AH36_4=1 4 Vietnamese


37

If AH36_5=1 5 Tagalog If AH36_7=1 6 Korean If AH36_8=1 7 Asian Indian Languages If AH36_9=1 8 Russian If AH36_12=1 9 Japanese If AH36_13=1 10 Other Asian Language If AH36_14=1 or AH36_15=1 or AH36_16=1

11 European

If AH36_18=1 12 Farsi If AH36_19=1 13 Armenian If AH36_20=1 14 Arabic If AH36_21=1 15 African/Afro-Asiatic If AH36_22=1 16 Other language

5. For cases in which respondents speak two or more languages, values for LANGTEMP were

assigned based on the following criteria:

If AH36_1=1 and AH36_2=1 17 English and Spanish If AH36_1=1 and (AH36_3=1 or AH36_6=1)

18 English and Chinese

If AH36_1=1 and (AH36_9=1 or AH36_14=1 OR AH36_15=1 OR AH36_16=1 or AH36_17=1)

19 English and European language

If AH36_1=1 and (AH36_4=1 OR AH36_5=1 or AH36_7=1 OR AH36_8=1 or AH36_13=1)

20 English and another Asian Language

If sum of AH36_1_22=2 and AH36_1=1

21 English and one other language

If sumAH36_1-22>=2 22 Other two or more languages/ three or more languages

6. Next, each case is tested through the following steps until a LNGHM_P1 value can be assigned:

Condition: LNGHM_P1

Value:

LNGHM_P1 Label:

If LANGTEMP=1 1 English If LANGTEMP=2 2 Spanish If LANGTEMP=3 3 Chinese If LANGTEMP=4 4 Vietnamese If LANGTEMP=6 5 Korean If LANGTEMP= Tagalog (5) Asian Indian Languages (7) Japanese (9) Other Asian language (10) Russian (8) European (11) Farsi (12) Armenian (13) Arabic (14) African/Afro-Asiatic (15) Other language (16)

6 Other one language


38

If LANGTEMP=17 8 English and Spanish If LANGTEMP=18 9 English and Chinese If LANGTEMP=19 10 English and a European

language If LANGTEMP=20 11 English and another Asian

language If LANGTEMP=21 12 English and one other language If LANGTEMP=22 13 Other two or more languages

SPK_ENG English Use and Proficiency The SPK_ENG variable is derived from questionnaire items AH36 and AH37. This variable measures the strength and use of the English language by the adult respondent. Adults who indicate speaking only English (AH36_1=1) were assigned a value of SPK_ENG=1. Of those who speak another language, those who indicate speaking English very well (AH37=1) or well (AH37=2) were assigned a value of SPK_ENG=2. Of those who speak another language, those who indicate speaking English not well (AH37=3) or not at all (AH37=4) were assigned a value of SPK_ENG=3. Cases for which the use of English cannot be determined are imputed to assign a value to SPK_ENG. CITIZEN2 Citizenship Status for Adults The CITIZEN2 variable collapses green card status and is created in order to provide another indication of citizenship. This variable also reflects a definition from UCLA’s Center for Health Policy Research. CITIZEN2 is derived from questionnaire items AH33, AH39 and AH40. Each case is tested through the following conditions until a CITIZEN2 value is assigned:

Condition: CITIZEN2 Value: CITIZEN2 Value:

If AH33=1, 2, 9, 22, or 26 (respondent reports that they were born in the U.S., American Samoa, Guam, Puerto Rico, or the Virgin Islands

1 U.S.-Born Citizen

If AH39=1 (respondent reports that they are U.S. citizens, but were not born in the U.S., American Samoa, Guam, Puerto Rico, or Virgin Islands

2 Naturalized Citizen

If AH40=1, 2, or 3 (permanent resident with green card, not permanent resident, or have application pending)

3 Non-Citizen

YRUS_P1 Years Lived in the U.S. (PUF 1-Year Recode) YRUS_P1 is constructed with questionnaire item AH41. YRUS_P1 assigns the number of years the respondent has lived in the U.S. (AH41) into 4 range levels. In addition, YRUS_P1 standardizes the number of years for those who reported a particular year.


39

The value for this variable is calculated for the respondents who report a particular year by subtracting the year they report from the current survey year (e.g., 2017 - AH41). A skip value (-1) is assigned for all persons who were born in the U.S., Guam, Samoa, or the Virgin Islands.

PCTLF_P Percent Life in U.S. (PUF Recode) The PCTLF_P is a recode of the continuous construct variable PCTLF. This variable provides categorical measure of the percentage of the adult respondent’s life spent in the United States. Values are assigned as follows:

Condition:

PCTLF_P Value:

PCTLF_P Label:

If 0<=PCTLF<=20 1 0%-20% If 21<=PCTLF<=40 2 21%-40% If 41<=PCTLF<=60 3 41%-60% If 61<=PCTLF<=80 4 61%-80% If 81<=PCTLF 5 81%+

WRKST Working Status This variable provides a categorical measure of current working status for adults and is derived from questionnaire items AK1, AK2, AK3, and AG10. WRKST values are assigned in the following hierarchical manner:

1. Respondents who currently/usually work per week, were working at a job last week, or were looking for work last week were assigned values based on the following criteria:

Condition: WRKST Value: WRKST Label: If AK3>=21 1 Full-time employed (21+

hours/week) If 0<AK3<=20 2 Part-time employed (0-20

hours/week) If AK1=1 and AK3=0 3 Employed but not at work If AK1=3 or AG10=3 4 Unemployed, looking for

work.

2. All cases that reported having a job but were not at work last week were coded as being unemployed and looking for work:

Condition: WRKST Value: WRKST Label: AK1=2 and AK3=0 3 Employed, not at work

3. Respondents who reported not working at a job or business last week, usually work, and were

either laid off/on strike or on family or maternity leave were also coded as being unemployed and looking for work:


40

Condition: WRKST Value: WRKST Label: If AK1=4 and AK2 in (8 or 9) and AG10=1

4 Unemployed, and looking for work

4. Respondents who reported not working at a job last week due to reasons other than being laid off

or maternity leave were coded as unemployed and not looking for work:

Condition: WRKST Value: WRKST Label: If AK1=4 5 Unemployed, and not looking

for work Cases in which current working status could not be ascertained are imputed to assign a value to WRKST.

WRKST_P1 Working Status (PUF 1-Year Recode) This variable provides a categorical measure of current working status for adults and is derived from questionnaire items AK1, AK2, AK3, and AG10. WRKST_P1 values are assigned in the following hierarchical manner:

5. Respondents who currently/usually work per week, were working at a job last week, or were looking for work last week were assigned values based on the following criteria:

Condition: WRKST_P1 Value: WRKST_P1 Label: If AK3>=21 1 Full-time employed (21+


hours/week) If AK1=1 and AK3=0 3 Other Employed If AK1=3 or AG10=3 4 Unemployed, looking for work.

6. All cases that reported having a job but were not at work last week were coded as being

unemployed and looking for work:

Condition: WRKST_P1 Value: WRKST_P1 Label: AK1=2 and AK3=0 3 Employed, not at work

7. Respondents who reported not working at a job or business last week, usually work, and were

either laid off/on strike or on family or maternity leave were also coded as being unemployed and looking for work:

Condition: WRKST_P1 Value: WRKST_P1 Label: If AK1=4 and AK2 in (8 or 9) and AG10=1

4 Unemployed and looking for work

8. Respondents who reported not working at a job last week due to reasons other than being laid off

or maternity leave were coded as unemployed and not looking for work:

Condition: WRKST_P1 Value: WRKST_P1 Label: If AK1=4 5 Unemployed and not looking

for work


41

Cases in which current working status could not be ascertained are imputed to assign a value to WRKST_P1.

WRKST_S Spouse Working Status The WRKST_S variable is derived from questionnaire items AK20, AG8, AG11, and AH43. This variable categorizes the working status of the adult respondent’s spouse. Values are assigned based on the following criteria:

Condition: WRKST_S Value: WRKST_S Label: If AK20>=21 1 Full-time employed (21+


hours/week) If AG8=(1 or 2) and AK20=0 3 Employed but not at work If AG8=3 or AG11=3 4 Unemployed, looking for work If AG8=4 5 Unemployed, not looking for

work Respondents with no spouse (AH43~=1) are assigned an inapplicable value of WRKST_S= (-1). Cases for which current working status of the spouse could not be ascertained are imputed to assign a value to WRKST_S.

AHEDC_P1 Educational Attainment (PUF 1-Year Recode) AHEDC_P1 is constructed with questionnaire item AH47. AHEDC_P1 is constructed by combining values in AH47 in order to create more general categories for education levels. The values for the educational attainment variable are assigned in the following manner: Condition:

AHEDC_P1 Value: AHEDC_P1 Label:

If AH47=1, 2, 3, 4, 5, 6, 7, 8 (grades), or 30 (no formal education

1 Grade 1 though 8

If AH47=9, 10, or 11 (grades) 2 Grade 9 through 11 If AH47=12 (grade) 3 Grade 12/HS diploma If AH47=13, 14, 15, or 22 4 Some college If AH47=24, 25, or 26 5 Vocational school If AH47=23 6 AA or AS degree If AH47=16 or 17 (4th or 5th year at university)

7 BA or BS degree

If AH47=18 8 Some grad school If AH47=19 or 20 9 MA or MS degree If AH47=21 10 PhD or equivalent

AHEDUC Educational Attainment AHEDUC is constructed with questionnaire item AH47.


42

AHEDUC is constructed by combining values in AH47 in order to create more general categories for education levels. The values for the educational attainment variable are assigned in the following manner: Condition:

AHEDUC Value: AHEDUC Label:

If AH47=1, 2, 3, 4, 5, 6, 7, 8 (grades), or 30 (no formal education

1 Grade 1 though 8

If AH47=9, 10, or 11 (grades) 2 Grade 9 through 11 If AH47=12 (grade) 3 Grade 12/HS diploma If AH47=13, 14, 15, or 22 4 Some college If AH47=24, 25, or 26 5 Vocational school If AH47=23 6 AA or AS degree If AH47=16 or 17 (4th or 5th year at university)

7 BA or BS degree

If AH47=18 8 Some grad school If AH47=19 or 20 9 MA or MS degree If AH47=21 10 PhD or equivalent

AKWKLNG Time at Main Job AKWKLNG is constructed with AK7 and AK7UNT. AKWKLNG standardizes the measurement unit of questionnaire item AK7 into years on the job in decimals. The respondents who report the length of time they have worked at their main job in months have their answers converted to years by dividing by 12 and rounding to the nearest whole number. This variable presents the values in range levels. Those who are not employed or who usually work zero hours are assigned a skip value (-1) for this variable. Any remaining cases are imputed to assign an AKWKLNG value.

SERVED Length of Time Served in Active Duty SERVED is constructed with AG22-AG24. The responses provided by these questions are first coded into an intermediate variable, SERVEMON, which standardizes the responses in months of active duty. AG23 responses are converted into months by subtracting the start date from the end date of active duty and then multiplying by 12. AG24 responses given in years are converted to months by multiplying the response by 12. The continuous responses for SERVEDMON are then coded into to time intervals for SERVED.

