Multiple Indicator Cluster Surveys Data Dissemination and Further Analysis Workshop

Multiple Indicator Cluster SurveysData Dissemination and Further Analysis Workshop

Sample and Survey Characteristics, Data Quality and Sampling Error Tables in

MICS ReportsMICS4 Data Dissemination and Further Analysis Workshop

Response rates and background characteristics

• Set of 6 tables that:

• Presents sample coverage and characteristics of households and respondents

• Age and sex distribution of survey population

• Characteristics of Respondents

Table HH.1: Results of household, women's, men's and under-5 interviewsNumber of households, women, men, and children under 5 by results of the household, women's, men's and under-5's interviews, and household, women's, men's and under-5's response rates, Country, Year

Residence Region Urban Rural Region 1 Region 2 Region 3 Region 4 Region 5 Total

Households Sampled Occupied Interviewed Household response rate

Women Eligible Interviewed Women's response rate Women's overall response rate

Men Eligible Interviewed Men's response rate Men's overall response rate

Children under 5 Eligible Mothers/caretakers interviewed Under-5's response rate Under-5's overall response rate

The denominator for the household response rate is the number of households found to be occupied during fieldwork (HH9 = 01, 02, 04, 07); the numerator is the number of households with complete household questionnaires (HH9 = 01). The denominator for the women’s response rate is the number of eligible women enumerated in the household listing form (HH12); the numerator is the number of women interviewed (HH13). The denominator for the men's response rate is the number of eligible men enumerated in the household listing form (HH13A); the numerator is the number of men interviewed (HH13B). The denominator for the response rate for the questionnaire for children under 5 is the number of under-5 children identified in the household listing form (HH14); the numerator is the number of complete questionnaires for children under 5 (HH15).

Overall response rates are calculated for women, men and under-5's by multiplying the household response rate with the women's, men's and under-5's response rates, respectively.

Table HH.2: Household age distribution by sexPercent and frequency distribution of the household population by five-year age groups, dependency age groups, and by child (age 0-17 years) and adult populations (age 18 or more), by sex, Country, Year

Males Females Total

Number Percent Number Percent Number PercentAge

0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64

65-69 70-74 75-79 80-84 85+ Missing/DK

Dependency age groups

0-14 15-64 65+ Missing/DK

Child and adult populations Children age 0-17 years

Adults age 18+ years Missing/DK

Total 100.0 100.0 100.0

Table HH.3: Household compositionPercent and frequency distribution of households by selected characteristics, Country, Year

Weighted percentNumber of households

Weighted UnweightedSex of household head Male Female

Region Region 1 Region 2 Region 3 Region 4 Region 5

Residence Urban Rural

Number of household members 1 2 3 4 5 6 7 8 9 10+

Education of household head None Primary Secondary Higher

Religion/Language/Ethnicity of household head Group 1 Group 2 Group 3

Total 100.0 Households with at least One child age 0-4 years One child age 0-17 years One woman age 15-49 years One man age 15-59 years

Mean household size

Total weighted and unweighted numbers of households should be equal when normalized sample weights are used.

Table HH.5: Under-5's background characteristicsPercent and frequency distribution of children under five years of age by selected characteristics, Country, Year

Weighted percentNumber of under-5 children

Weighted UnweightedSex Male Female


Residence Urban Rural

Age 0-5 months 6-11 months 12-23 months 24-35 months 36-47 months 48-59 months

Mother’s education* None Primary Secondary Higher

Wealth index quintile Poorest Second Middle Fourth Richest


Total 100.0

* Mother's education refers to educational attainment of mothers and caretakers of children under 5.

Total weighted and unweighted numbers of children under 5 should be equal when normalized sample weights are used.

Data quality tables

• One of the MICS primary goals is to produce high quality, statistically sound and internationally comparable estimates of indicators.

• The quality of MICS data is assured by several processes: • Recommended training and field work supervision• Double data entry, consistency checks, secondary editing• Field check tables generated on a regular bases with goal to indicate

potential problems in the field, etc.

Data quality tables

• After field work is completed 16 tables are produced for assessment of data quality.

• Intended to check distributions, heaping, understatement or overstatement, sex ratios, eligibility and coverage, out-transference of eligible persons, the extent of missing information, outliers, sex ratios, quality of anthropometric measurements.

• Useful for understanding quality issues, familiarity with issues in data sets, indicative of the quality of training and implementation.