FAMTYP_P Family Type (PUF Recode) The FAMTYP_P is a recoded version of FAM_TYPE, which is constructed using a number of questionnaire items that measure marital status, age of the respondent and parenthood. This variable is constructed for purposes of determining eligibility for public-based health care programs. If a respondent is not married (AH43~=1), is older than 21 years (SRAGE>=21) and has no children, that case is assigned a value of FAMTYP_P =1. Similarly, for those 21 and older who are married (AH43=1) but do not live with their spouse in the same household (AH44=2), the same value (FAMTYP_P =1) is also assigned. The same logic holds true for adult respondents ages 19 and 20 years old (FAMTYP_P =2). Respondents who are married (AH43=1) and live with their spouse in the same household (AH44=1) but have no children are assigned the value of FAMTYP_P =3. Respondents who are married (AH44=1)


43

and live with their spouse (AH44=1) and have one child or more or are co-parents of one child or more are assigned the value of FAMTYP_P =4. A respondent is considered to be single with kids (FAMTYP_P =5) if he/she is not currently married (AH43~=1) or is married but does not live in the same household (AH44=2), and has one child or more, excluding co-parentship. Finally, a respondent is considered to be 18 years old and single (FAMTYP_P =6) if he/she does not have any children, is not married (AH43~=1) or is married but does not live in the same household as the spouse (AH44=2) and is age 18. FAMT4 Family Type – 4 Levels FAMT4 is a recoded version of FAM_TYPE, which is constructed using a number of questionnaire items that measure marital status, age of the respondent, and parenthood. This variable displays a 4-level version of the six-category FAM_TYPE variables. Values for FAMT4 are determined as follows:

Condition:

FAMT4 Value: FAMT4 Label:

If FAM_TYPE = 1 or 2 or 6 1 Single, No Kids If FAM_TYPE = 3 2 Married, No Kids If FAM_TYPE = 4 3 Married with Kids If FAM_TYPE = 5 4 Single with Kids

FAMSIZE2_P1 Family Size Supported by Household Income (PUF 1-Year recode)

The FAMSIZE2_P1 variable is a recoded version of FAMSIZE2_P, which assigns family size (including all supported by household income) to the respondent. Note: Top code: 8

AK33_P1 Number of Persons in U.S. Supported by Household Income Not Currently Residing in Household (PUF 1-Year Recode)

AK33_P1 is a top coded version of questionnaire item AK33. Note: Top code is 3.

HEALTH INSURANCE

COVRDCA Private/Employer Based Coverage from Covered California The COVRDCA variable is a construct that summarizes Covered California participation and is primarily based on questionnaire items, AH104 and AH105. These questions ask individuals with private health coverage if they had directly purchased their plans from an insurance company or HMO or from Covered California (AH104=2) or individuals with employer sponsored coverage whether they signed up for their health insurance through an employer, union, or through Covered California’s Small Business Health Options Program/SHOP (AH105=2). Respondents who identify themselves as being covered through premium payment for a health plan (if AH56_12=1) are considered to be covered for this variable (COVRDCA=1). COVRDCA also captures individuals who reported Covered California or SHOP health coverage through a health insurance plan that was missed (AH19_12=1 or AI19_13=1).


44

Those who report that they do not have coverage through Covered California are coded as COVRDCA=2. This includes currently uninsured individuals. Note 1: The COVRDCA_S variable measures spouse’s participation in Covered California. Construction of this variable uses the same logic as COVRDCA using items AH104=2, AI19A=1, AH108=1, AH109, AI47_12=1, AI47_13=1, AI49_12=1, and AI49_13=1. Adjustment: Individuals who report AH106=5 are recategorized into Medi-Cal.

AI45_P1 Main Reason Spouse Not in Employer Health Plan (PUF 1- Year Recode)

The AI45_P1 is a recoded version of questionnaire item AH45.

Condition:

AI45_P1 Value: AI45_P1 Label:

If AI45=4, 5, 6, 7, or 91 4 Other Reason If AI45=1 1 Covered by Another Plan If AI45=2 2 Too Expensive If AI45=3 3 Didn’t Like Plan Offered

Some of the reasons captured in A145=4 include respondents reporting that they don’t need or believe in health insurance, that they are in the process of getting insurance, or that they receive extra pay/benefits by not enrolling.

AH45A_P1 Main Reason Spouse Ineligible for Employer Health Plan (PUF 1-Year Recode)

The AI45A_P1 is a recoded version of questionnaire item AH45A.

Condition:

AI45A_P1 Value: AI45A_P1 Label:

If AI45A=4, 5, 6, 7, or 91 4 Other Reason If AI45A=1 1 Haven’t Worked Here Long Enough If AI45A=2 2 Contract or Temp Employee If AI45A=3 3 Don’t Work Enough Hours Per Week

Some of the reasons captured in A145A=4 include respondents reporting that they don’t meet age requirements or that they have left their position, retired, or were laid off.

AH52_P1 Sign for Medicare HMO Directly or Otherwise (PUF 1-Year Recode)

The AH52_P1 is a recoded version of questionnaire item AH52:

Condition:

AH52_P1 Value: AH52_P1 Label:

If AH52=7 or 8 7 Spouse’s Employer/Spouse’s Union If AH52=91, 5, or 9 8 Other If AH52=1 1 Directly If AH52=2 2 Current Employer


45

If AH52=3 3 Former Employer If AH52=4 4 Union If AH52=6 6 AARP

HMO HMO Status The HMO variable is constructed from the constructed variable INSTYPE (Type of Current Health Coverage Source for all Ages) and several questionnaire items: AH49 (Medicare Coverage Provided through HMO), AI22C (Main Health Insurance is HMO), and AI22A (Name of Main Health Plan). Respondents were coded as Yes for HMO status under three conditions: 1) Are covered by a Medicare coverage provided through HMO or have HMO as a main health insurance; 2) Respondents who were skipped out of AH49 and AI22C, but identified as having Medi-Cal as a main health plan or Medicare HMO plan; and, 3) Respondents who identified Healthy Families as the current health coverage source. HMO values are assigned as follows:

Condition:

HMO Value: HMO Label:

If INSTYPE=1 3 Uninsured AH49=1 or AI22C=1 1 HMO (AH49=-1 and AI22C=-1) and (AI22A=38 or 54)

1 HMO

INSTYPE=6 1 HMO AH49=2(no) or AI22C=2(no) 2 Non-HMO All other conditions -9 Not Ascertained

Note: Due to the elimination of Healthy Families in late 2013, the third condition listed above is not applicable in CHIS 2014 and beyond.

AI22A_P Name of Health Plan (PUF Recode) The AI22A_P is a recoded version of questionnaire item AI22A, indicating the name of the respondent’s reported health plan. This variable collapses reported health plans into more general health insurance plans. It is not asked of respondents who have Medicare HMO plans.

AI25NEW RX Coverage Edited for Medi-Cal/HF The AI25NEW variable is constructed from source variables AI25 and INSMD. AI25NEW indicates whether or not respondents who are covered by Medi-Cal are covered for prescription drugs. Please note that due to the elimination of Healthy Families in late 2013, the Healthy Families coverage mentioned in previous years is not applicable in CHIS 2014 and beyond. IHS Covered by Indian Health Services The IHS variable is derived from questionnaire items AI19_8 and AI20. Respondents who report that the type of health coverage they have is through the Indian Health Service, Tribal Health Program, or Urban Indian Clinic (if AI19_8=1 or AI20=1) are considered to be covered for this variable (IHS= 1). Those who report that they are not covered by the Indian Health Service, Tribal Health Program, or Urban Indian Clinic (if AI20=2), are considered to be not covered for this variable (IHS= 2). Note: Only the respondents who report that they are American Indian/Alaska Native (if AA5A_3=1) are asked item AI20.


46

The IHS_S variable measures whether or not the adult respondent’s spouse is covered by Indian Health Services. Construction of this variable uses the same logic as IHS using items AI47_8 and AI49_8.

INS Currently Insured This variable indicates the current insurance status of the respondent. INS is created with other constructed insurance variables. Cases that are assigned a value of 1 (covered) for any of the following variables are considered to be currently insured (INS=1): INSMC, INSMD, INSEM, INSPR, INSML, INSOG, and INSOT. The cases assigned a value of 2 (not covered) for all of those variables are considered to not be currently insured (INS=2). Note: The INS_S variable provides a dichotomous measure of the respondent’s spouse’s current insurance status. Construction of this variable uses the same logic as INS.

INS12M Number of Months Covered by Health Insurance in Past 12 Months

This variable indicates the number of months a respondent has been insured during the past 12 months. The INS12M variable is derived from items AI31, AI34, AI35, AI27, and AI29. Each case is tested through the following series of conditions until a value for INS12M can be assigned:

1. The INS12M values are first assigned to the cases with respondents who report that they have current health coverage during the administration of the questionnaire:

Condition:

INS12M Value: INS12M Label:

If AI31=1 (have had current health insurance for all of the past 12 months)

12 Insured 12 months

If AI34=2 (have current coverage, some kind of health insurance for all of the past 12 months)

12 Insured 12 months

If AI35 >= 0 (months with no health insurance at all)

12 – AI35 value (#)

Insured # months

2. The INS12M values are then assigned to the cases with respondents who do not report current

health coverage during the administration of the questionnaire:

Condition: INS12M Value: INS12M Label:

If AI27=2 (no health insurance for all of the past 12 months)

0 Insured 0 months

If AI29 >= 0 (months with health insurance)

AI29 value (#) Insured # months


47

Note: This variable is constructed in an identical manner in the adult, adolescent, and child data files. INS64 Type of Current Health Coverage Source – Under 65 The INS64 variable indicates the type of current health insurance coverage for persons under 65 years old. INS64 is created with other constructed insurance variables. Each case with an adult who is under 65 years old (if SRAGE < 65) is tested through the following series of conditions until a respective INS64 value is assigned:

Condition:

INS64 Value: INS64 Label:

If INSMD=1 2 Medi-Cal (Medicaid) If INSMC=1 4 Medicare If INSEM=1 5 Employment-based If INSPR=1 6 Privately purchased If INSML=1 or INSOG=1 7 Other public If INSOT=1 6 Privately purchased INS=2 1 Uninsured

Any cases with an adult who is 65 years or older (if SRAGE >= 65) are assigned a skip value (-1) for this variable:

Condition: INS64 Value: INS64 Label:

If SRAGE >= 65 -1 Skipped - Age >= 65 Note: The corresponding constructed variable for adults 65 and older (if SRAGE => 65) is INS65. The INS64_S variable measures type of insurance coverage for the spouse of the respondent age 64 years and younger. Construction of this variable uses the same logic as INS64.

INS64_P Current Health Coverage Under 65 Yrs Old (PUF Recode) This INS64_P variable is re-coded from the constructed variable INS64 and is used in the public use file. This variable also indicates the type of current health insurance coverage for persons under 65 years old, but collapses CHIP coverage with other public health insurance coverage. Cases are assigned based on the following INS64 criteria:

Condition:

INS64_P Value: INS64_P Label:

If INS64=2 2 Medicaid If INS64=4 3 Medicare If INS64=5 4 Employment-based If INS64=6 5 Privately purchased If INS64=3 or INS64=7 6 CHIP/Other public If INS64=1 1 Uninsured

Any cases with an adult who is 65 years or older (if INS64= -1) are assigned a skip value (-1) for this variable:

Condition: INS64_P Value: INS64_P Label:


48

If INS64=-1 -1 Skipped >= 65

Note: The INS64S_P variable measures current health coverage for the spouse of the respondent age 64 years and younger (PUF recode). Construction of this variable uses the same logic as INS64_P.