Table DQ.1: Age distribution of household populationSingle-year age distribution of household population by sex, Country, Year Males Females Males Females Number Percent Number Percent Number Percent Number Percent

0 45

1 46

2 47

3 48

4 49

5 50

6 51

7 52

8 53

9 54

10 55

11 56

12 57

….. ….

37 82

38 83 39 84 40 85+ 41 42 DK/Missing 43 44 Total 100.0 100.0

Typical data quality issues: Heaping on ages with digits ending with 0 and 5. If age reporting is good, the distribution should be smooth. The table should also provide insights into overreporting or underreporting at certain age groups or intervals, and the extent of missing information on age. Deficits at ages 4, 15, and 49, excesses at ages 5 and 6, 14, and 50 might be indicative of out-transference of ages to avoid administering individual questionnaires.

Age distribution of household population, example country, 2010

Table DQ.2: Age distribution of eligible and interviewed womenHousehold population of women age 10-54, interviewed women age 15-49, and percentage of eligible women who were interviewed, by five-year age groups, Country, Year

Household population of women age 10-54

yearsInterviewed women age

15-49 years

Percentage of eligible women

interviewed (Completion rate)Number Number Percent

Age 10-14 na na na15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 na na na

Total (15-49) 100.0 Ratio of 50-54 to 45-49

Typical data quality issues: In countries with growing populations, the percentages in each age group of women should decline with age (Column B). The last column shows whether the survey was equally effective in interviewing women in all age groups - typically, some surveys fail to interview the younger women, sometimes because of problems in sample implementation, sometimes because of interviewers' reluctance to interview young women. These figures should be high, preferably over 95 percent, or at least 90 percent, and should not vary much by age. The distribution in Column D should be similar to the distribution in Column B

If completion rates vary greatly by age and fall below 85 percent in 2 or 3 groups, say for groups age 15 to 24, it may be necessary to re-calculate sample weights by taking age-specific non-response into account. Failure to do so may lead to biased estimates of indicators which typically vary by age of women.

Weights used for both household population of women (Column B) and interviewed women (Column D) are household weights. Age is based on the household schedule. Table should be run unweighted if major problems are identified.

Table DQ.2: Age distribution of eligible and interviewed women

Household population of women age 10-54, interviewed women age 15-49, and percentage of eligible women who

were interviewed, by five-year age groups, example country, year

Household population of women age 10-54

Interviewed women age 15-49

Percentage of eligible women

interviewed (Completion

rate)Number Number Percent

Age 10-14 6011 . . .15-19 3950 3318 20.8 84.020-24 3423 3011 18.9 88.025-29 3418 3073 19.3 89.930-34 2607 2350 14.7 90.235-39 2104 1919 12.0 91.240-44 1473 1289 8.1 87.545-49 1121 1000 6.3 89.250-54 1407 . . .

Total (15-49) 18095 15959 100.0 88.2Ratio of 50-54 to 45-49 1.26

Age distribution of eligible and interviewed women, example country, 2010

Table DQ.3: Age distribution of under-5s in household and under-5 questionnaires

Household population of children age 0-7, children age 0-4 whose mothers/caretakers were interviewed, and percentage of under-5 children whose mothers/caretakers were interviewed, by single ages, Country, Year

Household population of

children 0-7 years Interviewed under-5 children Percentage of eligible under-5s interviewed

(Completion rate) Number Number Percent Age

0 1 2 3 4 5 na na na6 na na na7 na na na

Total (0-4) 100.0 Ratio of 5 to 4

Typical data quality issues: In countries with growing populations, the numbers of children at each age (Column B) should be declining, The table is intended to provide information on the efficiency of the survey in collecting information on under-5s. Distribution of children by age in the household questionnaire should be smooth, with little or no heaping on age 5. Heaping on age 5 may be indicative of out-transference of children age 0-4 to outside the eligibility range. Percentages in the last column (completion rates) should be over 90, preferably over 95.

Weights used for both household population of children and under-5 interviews are household weights. Age is based on the household schedule. Table should be run unweighted if major problems are identified.