INS65 Type of Current Health Coverage Source for the Elderly The INS65 variable specifies the type of current health insurance coverage for adults 65 years and older. This variable also indicates whether or not the adult is covered simultaneously by Medicare and some other type of insurance. INS65 is created with other constructed insurance variables. Each case with a respondent who is 65 years or older (if SRAGE >= 65) is tested through the following series of conditions until a respective INS65 value is assigned:


If INSMC=1 (Medicare) and INSMD=1 (Medi-Cal)

1 Medicare + Medicaid

If INSMC=1 (Medicare) and another type of coverage: [INSEM=1 (employer based) or INSPR=1 (private) or INSML=1 (military) or INSOG=1 (other government) or INSOT=1 (other non-government)]

2 Medicare + Other

If INSMC=1 (Medicare) and no other type of coverage: [If INSMD=2 and INSEM=2 (no employer-based) and INSPR=2 (no private) and INSML=2 (no military) and INSOG=2 (no government) and INSOT=2 (no non-government)]

3 Medicare Only

INS=1 4 Other Only INS=2 5 Uninsured

Any cases with an adult, adolescent or child who is under 65 years old (if SRAGE < 65) are assigned a skip value (-1) for this variable:


If SRAGE < 65 (under 65 years old) -1 Skipped – Age < 65 Adjustment 1: Those who are initially not included in the “Medicare + Medi-Cal” category (if INS65~=1), but are covered by Medicare (if INSMC=1) and a supplemental Medicare policy (if AI4=1), are considered to be covered by “Medicare + Other” (INS65=2).


49

Adjustment 2: The respondents with Medicare (if INSMC=1) who are also in a managed care program [(if AI25=1 (covered for Rx), AI21=1 (have to sign up with PCP, group or clinic that must go to), and AI22=1 (have to get referrals)] are assigned to the “Medicare + Other” category. Note: The corresponding constructed variable for the type of current health coverage for persons under 65 years old (if SRAGE < 65) is INS64. The INS65_S variable measures type of insurance coverage for the spouse of the respondent age 65 years and older. Construction of this variable uses the same logic as INS65.

INSANY Any Health Insurance in the Last 12 Months The purpose of the INSANY variable is to indicate whether or not respondents have had any health insurance in the last 12 months. Instead of using the source variables from the questionnaire, INSANY is derived from other constructed insurance variables, including INS64 (Type of current health coverage source – under 65 years old), INS65 (Type of current health coverage source for the elderly), and INS12M (Number of months covered by health plans in past 12 months). Each case is tested through the following series of conditions until an INSANY value is assigned:

1. The respondents under 65 years old (if SRAGE < 65) are first assigned INSANY values:

Condition: INSANY Value:

INSANY Label:

If INS64=1 1 Currently uninsured If 0<=INS12M<12 2 Uninsured any of the last 12

months If INS12M=12 and INS64>1 3 Insured all of the last 12

months

2. Next, the respondents 65 and older (if SRAGE >= 65) are assigned INSANY values:

Condition: INSANY Value:

INSANY Label:

If INS65= 5 1 Currently uninsured If 0<=INS12M<12 2 Uninsured any of the past 12

mo If INS12M=12 and 1<=INS_INS65<5 3 Insured all of the past 12 mo

INSEM Covered by Employer-Based Plans The INSEM variable is derived from questionnaire item AI8, AH56 and AI19. Respondents who identify themselves as covered by a health insurance plan or HMO through a current or former employer/union (if AI8=1) or covered through premium payment for a health plan (if AH56_1, AH56_2, AH56_3, AH56_4 or AH56_5=1) are considered to be covered for this variable (INSEM=1). Those who report that they are not covered by an employer-based plan (AI8=2) are considered to be not covered (INSEM=2). In addition, the cases with INSEM~=1 who report that they are covered by a plan that was missed through a current or former employer/union, school, professional association trade group, or other organization (if AI19_1=1 or AI19_2=1), are also considered to be covered by an employer-based plan for this variable (INSEM=1).


50

Note: The INSEM_S variable provides a dichotomous measure of spouse coverage by employer-based health plans. Construction of this variable uses the same logic as INSEM using items AI40, AI40A, AI47_1, AI47_2, AI49_1 and AH49_2.

INSMC Covered by Medicare The INSMC variable is derived from questionnaire item AI1, AI2, AH56_9 and AI19_4. Respondents who identify themselves as covered by Medicare (if AI1=1 or AI2=2) or covered through a premium payment for a health plan (AH56_9) are considered to be covered for this variable (INSMC=1). Those who identify themselves as not covered (if AI1=2) are considered to be not covered (INSMC=2). In addition, the cases with INSMC~=1 who report that they are covered by Medicare through a plan that was missed (if AI19_4=1) are also considered to be covered by Medicare (INSMC=1). Data editing adjustment 1: Respondents who were younger than 65 years old and reported having Medicare coverage, but were working and not disabled or legally blind, are reassigned as not having Medicare coverage (INSMC=2). Data editing adjustment 2: Respondents who were younger than 65 years old and reported having Medicare coverage, but were not on SSDI, are reassigned as not having Medicare coverage (INSMC=2). Note: The INSMC_S variable measures spouse’s coverage by Medicare. Construction of this variable uses the same logic as INSMC using items AI37, AI47_4, and AI49_4. INSMD Covered by Medi-Cal The INSMD variable is derived from questionnaire items AI6, AH55_7, AH56_7 and AI19_5. Respondents who identify themselves as being covered by Medi-Cal (if AI6=1) or covered through premium payment for a health plan (if AH55_7 or AH56_7=1) are considered to be covered for this variable (INSMD=1). Those who report that they are not covered by Medi-Cal (if AI6=2) are considered to be not covered (INSMD=2). In addition, the cases with INSMD~=1, who report that they are covered by Medi-Cal through a plan that was missed (if AI19_5=1), are considered to be covered for this variable (INSMD=1). Data editing adjustment 1: The respondents with INSMD~= 1 (no Medi-Cal), who report that they have SSI/ AFDC/TANF/CalWorks, are considered to be covered by Medi-Cal (INSMD=1). Data editing adjustment 2: Respondents who received Medi-Cal coverage after signing up through Covered California (AH106=5) are considered to be covered by Medi-Cal (INSMD=1). Note: The INSMD_S variable measures spouse’s coverage by Medi-Cal. Construction of this variable uses the same logic as INSMD using items AI38, AI47_5 and AI49_5.

INSOG Covered by Other Government Plans The INSOG variable is derived from questionnaire items AI17 and AI19_9. Respondents who report that they are covered by some other government plan (such as AIM, Mister MIP, the Family PACT program, or something else) (if AI17=1) are considered to be covered for this variable (INSOG=1). The respondents who skip out of item AI17 (-1) or report that they are not covered by some other government plan (if AI17=2) are considered to be not covered for this variable (INSOG=2).


51

In addition, cases with INSOG ~=1 who report that they are covered by a plan that was missed, which is some other government health plan (if AI19_9=1), are considered to be covered for this variable (INSOG= 1). Note: This variable cannot be used as a count of respondents with other government plans. Only those without Medicare, Medi-Cal, employer, private, or military coverage are asked this question. The INSOG_S variable measures spouse’s coverage by other government plans. Construction of this variable uses the same logic as INSOG using items AI42A, AI47_9, and AI49_9. INSOT Covered by other non-government insurance plan The INSOT variable assigns if the adult respondent’s insurance type is an other, non-government plan. INSOT is constructed from questionnaire variables AH56_6, AH56_10, AI18, and AI19_10. INSOT yes or no category values are assigned as follows:

Condition:

INSOT Value: INSOT Label:

If AH56_6=1 or If AH56_10=1 or If AI18=1 AND AI19_10=1

1 Yes

All other responses to input variables

2 No

INSOT_S Spouse covered by other insurance (non-government plans) The INSOT_S variable assigns if the adult spouse’s insurance type is an other, non-government plan. INSOT_S is constructed from questionnaire variables AI47_10 and AI49_10. INSOT_S yes or no category values are assigned as follows:

Condition:

INSOT_S Value: INSOT_S Label:

If AI47_10=1 or AI49_10=1 1 Yes If AI47_10=2 or -1 And If AI49_10=2 or -1 or -9

2 No

INSPR Covered by Plans Purchased On Own The INSPR variable is derived from questionnaire item AI11 and AI19_3. If respondents report that they are covered by a health insurance plan that was purchased directly from an insurance company or HMO (if AI11=1), they are considered to be covered by a plan purchased on their own (INSPR=1). The respondents who skip out of AI11 (-1), or report that they are not covered by a plan purchased directly (if AI11=2), are considered to be not covered by a plan that was purchased on their own (INSPR=2). In addition, cases with INSPR ~=1 who report that they are covered by a plan purchased directly that was missed (if AI19_3=1) are considered to be covered for this variable (INSPR=1). Note: This variable cannot be used as a count of respondents with private insurance. Only those without Medicare, Medi-Cal, or employer coverage are asked this question.


52

The INSPR_S measures spouse’s coverage by plans purchased directly from an insurance company or HMO. Construction of this variable uses the same logic as INSPR using items AI41, AI47_3 and AI49_3.

INSPS Covered by Employer-Based Plans as Primary Coverage The INSPS variable is derived from questionnaire items AI8, AI9, AH56, AH59, AI19_1 and AI19_2. Respondents who identify themselves as covered by a health insurance plan or HMO through a current or former employer/union (if AI8=1) and report that the plan was obtained in their own name (if AI9=1 or AH59=1), are considered to covered for this variable (INSPS=1). Those who report that they are covered by a health insurance plan or HMO through a current or former employer/union (if AI8=1), but report that the plan was obtained in someone else’s name (if AI9=2 or AH59=1), are considered to have secondary coverage (INSPS=2). Those who report their current or former employers or unions pay the health plan premium (if AH56_1=1, AH56_2=1 or AH56_3=1) are considered to have primary coverage for employer-based plan (INSPS=1). Those who report their spouse’s current or former employers or unions pay the health plan premium (if AH56_4=1 or AH56_5=1) are considered to have secondary coverage (INSPS=2). In addition, those who report having an employer-based plan by a plan that was missed (if AI19_1=1 or AI19_2=1) were skipped out of the question of primary or secondary coverage (AI9). These cases were imputed to assign an INSPS value. Those who are skipped out of item AI9 because they do not have an employer-based plan (AI8=2) are assigned a skip value (-1). Note: The INSPS_S measures spouse’s primary or secondary coverage plans. Construction of this variable uses the same logic as INSPS using items AI40, AI40A, AI9, AI9A, AI40, AI40A, AH55_4, AH55_5, AH56_4, AH56_5, AH60 and AH62.

INSTYPE Insurance Type The INSTYPE variable assigns the adult’s insurance type based on various insurance construct variables. Please note that INSHK and INSHF are no longer used in CHIS 2014 and beyond due to the elimination of Healthy Kids and Healthy Families in late 2013. Due to this, beginning in 2014, Healthy Families is no longer a category for INSTYPE.