Table DQ.4: Women's completion rates by socio-economic characteristics of householdsHousehold population of women age 15-49, interviewed women age 15-49, and percentage of eligible women who were interviewed, by selected social and economic characteristics of the household, Country, Year

Household population of women

age 15-49 yearsInterviewed women age 15-49

yearsPercent of eligible women interviewed (Completion

rates) Number Percent Number Percent Region

Region 1 Region 2 Region 3 Region 4 Region 5

Area Urban Rural

Household size 1-3 4-6 7+

Education of household head None Primary Secondary +

Wealth index quintiles Poorest Second Middle Fourth Richest


Total 100.0 100.0

Typical data quality issues: Completion rates by socio-economic background characteristics should be similar across socio-economic groups. In cases when completion rates vary greatly by background characteristics, the sample may be biased.

Completion rates by regions and urban-rural residence are reflected in sample weights when the sample design is based on regions and urban-rural residence. While this "corrects" for differential completion rates by these characteristics, it does not necessarily mean that the sample is no longer biased in terms of other socio-economic characteristics.

Weights for both household population of women and interviewed women are household weights. Table should be run unweighted if major problems are identified.

Table DQ.5: Completion rates for under-5 questionnaires by socio-economic characteristics of householdsHousehold population of under-5 children, under-5 questionnaires completed, and percentage of under-5 children for whom interviews were completed, by selected socio-economic characteristics of the household, Country, Year


under-5 childrenInterviewed under-

5 children Percent of eligible under-5s with completed under-5 questionnaires (Completion rates) Number Percent Number Percent


Area Urban Rural

Household size 1-3 4-6 7+

Education of household head None Primary Secondary +



Total 100.0 100.0

Typical data quality issues: Completion rates by socio-economic background characteristics should be similar across socio-economic groups. In cases when completion rates vary greatly by background characteristics, the sample may be biased.

Completion rates by regions and urban-rural residence are reflected in sample weights when the sample design is based on regionsand urban-rural strata. While this "corrects" for differential response rates by these characteristics, it does not necessairly mean that the sample is no longer biased in terms of other socio-economic characteristics.

Weights for both household population of children and interviewed children are household weights. Table should be run unweighted if major problems are identified.

Table DQ.5: Completion rates for under-5 questionnaires by socio-economic characteristics of households

Household population of under-5 children, under-5 questionnaires completed, and percentage of under-5 children

for whom interviews were completed, by selected socio-economic characteristics of the household, example country,

2010


under-5 childrenInterviewed

under-5 children

Percent of eligible under-5s with

completed under-5 questionnaires

(Completion rates)Area Urbain 3764 20.9 3532 20.5 93.8

Rural 14235 79.1 13728 79.5 96.4Household size

1-3 3599 20.0 1425 8.3 96.04-6 9279 51.6 7258 42.0 96.67+ 5120 28.4 8577 49.7 95.3

Mother's education

None 10978 61.0 10508 60.9 95.7Primary 3878 21.5 3733 21.6 96.3Secondary + 3027 16.8 2905 16.8 96.0Missing/DK 116 .6 114 .7 98.1

Wealth index quintiles

Poorest 3490 19.4 3379 19.6 96.8Second 3710 20.6 3561 20.6 96.0Middle 3828 21.3 3704 21.5 96.8Fourth 3799 21.1 3652 21.2 96.1Richest 3172 17.6 2965 17.2 93.5

Total 17999 100.0 17260 100.0 95.9

Completion rates for under-5 questionnaires by socio-economic characteristics of households, example country, 2010

Table DQ.6: Completeness of reportingPercentage of observations that are missing information for selected questions and indicators, Country, Year

Questionnaire and type of missing information Reference group

Percent with missing/incomplete

information*Number of

casesHousehold

Age All household members Salt test result All households interviewed that have salt Starting time of interview All households interviewed Ending time of interview All households interviewed

Women

Woman's date of birth All women age 15-49 Only month Both month and year

Date of first birth All women age 15-49 with at least one live birth Only month Both month and year

Completed years since first birth All women age 15-49 with at least one live birth with year of first birth unknown

Date of last birth All women age 15-49 with a live birth in last 2 years Only month Both month and year

Date of first marriage/union All ever married women age 15-49 Only month Both month and year

Age at first marriage/union All ever married women age 15-49 with year of first marriage not known Age at first intercourse All women age 15-24 who have ever had sex Time since last intercourse All women age 15-24 who have ever had sex Starting time of interview All women interviewed Ending time of interview All women interviewed

Under-5

Date of birth All under-5 children Only month Both month and year

Anthropometric measurements All under-5 children Weight Height Both weight and height

Starting time of interview All under-5 children Ending time of interview All under-5 children

* Includes "Don't know" responses

Typical data quality issues: Surveys always have cases with missing information. The extent of missing information is important, because it can result in biased results if such proportions are high. Particularly informative about the quality of survey is the extent of missing information on measurements, ages, and dates of events.

Completeness of reporting, example country, year

Under 5 questionnaireWomen questionnaire

Table DQ.7: Completeness of information for anthropometric indicatorsDistribution of children under 5 by completeness of information for anthropometric indicators, Country, Year

Valid weight and date of birth

Reason for exclusion from analysis

Total

Percent of children excluded from

analysisNumber of

children under 5 Weight not measured

Incomplete date of birth

Weight not measured, incomplete date of birth

Flagged cases (outliers)

Weight by age <6 months 100.0 6-11 months 100.0 12-23 months 100.0 24-35 months 100.0 36-47 months 100.0 48-59 months 100.0

Total

Valid height and date of birth


Total


analysisNumber of

children under 5 Height not measured

Incomplete date of birth

Height not measured, incomplete date of birth


Height by age <6 months 100.0 6-11 months 100.0 12-23 months 100.0 24-35 months 100.0 36-47 months 100.0 48-59 months 100.0

Total

Valid weight and height


Total


analysisNumber of

children under 5 Weight not measured

Height not measured

Weight not measured, height not measured


Weight by height <6 months 100.0 6-11 months 100.0 12-23 months 100.0 24-35 months 100.0 36-47 months 100.0 48-59 months 100.0

Total

Typical data quality issues:Under-5 children may be excluded from anthropometric analysis due to a number of reasons. Column B shows the percentage of under-5 children who are included in anthropometric analysis for each of the three anthropometric indicators (underweight, stunting and wasting). Both in terms of the total rows and across age groups, these percentages should be above 90 percent, preferably 95 percent. Column H shows the percentage of under-5 children excluded from analyses.

Completeness of information for anthropometric indicators, example country, year

Table DQ.7: Completeness of information for anthropometric indicatorsDistribution of children under 5 by completeness of information for anthropometric indicators, country, year

Valid weight and

date of birth


Total

Percent of children excluded

from analysis

Number of children under 5

Weight not measured

Incomplete date of

birth

Weight not measured, incomplete

date of birth

Flagged cases

(outliers)Weight by age <6 months 84.2 .3 .3 .0 15.1 100.0 15.8 304

6-11 months 90.0 .0 .0 .0 10.0 100.0 10.0 35012-23 months 87.5 .4 .1 .0 12.0 100.0 12.5 71124-35 months 83.2 .3 .5 .0 16.1 100.0 16.8 65436-47 months 82.4 .9 .6 .0 16.2 100.0 17.6 69748-59 months 84.3 .2 .9 .0 14.7 100.0 15.7 586

Total 84.9 .4 .4 .0 14.2 100.0 15.1 3302

Valid weight and

date of birth


Total

Percent of children excluded

from analysis


Weight not measured

Incomplete date of

birth

Weight not measured, incomplete

date of birth

Flagged cases

(outliers)Weight by age <6 months 86.8 .1 3.8 .0 9.4 100.0 13.2 1867

6-11 months 86.7 .3 6.0 .0 6.9 100.0 13.3 161512-23 months 82.1 .1 11.1 .0 6.7 100.0 17.9 296424-35 months 72.8 .1 18.5 .0 8.6 100.0 27.2 342136-47 months 69.5 .1 21.6 .1 8.7 100.0 30.5 367048-59 months 65.8 .1 24.1 .0 10.0 100.0 34.2 3469

Total 75.1 .1 16.2 .0 8.5 100.0 24.9 17006

Example 1

Example 2

Table DQ.8: Heaping in anthropometric measurements

Distribution of weight and height/length measurements by digits reported for decimals, Country, Year Weight Height or length

Digits Number Percent Number Percent0 1 2 3 4 5 6 7 8 9 0 or 5 Total 100.0 100.0

Typical data quality issues: Under normal circumstances, approximately 10 percent of anthropometric measurements should be reported for each of the digits for the decimals. Significant excesses over 10 percent are indicative of heaping, and therefore quality problems in anthropometric measurements, either due to truncation or rounding.