Condition:

INSTYPE Value: INSTYPE Label:

If INS=2 1 Uninsured If INSMD=1 and INSMC=1 2 Medicare & Medicaid If [INSMC=1 and (INSEM=1 or INSPR=1 or INSML=1 or INSOT=1 or INSOG=1

3 Medicare & Others

If INSMC=1 4 Medicare Only If INSMD=1 5 Medicaid If INSEM=1 7 Employment-Based If INSPR=1 or INSOT=1 8 Privately Purchased If INSML=1 or INSOG=1 Or If INSMD=1 and INSML=1

9 Other Public


53

INST_12 Health Ins Coverage in Last 12 Mos, Incl Current Status: 8

Lvls The INST_12 variable is derived from the constructed variable INSLT12. This variable re-categorizes health insurance coverage over the past 12 months, including current status, into 8 distinct levels. The values of INST_12 are assigned as follows: Condition: INST_12 Value: INST_12 Label:

If INSLT12=1 1 Medi-Cal (Medicaid) only If INSLT12=2 2 Employer-based coverage only (EBI) If INSLT12=17 3 Private coverage only If INSLT12=10, 18,19 4 Other coverage only If INSLT12=6, 8, 9, 12, 13, 14, 16 5 Any 2 or more types (never uninsured) If INSLT12=3 6 Uninsured only If INSLT12=4 7 Uninsured + Employer-based only If INSLT12=5, 7, 11, 15 8 Any 1 or more types + Uninsured Age >= 65 -1 Inapplicable – Age >=65

INSLT_P Health Ins Coverage in Last 12 Mos, Incl Current Status: 8 Lvls (PUF Recode)

The INSLT_P variable is also derived from the constructed variable INSLT12. This variable re-categorizes health insurance coverage over the past 12 months, including current status, into 8 distinct levels. The values of INSLT_P are assigned as follows: Condition: INSLT_P Value: INSLT_P Label:

If INSLT12=1 1 Medi-Cal (Medicaid) only If INSLT12=2, 18 2 Employer-based coverage only (EBI) If INSLT12=3 3 Uninsured Only If INSLT12=17 4 Privately Purchased Only If INSLT12=5, 11 5 Partially Insured with Medicare If INSLT12=4, 7, 15 6 Partially Insured with Other If INSLT12=6, 8, 12, 13, 14 7 Insured All Year with Medicare If INSLT12=9, 10, 16, 19 8 Other Insured All Year Age >= 65 -1 Inapplicable – Age >=65 Beginning in 2014, Healthy Families is no longer a part of the above categories for INSLT_P, due to phase out of the program.

INS9TP Type of Current Health Coverage Source for All Ages – 9 Levels

The purpose of this variable is to provide a 9-level measurement of type of current health coverage. INS9TP is derived from 8 input variables (INS, INSMC, INSMD, INSEM, INSML, INSPR, INSOT, and INSOG), and its values are assigned as follows: Condition: INS9TP Value: INS9TP Label:

If INS=2 1 Uninsured


54

If INS /= 2 and INSMC=1 and INSEM=1 2 Medicare + Employer-Based If INS /= 2 and INSMC=1 3 Medicare and Others (No Employer-

Based) If INS /= 2 and INSMD=1 and INSEM=1 4 Medicaid + Employer-Based If INS /= 2 and INSMD=1 5 Medicaid (No Employer-Based) If INS /= 2 and INSEM=1 7 Employer Based (No Medicare or

Medicaid) If INS /= 2 and INSPR=1 or INSOT=1 8 Private Purchase If INS /= 2 and INSML=1; or If INS /= 2 and INSOG=1 or INSML=1

9 Other Public

INSLMT12M Reached Health Insurance Payment Limit in Past 12 Months The variable INSLMT12M is constructed from the questionnaire variables AH79B and AH80B. It displays if an adult respondent reached their health insurance payment limit in the past 12 months, displayed as yes or no categories, if the adult respondent is currently insured (excluding adults with Medi-Cal). If a respondent answered AH79B=2, then the respondent’s value for INSLMT12M is also equal to 2 (“No” – did not reach payment limit). If a respondent answered AH79B=1, then the respondent’s value for INSLMT12M is equal to the answered value for AH80B. If the respondent’s value for AH79B<0 (but is not -1), then the respondent’s value for INSLMT12M is equal to the answered valued for AH79B. If the respondent is inapplicable for AH79B (AH79B=-1), then the respondent is inapplicable (INSLMT12M=-1) for the constructed variable INSLMT12M.

HSAAMT_P1 Amount In Health Savings Accounts (PUF 1-Year Recode) The variable HSAAMT_P1 is a PUF recode of the variable HSAAMT, which is constructed from the questionnaire variables AH73B, AH130, and AH131. HSAAMT_P1 categorizes the continuous HSAAMT into 7 categories as follows, displaying amount of money in respondent’s Health Savings Account (HSA), with those respondents who do not have an HSA made Inapplicable for HSAAMT_P1:

HSAAMT_P1 Value: HSAAMT_P1 Label:

1 $0 2 $1 - $999 3 $1000 - $1999 4 $2000 - $2999 5 $3000 - $3999 6 $4000 - $4999 7 $5000+

PINSTYPE Type of Health Insurance Adult Had Before Current Plan The variable PINSTYPE is constructed from the questionnaire variables AI31, AI32, AH134, AH135, AI33_A, and AI33_B. It displays the type of health insurance an adult respondent had prior to the plan the respondent is currently on. If a respondent answered AI31=1 or AI32=2, -1, -7, -8, or -9, then the respondent’s value for PINSTYPE is equal to the answered value for AH135.


55

If the above does not apply and the respondent did not answer AH134=-1, then the respondent’s value for PINSTYPE is equal to the answered value for AH134. If neither of the above conditions apply and the respondent answered AI32=1 and AI33_B=-1, then the respondent’s value for PINSTYPE is equal to the answered value for AI33_A. If a none of the above apply and the respondent was recorded as AI31=-1, then the respondent’s PINSTYPE value will also equal -1 (Inapplicable). The categories for PINSTYPE are as follows:

PINSTYPE Value: PINSTYPE Label:

1 Medi-Cal 3 ESI 5 Private Purchase 6 Covered CA

91 Other Plan 95 No Previous Plan

Note: PINSTYPE is only coded for those respondents who were coded as currently insured during their interview. Respondents who were coded as uninsured, but had insurance in the past 12 months, were also excluded because we do not ask those who were uninsured for more than one year what type of insurance they last had. TCURPLAN Time Adult Has Been Enrolled in Current Plan: Months The variable TCURPLAN is constructed from the questionnaire variables AI31, AH132, and AH133. It displays the time an adult respondent has been enrolled in their current insurance plan in months. If a respondent answered AI31=1, then the respondent’s value for TCURPLAN is equal to the answered value for AH132. If a respondent answered AI31=2, then the respondent’s value for TCURPLAN is equal to the answered value for AH133. If the respondent did not answer AI31=1 or AI31=2, then the respondent is inapplicable (TCURPLAN=-1) for the constructed variable TCURPLAN. UNINSANY Uninsured in Past 12 Months The UNINSANY variable is derived from the constructed variable INSLT12R, which measures health insurance coverage in the last 12 months. This variable assigns values based on the adult’s insurance status during all or part of the year. Values are assigned as follows:

Condition: UNINSANY Value: UNINSANY Label:

If INSLT12R=3 1 Uninsured all year If INSLT12R=6, 7 2 Uninsured part year If INSLT12R=1, 2, 4, 5, 8, 9 3 Insured all year

OFFTK Offer, Eligibility, Acceptance of Employer-Based Insurance (EBI)


56

This variable is constructed from a series of other constructed insurance variables, as well as questionnaire items. OFFTK categorizes the offering, eligibility and acceptance of health insurance plans by the respondent from his or her employer. Adults working for wages that are covered by employer-based plans (INSEM=1) or who have employer-based coverage as their primary plan (INSPS=1) are considered to have accepted employer-based coverage (OFFTK=1). Otherwise, among persons who were working last week (AK1=1) or who usually work (AG10=1), case assignments are based on the following criteria:

Condition: OFFTK Value: OFFTK Label:

If AI13=1 and AI14=1 2 Did not accept EBI, but was offered and eligible

If AI13=1 and AI14=2 3 Was offered EBI, but was not eligible

If AI13=2 4 Was not offered EBI Adults indicating an unemployed status are assigned a skip value of OFFTK=(-1). Note: The OFFTK_S variable is constructed using the same logic as OFFTK. This variable also categorizes the eligibility and acceptance of health insurance plans offered to the respondent’s spouse by the spouse’s employer.

MC_TYPE Type of Medicare Coverage This variable displaying type of Medicare coverage is constructed from a series of other insurance-related variables, both constructed variables and questionnaire items. MC_TYPE is constructed using responses from INSMC, AH123, and AI4, as follows:

Condition: MC_TYPE Value: MC_TYPE Label:

If INSMC=2 -1 Not Enrolled in Medicare

If INSMC=1 and AH123=1 1 Medicare Advantage If INSMC=1 and AI4=1 2 Medicare + Medicare

Supplemental Plan (from AI4)

If INSMC=1 and all other responses to AH123 and AI4

3 Other Medicare

AH101_P Person who helped Find Health Plan (PUF Recode) AH101_P is a recoded version of AH101, which identifies the person who helped the adult respondent find their health plan while they were trying to purchase the insurance on their own. This question is only asked among those who are currently uninsured or have been uninsured in the past 12 months.

AH95_P # Times Visited ER in Past 12 Months (PUF Recode) AH95_P is a categorical recode of the continuous variable AH95, which indicates the number of times the adult respondents visited the ER in the past year. Values greater than 7 were combined and collapsed into different categories in order to combine potentially identifying data into less identifiable categories. Categories were assigned as follows:


57

Condition:

AH95_P Value: AH95_P Label:

If 7 <= AH95 < = 8 7 7-8 times If 9 <= AH95 < = 12 8 9-12 times If 13 <= AH95 < = 24 9 13-24 times If AH95 > = 25 10 25+ times

AH95_P1 # Times Visited ER in Past 12 Months (PUF 1-Year Recode) AH95_P is a categorical recode of the continuous variable AH95, which indicates the number of times the adult respondents visited the ER in the past year. Values greater than 5 were combined and collapsed into a single category in order to mask potentially identifying data, withAH95_P1=5 (5 or more times).

AH71_P1 Health Plan Deductible More than $1,000 (PUF 1-Year Recode)

AH71_P is a recode of the variable AH71, which indicates whether the adult respondent’s health plan has a deductible that is more than $ 1,000. Cases in the third category under AH71 (“Yes, but only when I go out of network”) were re-categorized as not having health plan deductible greater than $1,000.

AH72_P1 Health Plan Deductible More than $2,000 (PUF 1-Year Recode)

AH72_P is a recode of the variable AH72, which indicates whether the adult respondent’s health plan has a deductible for all covered persons that is more than $ 2,000. Cases in the third category under AH72 (YES, BUT ONLY WHEN I GO OUT OF NETWORK) were re-categorized as not having health plan deductible greater than $2,000.

AH98_P1 Difficulty Finding Plan with Needed Coverage (PUF 1-Year Recode)

AH98_P1 is a recoded version of AH98, a Likert scale measure which asks adult respondents about difficulty finding a private health plan with the coverage they needed. This question was asked of adults with private health coverage in past 12 months (current and past coverage) and currently uninsured adults who tried to purchase health insurance directly from an insurance company or HMO (AH110=1 or 3). The construct combines AH98=3 and AH98=4 to create a three-level variable with the following categories: Very Confident (AH98_P1=1), Somewhat Confident AH98_P1=2), Not too Confident/Not at all Confident (AH98_P1=3).