Typically, more heaping is expected in height/length than weight measurements.

Table DQ.8: Heaping in anthropometric measurementsDistribution of weight and height/length measurements by digits

reported for decimals, country, year

Weight HeightNumber Percent Number Percent

Digits 0 1997 12.8 4234 27.11 1501 9.6 1213 7.82 1638 10.5 1885 12.13 1483 9.5 1527 9.84 1391 8.9 1162 7.45 1486 9.5 1876 12.06 1491 9.6 1004 6.47 1567 10.0 1004 6.48 1507 9.7 825 5.39 1539 9.9 887 5.70 or 5 3483 22.3 6110 39.1Total 15600 100.0 15617 100.0

Heaping in anthropometric measurements, example country, year

Table DQ.9: Observation of bednets and places for hand washingPercentage of bednets in all households interviewed observed by the interviewer, and percentage of places for handwashing observed by the interviewer in all interviewed households, Country, Year

Percentage of bednets

observed by interviewer

Total number of bednets

Place for handwashing

Total

Number of households interviewed

Observed

Not observed

Not in the dwelling, plot or

yardNo permission

to see OtherRegion


Area Urban Rural


Total

Typical data quality issues: Interviewers are required to observe and record the type of bednets in households. Observation of bednets is likely to lead to improved data quality. Interviewers are also required to observe the place for handwashing for the presence of water and soap. Both Columns B and D should not be less than 90 percent.

Household members may be reluctant to let interviewers observe places for handwashing or bednets in the rooms of the house, particularly bedrooms. This might in turn be related to cultural and social characteristics of the households. For this reason, percentages of bednets and places for handwashing are provided here by regions and urban-rural areas in this table.

Table DQ.10: Observation of women's health cardsPercent distribution of women with a live birth in the last 2 years by presence of a health card, and the percentage of health cards seen by the interviewers, Country, Year

Woman does not

have health card

Woman has health card

Missing/DK Total

Percent of health cards seen by the interviewer

(1)/(1+2)*100

Number of women with a live birth in the last two years

Seen by the

interviewer (1)

Not seen by the

interviewer (2)

Region Region 1 100.0 Region 2 100.0 Region 3 100.0 Region 4 100.0 Region 5 100.0

Area Urban 100.0 Rural 100.0

Wealth index quintiles Poorest 100.0 Second 100.0 Middle 100.0 Fourth 100.0 Richest 100.0

Total 100.0

Typical data issues: Interviewers are required to ask respondents if they have health cards, and if so, ask to see these cards (MN5 in Women;s Questionnaire). These cards are then used by the interviewer to record information on tetanus toxoid vaccinations during pregnancy, or any other useful information on the card. Observation of cards is likely to improve the quality of information collected, as the data collected becomes less dependent on the recall of the respondent.

Table DQ.11: Observation of under-5s birth certificates

Percent distribution of children under 5 by presence of birth certificates,and percentage of birth calendar seen, Country, Year

Child does not

have birth certificate

Child has birth certificate

Don't know/Missing Total

Percent of birth certificates seen

by the interviewer (1)/(1+2)*100

Number of children

under age 5 Seen by the

interviewer (1)Not seen by the interviewer (2)

Region Region 1 100.0 Region 2 100.0 Region 3 100.0 Region 4 100.0 Region 5 100.0


Child's age 0 100.0 1 100.0 2 100.0 3 100.0 4 100.0

Total 100.0

Typical data quality issues: Interviewers are required to ask and see the birth certificates of children. This is important for the completion of the Birth Registration module in the Under-5 questionnaire, but may also be useful for obtaining accurate information on children's dates of birth and ages.

Percent of birth certificates seen by the interviewer (Column G) are desired to be as high as possible, preferably over 90 percent.

Table DQ.12: Observation of vaccination cardsPercent distribution of children under 5 by presence of a vaccination card, and the percentage of vaccination cards seen by the interviewers, Country, Year

Child does not have

vaccination cardChild has

vaccination card

Total

Percent of vaccination cards

seen by the interviewer

(1)/(1+2)*100

Number of children

under age 5

Had vaccination

card previously

Never had vaccinatio

n card

Seen by the

interviewer (1)

Not seen by the

interviewer (2)

Don't know/Miss

ingRegion

Region 1 100.0 Region 2 100.0 Region 3 100.0 Region 4 100.0 Region 5 100.0


Child's age 0 100.0 1 100.0 2 100.0 3 100.0 4 100.0

Total 100.0

Typical data quality issues: Interviewers are required to ask to see the vaccination cards of under-5s from the respondent, and copy the information on the cards to the under-5 questionnaire. Information on vaccination cards is believed to be more accurate than information that would be provided by mothers or caretakers, in the absence of vaccination cards. Percentages in Column G is desired to be as high as possible.