AH99_P1 Difficulty Finding Plan that is Affordable (PUF 1-Year Recode)

AH99_P1 is a recoded version of AH98, a Likert scale measure which asks adult respondents about difficulty finding a private health plan that was affordable. This question was asked of adults with private health coverage in past 12 months (current and past coverage) and currently uninsured adults who tried to purchase health insurance directly from an insurance company or HMO (AH110=1 or 3). The construct combines AH99=3 and AH98=4 to create a three-level variable with the following categories:


58

Very Confident (AH99_P1=1), Somewhat Confident AH98_P1=2), Not too Confident/Not at all Confident (AH99_P1=3).

ELGMAGI3 Medi-Cal Eligibility for Uninsured: ACA MAGI (2 Levels) ELGMAGI3 is variable asked of all uninsured adults, which is constructed from revised Medi-Cal eligibility criteria based on modified adjusted gross income (effective January 2014). The two levels created for this variable are 1=MEDI-CAL (MEDICAID) ELIGIBLE and 3=NOT ELIGIBLE. The 2010 Patient Protection and Affordable Care Act (ACA) included numerous changes to Medi-Cal eligibility. Effective January 1, 2014, income-based eligibility for Medi-Cal and Healthy Families (California’s Children’s Health Insurance Program) participation is determined by Modified Adjusted Gross Income (MAGI). CHIS 2014 and beyond included the following changes to approximate a respondent’s MAGI: a. Added two questions to identify households receiving workers’ compensation in the past month and the amount received in order to exclude workers’ compensation income from total income (AL32, AL33); b. Modified questions on child support to eliminate “other” income sources (government or veterans’ programs) (AL15, AL16) to exclude child support income from total income; and c. Added two questions clarifying whether there is anyone else not living in the household, but living in the U.S., who is supported by the total household income reported (AK32, AK33). The MAGI approximation informed the construction of the CHIS variable, ELGMAGI3, capturing uninsured individuals who are “newly eligible” for Medi-Cal under the ACA. ELGMAGI3 also categorizes many uninsured individuals eligible for Healthy Families as Medi-Cal eligible due to the transition of Healthy Families enrollees into the Targeted Low-Income Medicaid Program. See the Analyze CHIS Data forum for more information (http://healthpolicy.ucla.edu/forum/Pages/Forum.aspx). Additionally, changes may have been made to this variable across years due to the phase out of Healthy Families programs.

HEALTH CARE UTILIZATION AND ACCESS

ACMDNUM Number of Doctor Visits in the Past Year The ACMNDUM variable is derived from the continuous AH5 variable, which assigns the number of doctor visits in the last year as reported by the respondent. The ACMDNUM variable provides 10 categories for the number of visits reported. ACMDNUM values are assigned as follows:

Condition:

ACMDNUM Value: ACMDNUM Label:

If AH5=0 0 0 visits If AH5=1 1 1 visit If AH5=2 2 2 visits If AH5=3 3 3 visits If AH5=4 4 4 visits If AH5=5 5 5 visits If AH5=6 6 6 visits If AH5=7, 8 7 7-8 visits If AH5=9 to 12 8 9-12 visits If AH5=13 to 24 9 13-24 visits


59

If AH5=25+ 10 25+ visits

DOCT_YR Visited a Doctor During the Past 12 Months The DOCT_YR variable is derived from the questionnaire item AH5. The DOCT_YR variable is a dichotomous variable that ascertains whether or not the adult respondent visited a doctor at least once during the past 12 months. Those who indicated one or more visits (AH5>=1) were assigned the value DOCT_YR=1. Those indicating 0 visits (AH5=0) were assigned the value DOCT_YR=2. Cases in which the number of visits could not be ascertained were imputed to assign a value to DOCT_YR.

AJ50_P Language Doctor Speaks To Respondent (PUF Recode) The AJ50_P variable is a recoded version of AJ50 which assigns the language used by the respondent’s doctor to communicate with the respondent. The purpose is to collapse categories into less identifiable categories. The values are assigned as follows:

Condition:

AJ50_P Value: AJ50_P Label:

If AJ50=1 1 English If AJ50=2 2 Spanish If AJ50=4 3 Vietnamese If AJ50=7 4 Korean If AJ50=3 5 Cantonese If AJ50=6 6 Mandarin If AJ50=5 or AJ50>=8 7 Other

USOC Usual Source of Care Other Than ER The USOC variable is derived from constructed variable USUAL_TP. USOC provides a dichotomous measure of whether an adult respondent has a usual source of care other than emergency room services. Each case is first tested through the following conditions until a USOC value is assigned: Condition:

USOC Value: USOC Label:

If USUAL_TP=1 (Doc Office/HMO/Kaiser) 2 (Commun/Gov Clinic) 5 (Some Other Place) or 6 (No One Particular Place)

1 Yes

If USUAL_TP=3 (Emergency Room) or 7 (No Usual Source of Care)

2 No

Cases in which respondents refused to report their usual source of care or did not know their usual source of care are imputed to assign a value to USOC.

USUAL Have Usual Place to Go When Sick or Needing Health Advice

USUAL is constructed with questionnaire item AH1.


60

USUAL is constructed by combining the responses in AH1 in order to create a dichotomous variable for usual source of care. The respondents who report in questionnaire item AH1 that they have a usual place (AH1=1), have a doctor (AH1=3), go through Kaiser (if AH1=4), or usually go to more than one place (if AH1=5), are considered to have a usual place to go when sick or needing health advice (USUAL=1). Respondents who indicate not having a usual place to go to when sick (AH1=2) are assigned the value of USUAL=2. AH3_P1 Kind of Place for Usual Source of Health Care (PUF 1-Year

Recode) AH3_P1 is a four-level variable constructed from questionnaire item AH3:

Condition:

AH3_P1 Value: AH3_P1 Label:

If AH3=4, 5, 6, 91, or 94 4 Other/No One Place If AH3=1 1 Doctor’s Office/Kaiser/HMO If AH3=2 2 Clinic/Health

Center/Hospital Clinic If AH3=3 3 Emergency Room

USUAL_TP Usual Source of Care – 7 Levels The USUAL_TP variable is derived from questionnaire items AH1, AH3, and AH4, which measure source of health care for the adult respondent. The constructed USUAL_TP variable categorizes cases based on the place most often sought for source of health care. Values are assigned according to the following criteria:

Condition:

USUAL_TP Value: USUAL_TP Label:

If AH1=3 or 4 or AH3=1, 2, or 3

1 Doc Office/HMO/Kaiser

If AH3=4 2 Community/Gov. Clinic, Community Hospital

If AH3=5 3 Emergency Room If AH3=6, 7 or 8 5 Some Other Place If AH3=94 6 No one particular place If AH1=2 7 No usual source of care

Cases in which respondents did not know their usual source of care or the usual source of care cannot be determined are imputed to assign a value to USUAL_TP.

USUAL5TP Usual Source of Care – 5 Levels The USUAL5TP variable is derived from the constructed variable USUAL_TP, which categorizes the most often visited place the adult respondent goes to for health care. Five levels are assigned to USUAL5TP that re-categorize the usual source of care for the respondent. Values are assigned according to the following criteria:

Condition:

USUAL5TP Value: USUAL5TP Label:


61

If USUAL_TP=1 1 Doc Office/HMO/Kaiser If USUAL_TP=2 2 Community/Gov. Clinic,

Community Hospital If USUAL_TP=3 3 Emergency Room/Urgent

Care If USUAL_TP=5 or 6 4 Other place, no one place If USUAL_TP=7 5 No usual source of care

Cases in which a usual source of care cannot be determined are imputed to assign a value of USUAL5TP.

AJ131_P1 Main Reason for Delaying Needed Care (PUF 1-Year Recode) AJ131_P1 is a recode of the variable AJ131, which provides the main reason why (other than cost/lack of insurance) respondents had to delay or forgo other needed medical care. This question is asked of respondents who delayed or did not get care in the past 12 months (AH22=1) and indicated that cost or lack of insurance was not the main reason (AJ130=2).

Condition:


If AJ131=1 1 Couldn’t Get Appointment If AJ131=2,3, 10 or 11 2 Insurance Not Accepted/Did

not Cover/No Insurance If AJ131=5 5 Transportation Problems If AJ131=6 6 Hours not convenient If AJ131=9 9 I Didn’t Have Time If AJ131=12 12 Did Not Think Serious

Enough/Needed If AJ131=13 13 Not Satisfied with Health

Care Received If AJ131=4,7,8, 14, or 16 14 Other Reason If AJ131=15 15 Anxiety, Fear, or Avoidance

of Medical Care

FORGO Had to Forgo Necessary Care The construct FORGO is constructed using questionnaire items AH22 and AJ129. This dichotomous variable indicates whether or not adults had to forgo necessary medical care in the past 12 months. Individuals who reported that they delayed care in the past 12 months (AH22=1) and that they did not get the care eventually (AJ129=2) were categorized as forgoing necessary care.

RN_FORGO Reasons Forgone Necessary Care The construct RN_FORGO is based on questionnaire items AJ20, AJ130, and AJ131. This variable classifies the various reasons why adults had to forgo necessary medical care in the past 12 months into 4 summary categories. Among adults who reported cost/lack of insurance as a reason why they delayed/had forgone care (AJ20=1) and said that was the main reason (AJ130=1), they were classified as “Cost or Lack of Insurance” as the main reason (RN_FORGO =1). Those who reported that cost wasn’t the main reason (AJ130=2) moved on to the next question (AJ131) to provide the main reason. Those who reported AJ131=10 or 11 were also categorized as RN_FORGO=1. The table summarizes the category assignments for RN_FORGO.

Condition: RN_FORGO RN_FORGO Label:


62

Value: If AJ20=1 and AJ130=1 1 Cost or Lack of Insurance If AJ130=2 and/or AJ131=10 or 11

1 Cost or Lack of Insurance

If AJ131=1 2 Couldn’t Get Appointment If AJ131=6,9 3 Didn’t Have Time If AJ131=2, 3, 4, 5, 7, or 8 4 Others

ER Emergency Room Visit in the Past Year The ER variable is constructed using questionnaire items AH12, AH13A, AB67, AB109 and AB115. This dichotomous variable indicates whether or not the adult visited an emergency room for any reason within the past year. Respondents who indicated that they visited an emergency room within the last year (AH12=1) are assigned a value of ER=1. Current asthmatics who visited the ER for asthma in the past year (AH13A=1) are assigned a value of ER=1. Respondents who have been diagnosed with asthma but who do not currently have asthma and visited the ER for asthma in the past year (AB67=1) are assigned a value of ER=1. Diabetics who visited the ER for diabetes in the past year (AB109=1) are assigned a value of ER=1. Finally, respondents who visited the ER for heart disease in the past year (AB115=1) are also assigned a value of ER=1. Those respondents who did not visit the ER for any reason (AH12=2) or who did not visit the ER for asthma in the past year (among current asthmatics and those who used to have asthma) (AH13A=2 or AB67=2) or who did not visit the ER for diabetes or heart disease in the past year (AB109=2 or AB115=2) are assigned a value of ER=2. Those who were coded as “Inapplicable” (-1) for questionnaire items AH12, AH13A, AB67, AB109 and AB115 were also assigned a value of -1 or “Inapplicable” for the variable ER.

CARE_PV Had a Preventive Care Visit in the Past Year The CARE_PV variable is a dichotomous indicator and is constructed using AJ114, which asks respondents how long since they had last seen a medical provider for a routine check-up. Those reporting “One Year Ago or Less” (AJ114=0) were classified as having a preventative care visit in the past year (CARE_PV=1). All other adults were classified as not having a preventative care visit in the past year (CARE_PV=2), including adults who reported that they have never seen a doctor about their own health (AH6 = 4). TIMAPPT Able to Get an Appointment in a Timely Way The TIMAPPT variable is a dichotomous indicator constructed from questionnaire items AJ102 and AJ103. Individuals who needed a doctor appointment within two days in the past 12 months (AJ102=1) were asked how often they were able to get a timely appointment. Respondents who reported sometimes, usually, or always (AJ103=2, 3, or 4) were categorized as Able to Get an Appointment in a Timely Way (TIMAPPT=1); respondents who reported never (AJ103=1) were categorized as Unable to Get an Appointment in a Timely Way (TIMAPPT=2).

EMPLOYMENT, INCOME, POVERTY, & FOOD SECURITY

AG9_P1 Type of Employer on Spouse’s Main Job (PUF 1-Year Recode)


63

AG9_P1 is a recode of the variable AG9, which indicates the type of employer for respondent spouses and combines AG9=3 (Self-Employed) and AG9=4 (Family Business or Farm). AG9_P1 is a three-level variable. AK10_P Earnings Last Month Before Taxes and Deductions (PUF Recode) AK10_P is a top coded version of questionnaire item AK10. Note: Top code is $30,000

AK10A_P Spouse’s/Partner’s Earnings Last Month (PUF Recode) AK10A_P is a top coded version of questionnaire item AK10A. Note: Top code is $30,000

AK2_P1 Main Reason for Not Working Last Week (PUF 1-Year Recode)

AK2_P1 is a recoded variable of AK2 that recategorizes AK2=9 (Family or Maternity Leave) into AK2_P1=91 with the additional “Other” reasons among respondents with jobs who did not work last week and respondents without work (this includes adults who were not looking for work).

AK7_P1 Length of Time Working at Main Job (PUF 1-Year Recode) AK7_P1 is a top coded version of questionnaire item AK7, which captures the length of time at main job in months among respondents who currently work. Note: Top code is 456.

AK22_P Household’s Total Annual Income (PUF Recode) AK22_P is a top coded version of questionnaire item AK22. Note: Top code is 300,000.

FPG Federal Poverty Guideline The FPG construct uses the continuous family poverty threshold variable, POVGWD, to create a 4-level indicator that categorizes family income into groupings aligning with public health coverage income eligibility: FPG 0 to 138 (FPG=1), FPG 139 to 200 (FPG=2), FPG 201 to 400 (FPG=3), and FPG 400+ (FPG=4).

POVGWD_P Family Poverty Threshold Level (PUF Recode) The POVGWD_P construct is a recoded variable of POVGWD that measures family poverty threshold level. The POVGWD variable uses AK10, AK10A, and other items to construct adjusted gross income.


64

The following income adjustments are made before income is finalized:

• Add in money received from alimony and child support (AL15, AL16). • Subtract money paid for child support or alimony (AL17, AL18). • Subtract $50/month for standard dependent support deduction (AL16) • Subtract $90/month for standard work expenses/earned income deduction (AK10, AK10A). • Subtract up to $375/month depending on number/age of children in family for child care expenses

(AH44A, AH44B). For cases where respondent/spouse monthly earnings are not provided, reported annual household income (AK22) is converted to estimate monthly earnings. As this questionnaire item asks respondents to consider income from all sources, which may already include child support received, the first adjustment (to add child support payments to monthly earnings) is not applied. The variable categorizes the adjusted income according to Health and Human Services (HHS) poverty guidelines to provide a family poverty level measure for the selected adult’s family based on family count. POVGWD is used to determine the income level for Medi-Cal eligibility in variable ELIGPRG3 which is based on pre-ACA guidelines. Top code: 24.

POVGWD2 Family Poverty Threshold Level The POVGWD2 variable uses AK10, AK10A, and other items to construct adjusted modified gross income (MAGI) to align with the Affordable Care Act. The respondent/spouse’s monthly earnings are used as the primary income source to determine income eligibility (AK10, AK10A). There are differences in how supplemental income sources or expenses are treated compared to POVGWD:

• Child support received is no longer counted as income and child support paid is no longer considered a deduction (Al15, AL16, Al17, and AL18).

• Alimony received is counted as income (Al15, AL16, Al17, and AL18). • Child Care no longer counts as an income deduction. • Worker’s compensation is not counted as income (AL32 and AL33). • No deductions made for work expenses.

For cases where respondent/spouse monthly earnings are not provided, reported annual household income (AK22) is converted to estimate monthly earnings. This questionnaire item asks respondents to consider income from all sources, which may already include child support and workers’ compensation received. Since the new guidelines no longer consider child support and workers’ compensation received as income, these amounts are subtracted from annual household income. While child support paid or received is no longer counted under MAGI, and alimony received is counted under MAGI, the questions combine both income sources. To address this issue, MAGI eligible groups are treated as follows: all alimony and child support income as child support for those who should not receive alimony (currently married, widowed, and never married not living with a partner). For those who are separated or divorced or living with a partner, if no children are present in the family, allocate the full amount to alimony. If children are present, allocate half to child support and half to alimony. The variable categorizes MAGI according to Health and Human Services (HHS) poverty guidelines to provide a family poverty level measure for the selected adult’s family based on family count. POVGWD2 is used to determine the income level for Medi-Cal eligibility in variable ELGMAGI3 following implementation of the Affordable Care Act.


65

POVLL Poverty Level The POVLL variable indicates the total annual income of the household as a percent of the Federal Poverty Level. In order for the data collection vendor to approximate the 100%, 200%, and 300% Federal Poverty Level cutoff points for each household, the respondents were asked to report the number of people living in their household who are supported by the total annual household income (AK17/HHINC), and if needed, how many of those people are children under 18 years old (AK18). The 100%, 200%, and 300% cutoff values for each household were calculated during the administration of the survey by multiplying the 2007 Census Poverty Threshold “size of family unit” by “related children under 18 years” table amounts by 1, 2, or 3 (U.S. Bureau of the Census: Current Population Survey). The income values were then rounded to the nearest 100 dollars. The three household income cutoff points for each household were then stored as CATI variables POVRT100, POVRT200, and POVRT300. A. First, the income values within the poverty variables (POVRT100, POVRT200, POVRT300) are categorized into the same income range levels as the household income variable (HHINC), creating three transitional variables (i.e. POVRT100n, 200n, 300n). B. Second, the POVLL values are assigned.

1. Each case with a POVRT100n value equal to (-9) is assigned a value of 4 (301% FPL and above) that indicates an income of 301% FPL and above.

2. Next, questionnaire items AK18A, AK18B, AK18C and the CATI variables POVRT100, POVRT200, and POVRT300 are used in order to assign POVLL values to the recoded cases. Each case is tested through the following series of conditions until a value is assigned:

Condition: POVLL Value: POVLL Label:

AK18A=1 (equal to or less than calculated POVRT100)

1 0-99 %FPL

AK18A=2 (more than POVRt100) or AK18B=1 (equal to or less than calculated POVRT 200)

2 100-199% FPL

AK18B=2 (more than POVRT200) or AK18C=1 (equal to or less than POVRT300)

3 200-299% FPL

AK18C=2 (more than POVRT300)

4 300% FPL and above

3. For the remaining cases, the actual household income values (HHINC) are compared to the

transitional poverty variables -- POVRT100n, POVRT200n, and POVRT300n -- which have the same range levels. Each case is tested through the following conditions until a respective POVLL value is assigned:

Condition: POVLL Value: POVLL Label:

If HHINC <= POVRT100n 1 0-99% FPL If HHINC <= POVRT200n 2 100-199% FPL


66

If HHINC <= POVRT300n 3 200-299% FPL If HHINC > POVRT300n 4 300% FPL and above

POVLL_ACA Poverty Level as Times of 100% FPL (PUF Recode) . The POVLL_ACA variable is based on the POVLL2 (source) variable (in turn based on questionnaire item AK22) a construct created by dividing POVLL into 100% of FPL. POVLL_ACA provides a recoded categorical measure of poverty that aligns with public health coverage income eligibility thresholds.

Condition: POVLL_ACA Value:

POVLL_ACA Label:

POVLL2 < 1.39 1 0-138% FPL 1.39 ≤ POVLL2 < 2.50 2 139-249% FPL 2.50 ≤ POVLL2 < 4.00 3 250-399% FPL POVLL2 ≥ 4.00 4 400% FPL and above

ELIGPRG3 Uninsured Medi-Cal/Healthy Families Eligible (3 levels) A series of eligibility variables was constructed to estimate and categorize the number of uninsured Californians who meet the eligibility criteria for the “full-scope” Medi-Cal or Healthy Families programs if they were to apply. The estimated number of uninsured eligible is used to calculate program participation rates for the Medi-Cal and Healthy Families programs. Criteria for assignment within these eligibility variables are based on a number of factors:

A. Categorical Eligibility: Persons eligible for program participation must meet a number of age-related and/or disability criteria. Questionnaire items are used to measure age, disability status, pregnancy status, and whether the respondent is a parent of a minor.

B. Family Composition: Questionnaire items are used to derive family composition

necessary for eligibility with these two programs. Variables used include the adult respondent’s marital status; the presence of a spouse in the household; and whether each child in the household is related by blood, guardianship to the adult respondent, their spouse or their unmarried partner with whom they share a biological child, or their unmarried partner with whom they share guardianship of a non-biological child.

C. Income Eligibility: Family income as a percent of the federal poverty guidelines

(POVGWD) is used for both Medi-Cal and Healthy Families income eligibility. The monthly earnings by the adult respondent and/or spouse of the adult respondent and the Federal Poverty guidelines are used as the primary income source in constructing the eligibility variable.

D. Immigration Status: In order to participate in the full-scope Medi-Cal and Healthy Families

programs, eligible persons must be citizens or legal residents. Questionnaire items related to immigration status are used to construct the eligibility variable.

E. Asset Test: Adults in the Medi-Cal program are subject to an asset test, but there is no

asset test for children in either the Medi-Cal or Healthy Families programs. The main questionnaire item used to construct this variable addresses the combined values of specific types of family assets exceeding $5,000.

Note: Other constructed eligibility variables include ELIGPRG4 (Uninsured Medi-Cal/Health Families Eligible--4 levels) and ELIGPP03, ELIGPP04, ELIGPP05, ELIGPP07 (Eligibility program including local child coverage expansions). These variables are constructed using the same logic and criteria as ELIGPRG4.


67

Note 2: Additional changes may have been made to this variable (as well as others mentioned above as having similar construct logic and criteria) across years (2014 and beyond) due to the elimination and phase out of Healthy Kids and Healthy Families programs in late 2013.

FSLEV Food Security Status Level The FSLEV variable provides a categorical measure of food security status for adults who fall below the 200% federal poverty line. This variable is derived from questionnaire items AM1 through AM5. Construction of this variable is based on the following criteria:

1. First, a temporary variable (FSLEVSCR) is created that represents an additive food insecurity score derived from items AM1-AM5. Adults indicating some type of food deprivation in the past 12 months were assigned a value of 1, whereas adults indicating no food deprivation in the past 12 months were assigned a value of 0. The range for this temporary variable is 0 to 6.

2. Next, scores were assigned to one of the following food security metric scale scores:

3. Finally, cases were assigned to one of the following food security status levels (FSLEV):

4. Cases with fewer than 3 missing values in AM1-AM5 (-7, -8 or –9) were imputed. Cases with more than 3 missing values are computed.

FSLEVCB Food Security Status (2 Levels) The FSLEVCB variable is derived from the constructed variable FSLEB. This variable provides a dichotomous measure of food security, whereby persons who are food secure (FSLEV=1) are assigned a value of FSLEVCB=1. Adults who are considered to be food insecure without hunger (FSLEB=2) and with (FSLEV=3) hunger are assigned a value of FSLEVCB=2. Skip values are assigned to cases that fall above 200% of federal poverty line (FSLEVCB=-1).

ELDER_IDX Elderly Single/Couple Income Below County Cost of Living Thresholds

Condition: FSLEVTEMP Value:

If FSLEVSCR=0 0 If FSLEVSCR=1 2.04 If FSLEVSCR=2 2.99 If FSLEVSCR=3 3.77 If FSLEVSCR=4 4.50 If FSLEVSCR=5 5.38 If FSLEVSCR=6 6.06

Condition: FSLEV Value:

FSLEV Label:

If FSLEVTEMP=0 or 2.04 1 Food secure If FSLEVTEMP=2.99, 3.77, or 4.50

2 Food insecure without hunger

If FSLEVTEMP=5.38 or 6.06 3 Food insecure with hunger


68

The ELDER_IDX variable is a dichotomous measure of income inadequacy for adults, aged 65 or older who reside in families with 2 or fewer persons. This variable is constructed based on the 2011 Elder Index amounts for single elders and older couples living alone. The Elder Index amounts are calculated for all counties in California and use publicly available data to calculate basic living costs for housing, food, healthcare, transportation and miscellaneous items; these costs do not assume any subsidies. For a detailed description of ELDER_IDX, please refer to the following methodology report and website: http://healthpolicy.ucla.edu/programs/health-disparities/elder-health/EIRD2011/Documents/elderindex_methodology2011.pdf http://healthpolicy.ucla.edu/programs/health-disparities/elder-health/Pages/elder-index-2011.aspx Values of ELDER_IDX were assigned in the following manner: Using FAM_SIZE, adults aged 65 or older were divided into six groups based on marital/partner status and whether they rented or own their home, and, for homeowners, whether they paid or do not pay for a mortgage(s) in 2011-2012: 1. Single Elders, homeowner with mortgages 2. Older Couples, homeowner with mortgages 3. Single Elders, homeowner without mortgages 4. Older Couples, homeowner without mortgages 5. Single Elders, renter 6. Older Couples, renter Individual annual income (AK22) was compared to the county-specific minimum annual family income needed for each of these six groups. If a respondent’s annual income was less than the minimum income value threshold for county of residence, then 1 (YES) was assigned to ELDER_IDX, otherwise a value of 2 (NO) was assigned. Note: Family size and household size are not interchangeable. Current ELDER_IDX construction may include elderly single/couple family types residing in single households with multiple families. ELDER_IDX was constructed using the source variable, FAM_SIZE, which assigns family size of 1 or 2 persons; however, these persons may reside in households which are larger in size than their family size. Please refer to AskCHIS to obtain population estimates for elderly singles/couples that live alone or subset ELDER_IDX where HH_SIZE=FAM_SIZE; FAM_SIZE is a confidential variable. Please refer to the following for more information: http://healthpolicy.ucla.edu/chis/data/Pages/confidential.aspx.

PUBLIC PROGRAM PARTICIPATION

AL18_P Total Alimony/Child Support Paid Last Month (PUF Recode) AL18_P is a top coded version of questionnaire item AL18. Note: Top code is 2,000.

GEOGRAPHICAL INFORMATION

UR_CLRT2 Rural and Urban – Claritas (By Zip Code) (2 levels) Four urbanization categories are defined for the ZIP codes in California by the commercial company Nielson, Inc (please see constructed variable UR_CLRT). The UR_CLRT2 variable is a modified version


69

of the constructed UR_CLRT variable. The UR_CLRT2 variable designates all ZIP codes as either rural or urban.

1. The cases assigned to the urban, 2nd city, or suburban UR_CLRT categories (if UR_CLRT=1, 2, or 3) are considered to be urban (UR_CLRT2=1).

2. The cases assigned to the small town or rural UR_CLRT category (if UR_CLRT=4) are

considered to be rural (UR_CLRT2=2). Note: This variable is particularly useful since it provides an estimate that seems to correspond to the Census definition of urbanized and non-urbanized areas. As Nielson Inc. states, “The rural and small town/exurban classifications are not far from the density cutoff of the Census definition that distinguishes urbanized from non-urbanized areas as those having densities above/below 1,000 persons/square mile.”

UR_IHS Rural and Urban – Indian Health Service The UR_IHS variable uses a county-level classification of rural and urban from the Indian Health Service. According to the IHS definition, counties are classified either as urban or rural. All counties (SRCNTY) are classified as either rural or urban using the IHS definition. In addition, the cities of San Diego, Santa Barbara, and Bakersfield are coded as urban.

1. Respondents who report that they live within an urban county are coded as urban (UR_IHS=1).

2. Respondents who report that they live in ZIP codes within the cities of San Diego, Santa Barbara, or Bakersfield are also assigned to the urban category for this variable (UR_IHS=1).

3. Respondents who report that they live within a rural county, but are not in ZIP codes for the cities

of San Diego, Santa Barbara or Bakersfield, are considered to be rural (UR_IHS=2).

UR_OMB Rural and Urban – OMB The UR_OMB variable reflects the Office of Management and Budget’s (OMB) classification of metropolitan statistical areas (MSAs). Counties are considered to be metropolitan or non-metropolitan depending on whether or not they are included in an MSA. All except one stratum level in the data file are composed entirely of either metropolitan or non-metropolitan counties. Each case is tested through the following series of conditions until an UR_OMB value is assigned:

1. The cases with respondents who report that they live within a metropolitan county (SRCNTY) are assigned to the metropolitan category (UR_OMB=1).

2. The cases with respondents who report that they live within a non-metropolitan county (SRCNTY)

are assigned to the non-metropolitan category (UR_OMB=2). UR_RHP Rural and Urban – Office of Rural Health Policy THE UR_RHP variable uses an operational classification of rural and urban from the Federal Office of Rural Health Policy (ORHP). The ORHP classifies counties as either rural or urban. The counties are classified with the same criteria that the Office of Management and Budget uses to determine metropolitan and non-metropolitan areas (see UR_OMB). However, to consider particular rural areas within large urban counties (>1225 square miles), certain census tracts within these counties are designated as rural.


70

Each case is tested through the following series of steps until an UR_RPH value is assigned:

1. Respondents who report that they live within counties that are designated as rural are coded as rural (UR_RPH=2).

2. The cases with census tracts that are designated as rural, within a large urban county, are

assigned to the rural category.

3. The remaining respondents who report that they live within a county that is classified as urban counties are coded as urban (UR_RHP=1).

UR_BG6 Rural and Urban – Claritas (By Block Group) (6 Levels) The UR_BG6 variable assigns rural and urban block group using the definition from the commercial company, Nielson, Inc (formerly Claritas). We obtain a file from Nielson Inc. that contains block groups in California and their associated urbanization categories. In order to construct UR_BG6, the block group for each case (using the CBLK variable) is assigned to its corresponding urbanization category as provided by Nielson. For cases with missing CBLK data, the block the respondent reports using questionnaire items AM8 and AM9 is used in order to make this assignment. Nielson assigns block groups in California to 4 urbanization categories based on the analysis of population density grids using projected geoboundaries, redistricting updates, and population estimates. The urbanization categories are defined by Nielson, Inc. as follows:

Urban

Blocks associated with dense neighborhoods that represent the central cities of most major metropolitan areas (more than 4,150 persons/square mile).

2nd City Blocks associated with moderate-density neighborhoods in population centers (more than 1,000 and fewer than 4,150 persons/square mile).

Suburban Blocks associated with moderate-density neighborhoods that are not surrounded by urban or second-city population centers (estimated to be more than 1,000 persons/square mile and not in an urban or 2nd city population center).

Town or Rural Blocks associated with isolated small towns or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950 persons/square mile). Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile).

Beginning in 2017, we expanded the number of categories to 6, to include new categories for “Mixed”, “Town”, and “Rural”. The “Mixed” category is derived from the previous 2nd City category and denotes locations that can be categorized as a mix between 2nd City and Suburban. The remaining previous 2nd City cases are still categorized as 2nd City. Additionally, the previous “Town or Rural” category has been split into two new categories: one for “Town” as defined by “Blocks associated with isolated small towns


71

or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950 persons/square mile)”; and one for “Rural” as defined by “Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile)”. Cases with no block information are imputed to assign a value to UR_BG6.

UR_CLRT6 Rural and Urban – Claritas (By Zip Code) (6 Levels) The UR_CLRT6 variable uses a definition of rural and urban from the commercial company Nielson, Inc (formerly Claritas). Nielson assigns the ZIP codes in California to 4 urbanization categories based on the analysis of population density grids using projected geoboundaries, redistricting updates, and population estimates. We obtained a file from Nielson Inc. that contains the ZIP codes in California and their associated urbanization categories. The urbanization categories are defined by Nielson, Inc. as follows:

Urban

ZIP codes associated with dense neighborhoods that represent the central cities of most major metropolitan areas (more than 4,150 persons/square mile).

2nd City ZIP codes associated with moderate-density neighborhoods in population centers (more than 1,000 and fewer than 4,150 persons/square mile).

Suburban ZIP codes associated with moderate-density neighborhoods that are not surrounded by urban or second-city population centers (estimated to be more than 1,000 persons/square mile and not in an urban or 2nd city population center).

Town or Rural ZIP codes associated with isolated small towns or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950 persons/square mile). Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile).

In order to create the UR_CLRT6 variable, the ZIP code for each case (within the BESTZIP variable) is assigned to its corresponding urbanization category as provided by Nielson. For cases with missing BESTZIP data, the ZIP code the respondent reports in questionnaire item AM7 is used in order to make this assignment (if AM7 > 90001). In addition, some respondents report the ZIP code of a PO Box location rather than a ZIP code for a residence. Nielson Inc. provided the “parent ZIP codes” for these PO Box locations. The urbanization categories assigned to the “parent” ZIP codes are used to classify these cases. Beginning in 2017, we expanded the number of categories to 6, to include new categories for “Mixed”, “Town”, and “Rural”. The “Mixed” category is derived from the previous 2nd City category and denotes locations that can be categorized as a mix between 2nd City and Suburban. The remaining previous 2nd City cases are still categorized as 2nd City. Additionally, the previous “Town or Rural” category has been split into two new categories: one for “Town” as defined by “ZIP codes associated with isolated small towns or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950


72

persons/square mile)”; and one for “Rural” as defined by “Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile)”. The cases with no ZIP code information are imputed to assign a value to UR_CLRT6. UR_TRACT6 Rural and Urban – Claritas (By Census Tract) (6 Levels) The UR_TRACT6 variable assigns the definition of rural and urban tract from the commercial company, Nielson, Inc (formerly Claritas). We obtained a file from Nielson, Inc that contains the tracts in California and their associated urbanization categories. Nielson assigns the tracts in California to 4 urbanization categories based on the analysis of population density grids using projected geoboundaries, redistricting updates, and population estimates. The urbanization categories are defined by Nielson, Inc as follows:

Urban

Tracts associated with dense neighborhoods that represent the central cities of most major metropolitan areas (more than 4,150 persons/square mile).

2nd City Tracts associated with moderate-density neighborhoods in population centers (more than 1,000 and fewer than 4,150 persons/square mile).

Suburban Tracts associated with moderate-density neighborhoods that are not surrounded by urban or second-city population centers (estimated to be more than 1,000 persons/square mile and not in an urban or 2nd city population center).

Town or Rural Tracts associated with isolated small towns or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950 persons/square mile). Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile).

In order to create the UR_TRACT6 variable, the tract for each case is assigned to its corresponding urbanization category as provided by Nielson. For cases with missing tract data, the tract of the respondent reports in questionnaire items AM8 and AM9 is used in order to make this assignment. Beginning in 2017, we expanded the number of categories to 6, to include new categories for “Mixed”, “Town”, and “Rural”. The “Mixed” category is derived from the previous 2nd City category and denotes locations that can be categorized as a mix between 2nd City and Suburban. The remaining previous 2nd City cases are still categorized as 2nd City. Additionally, the previous “Town or Rural” category has been split into two new categories: one for “Town” as defined by “Tracts associated with isolated small towns or less-developed areas on the exurban frontier (estimated to be more than 210 but fewer than 950 persons/square mile)”; and one for “Rural” as defined by “Small villages and rural hamlets surrounded by productive farmland or wide-open spaces (estimated to be 210 or fewer persons/square mile)”. The cases with no tract information are imputed to assign a value to UR_TRACT6.


73

OTHER CONSTRUCTED VARIABLES AM38_P1 Main Reason for Last Move (PUF 1-Year Recode) AM38_P1 is a recoded variable of AM38 that combines AM38= 3 and 4 to create AM38_P1=3 collapsed category (For Child’s Education/To Attend or Leave College) among respondents who have lived at their current address for less than 5 years.

HHSIZE_P Household Size (PUF Recode) The HHSIZE_P variable is a recoded variable based on the HH_SIZE variable that measures household size. The purpose of the household size variable is to combine the number of adults, children, and adolescents in the selected household. The HHSIZE_P variable is created by adding together counts derived from the temporary variables ADLTCNT, CHLDCNT, and TEENCNT. Note 1: Top code: 10 Note 2: This variable is available in the CHIS 2017-2018 Two-Year file only. HHSIZE_P1 Household Size (PUF 1-Year Recode) The HHSIZE_P1 variable is a recoded variable based on the HH_SIZE variable that measures household size. The purpose of the household size variable is to combine the number of adults, children, and adolescents in the selected household. The HHSIZE_P1 variable is created by adding together counts derived from the temporary variables ADLTCNT, CHLDCNT, and TEENCNT. Note: Top code: 10

PROXY Proxy Interview The PROXY variable provides an indicator of whether or not the respondent participated in a proxy interview on behalf of another household member. Respondents serving as proxies are assigned the value PROXY=1. Non-proxy interviews are assigned a value of PROXY=2.

TIMEAD_P1 Length of Time Lived at Current Address (Months) (PUF 1 Year Recode)

The TIMEAD_P1 variable is derived from TIMEAD, which comes from questionnaire item AM14. TIMEAD_P1 is a PUF recode of TIMEAD. This variable is a categorized version of the continuous variable TIMEAD, and it measures the length of time in months the adult respondent has lived at his/her current address. The first 12 response categories are individual months (i.e. 1=1 Month, 2=2 Months, etc.). After this, responses are grouped in multiple month categories, using 60 month intervals, as follows:

Condition: TIMEAD_P1 Value:

TIMEAD_P1 Label:

If TIMEAD = 13 months – 60 months

13 13-60 Months

If TIMEAD = 61 months – 120 14 61-120 Months


74

months If TIMEAD = 121 months – 180 months

15 121-180 Months


16 181-240 Months


17 241-300 Months


18 301-360 Months


19 361-420 Months


20 421-480 Months


21 481-540 Months


22 541-600 Months

If TIMEAD = 601 months or more

23 >600 Months

SRTENR Self-Reported Household Tenure (HH) SRTENR is a data collection vendor-generated variable and is constructed using the adult questionnaire item, AK25. Adult respondents who own their households are assigned a value of SRTENR=1. Adult respondents who rent their households are assigned a value of SRTENR=2. For missing values, AK25 is imputed. For this imputation method, please consult methodology reports by the CHIS data collection vendor.

INTVLANG Language of Interview The INTVLANG variable indicates the language spoken during the interview by the interviewer and the respondent. This variable was derived from ENGLSPAN (Source Data). Each case is reassigned to an INTVLANG value based on the following criteria:

Condition: INTVLANG Value:

INTVLANG Label:

If ENGLSPAN=1 1 English If ENGLSPAN=2 2 Spanish If ENGLSPAN=3 3 Vietnamese If ENGLSPAN=4 4 Korean If ENGLSPAN=5 5 Cantonese If ENGLSPAN=6 6 Mandarin

INTVLNGC_P1 Language of Interview (PUF 1-Year Recode) INTVLNGC_P1 is a three-level variable constructed from the variable INTVLANG, which displays the language that the child portion of the survey was conducted in.

Condition:

INTVLNGC_P1 Value:

INTVLNGC_P1 Label:


75

If INTVLANG>=3 3 Other Language If INTVLANG=1 1 English If INTVLANG=2 2 Spanish

AHCHLDC Amount Per Week Paid For Child Care Adults who have children under 14 and used paid child care in the last month (AH44A) are asked in the questionnaire item to report how much they paid for these services. Respondents reported the amount they paid in the last month or the amount they paid in a typical week (AH44B). In order to convert the amount paid in the last month to a weekly unit, the dollar amount (in whole numbers) was divided by 4.33 and rounded to the nearest whole number. Another reason for creating this variable was to present the amounts paid for childcare in range levels. Please see the data dictionary for details. A few respondents reported that they used paid childcare in the past month, but they did not make a payment. These cases are assigned a value of zero for this variable. Adults without children under 14, and those who did not use paid childcare in the last month, are assigned a skip value (-1) for this variable. Remaining respondents are imputed to assign a value. The below table shows the conversion from (converted) weekly dollar amount to AHCHLDC category:

Condition: AHCHLDC Value: AHCHLDC Label:

If childcare amount >=0 and <10 1 $0-9/week If childcare amount >=10 and <25 2 $10-24/week If childcare amount >=25 and <50 3 $25-49/week If childcare amount >=50 and <100 4 $50-99/week If childcare amount >=100 and <200 5 $100-199/week If childcare amount >=200 6 $200+/week

ACHLDC_P1 Amount Per Week Paid for Child Care (PUF 1-Year Recode) Adults who have children under 14 and used paid child care in the last month (AH44A) are asked in the questionnaire item to report how much they paid for these services. Respondents reported the amount they paid in the last month or the amount they paid in a typical week. In order to convert the amount paid in the last month to a weekly unit, the dollar amount (in whole numbers) was divided by four and rounded to the nearest whole number. Another reason for creating this variable was to present the amounts paid for childcare in range levels. Please see the data dictionary for details. A few respondents reported that they used paid childcare in the past month, but they did not make a payment. These cases are assigned a value of zero for this variable. Adults without children under 12, and those who did not use paid childcare in the last month, are assigned a skip value (-1) for this variable. Remaining respondents are imputed to assign a value.


76

PC_NEWP Primary care: have difficulty finding a provider that accepts new patient

PC_NEWP, first created in CHIS 2013, is equal to variable AJ134. AJ134 displays respondents who responded that a doctor’s office told the respondent that they would not take the respondent as a new patient in the past year.

PC_INS Primary care: have difficulty finding a provider that accepts their insurance

PC_INS, first created in CHIS 2013, is equal to variable AJ135. AJ135 displays currently insured respondents who responded that a doctor’s office told the respondent they did not take the respondent’s main health insurance in the past year.

SC_NEWP Specialty care: have difficulty finding a provider that accepts new patient

PC_NEWP, first created in CHIS 2013, is equal to variable AJ138. AJ138 displays respondents who responded that a medical specialist’s office told the respondent that they would not take the respondent as a new patient in the past year.

SC_INS Specialty care: have difficulty finding a provider that accepts their insurance

SC_INS, first created in CHIS 2013, is equal to variable AJ139. AJ139 displays currently insured respondents who responded that a medical specialist’s office told the respondent they did not take the respondent’s main health insurance in the past year.

VOTE_REG Currently Registered to Vote VOTE_REG is constructed from questionnaire variables AP70 and AP71 to display whether respondents are currently registered to vote. VOTE_REG category values are assigned as follows:

Condition:

VOTE_REG Value:

VOTE_REG Label:

If AP70=1 1 Yes If AP70=2 or 3 2 No/Not Sure If AP70=4 or -1 OR If AP71=6

3 Not Eligible

VOTE_RSNNOT Main Reason Not Registered to Vote VOTE_RSNNOT is constructed from questionnaire variables AP70 and AP71 to display the main reason why a respondent is not registered to vote. VOTE_RSNNOT category values are assigned as follows:

Condition:

VOTE_RSNNOT Value:

VOTE_RSNNOT Label:

If AP71=1 1 Too Busy


77

If AP71=2 2 Doesn’t Make a Difference If AP71=3 or 4 3 Don’t Know How/Where to

Register If AP71=7 or 8 4 Don’t Know Enough About

Candidates/Issues If AP71=9 or 11 5 Don’t Like Candidates/Don’t

Like Politics If AP71=12 6 Not Interested or Motivated

to Vote If AP71=6 OR If AP70=4 or -1

7 Not Eligible

If AP71=5 or 10 or 13 8 Other If AP71=-1 -1 Inapplicable

VOTE_PRES16 Voted in 2016 Presidential Election VOTE_PRES16 is constructed from questionnaire variables AP70, AP71, and AP72 to display whether respondents voted in the 2016 presidential election. VOTE_PRES16 also uses AA1YR and AA1MON to determine eligibility to vote in 2016 based on age. The variable’s category values are assigned as follows:

Condition:

VOTE_PRES16 Value:

VOTE_PRES16 Label:

If AP72=1 1 Yes If AP72=2 2 No If AP70=4 or -1 OR If AP71=6

3 Not Eligible

If AA1YR>1998 OR If AA1YR=1998 and AA1MON>=11

3 Not Eligible

VOTE_ENG1 Voter Engagement VOTE_ENG1 is constructed from questionnaire variables AP73, AP74, and AP75 designed to display respondent voters’ level of voting engagement. The variable only applies to those who are eligible to vote in at least one type of election (presidential, state, local). Otherwise, respondents are designated as inapplicable for the VOTE_ENG1 variable. VOTE_ENG1’s category values are assigned as follows:

Condition:

VOTE_ENG1 Value:

VOTE_ENG1 Label:

If AP73=1 and If AP74=1 and If AP75=1

1 Always Engaged

If (AP73=1) + (AP74=1) + (AP75=1) = 2

2 Frequently Engaged

If AP73=3 and If AP74=3 and If AP75=3

4 Never Engaged

All other combinations of responses to AP73, AP74,

3 Sometimes Engaged


78

and AP75 If AP73=-1 and If AP74=-1 and If AP75=-1

-1 Inapplicable

Constructed Variables CHIS 2017-2018 Adult...

Documents

Transcript of Constructed Variables CHIS 2017-2018 Adult...