Particularly important are the results for children age 1, as immunization indicators are based on these children in most countries.

Observation of vaccination cards, example country, 2010

Table DQ.13: Presence of mother in the household and the person interviewed for the under-5 questionnaire

Distribution of children under five by whether the mother lives in the same household, and the person interviewed for the under-5 questionnaire, Country, Year Mother in the household Mother not in the household

Total


Mother interviewed

Father interviewed

Other adult female

interviewed

Other adult male

interviewed Father

interviewed

Other adult female

interviewed

Other adult male

interviewedAge

0 100.0 1 100.0 2 100.0 3 100.0 4 100.0

Total 100.0

Typical data quality issues: The under-5 questionnaire should be administered to the mother, if the mother is listed the household roster. The table is informative on whether the questionnaire was administered to the right person during the fieldwork. Not all information will have been collected from mothers, but cases where the mother is in the household but somebody else was interviewed can be problematic (Columns C, D, and E).

"Adult" males and females are defined as those age 15 and above.

Table DQ.14: Selection of children age 2-14 years for the child discipline module

Percent of households with at least two children age 2-14 years where correct selection of one child for the child discipline module was performed, Country, Year

Percent of households where correct selection was

performed

Number of households with 2 or more children age 2-14

yearsRegion


Area Urban Rural

Number of children age 2-14 years 2 3 4 5+

Total

Typical data quality issues: In households where 2 or more children age 2-14 years live, interviewers are required to select, according to pre-determined random selection procedures, one child for the child discipline module. Percentages with correct selection should be close to 100.0

Table DQ.14: Selection of children age 2-14 years for the child discipline module

Percent of households with at least two children age 2-14 years where correct selection of one child for the child discipline module was performed, country, year

Percent of households

where correct selection was

performed

Number of households with 2 or more children

age 2-14 yearsArea Urban 83.5 3656

Rural 85.5 6183Number of households by number of children 2-14

2 89.3 27463 89.1 24304 79.8 4663

Total 84.7 9839

Selection of children age 2-14 years for the child discipline module,example country, 2010

Table DQ.15: School attendance by single ageDistribution of household population age 5-24 by educational level and educational level and grade attended in the current (or most recent) school year, Country, Year Currently attending

Number of household members

Not attending

schoolPrescho

ol

Primary schoolGrade

Secondary schoolGrade

Higher than

secondaryMissing/

DK 1 2 3 4 5 6 1 2 3 4 5 6 TotalAge at beginning of school year

5 100.0 6 100.0 7 100.0 8 100.0 9 100.0 10 100.0 11 100.0 12 100.0 13 100.0 14 100.0 15 100.0 16 100.0 17 100.0 18 100.0 19 100.0 20 100.0 21 100.0 22 100.0 23 100.0 24 100.0

Typical data quality issues: The table could be used to look at outliers. Data entry programs do not check age versus educational grade in detail. If data has been collected and entered correctly, one should see cases concentrated over the diagonal, and should not expect such cases as 22 year old persons attending grades in primary school, very young people at grade 6 of secondary school etc. Many cases outside the diagonal would be indicative of both poor fieldwork supervision, as well as poor data entry and (lack of) verification.

Before running the table, grades should be adapted to the system in the country. The table assumes 6 years of primary school and 6 years of secondary school.

Age at the beginning of the school year is calculated from dates of birth of household members or by rejuvenating household members based on the date of the survey and current age. Levels and grades refer to the current school year, or the most recent school year if data collection was completed between school years.

Table DQ.16: Sex ratio at birth among children ever born and livingSex ratio (number of males per 100 females) among children ever born (at birth), children living, and deceased children, by age of women, Country, Year Children Ever Born Children Living Children Deceased

Number of

women

Number of sons

ever born

Number of

daughters ever born

Sex ratio at birth

Number of sons living

Number of daughters

living Sex ratio

Number of

deceased sons

Number of deceased daughters Sex ratio

Age 15-19 20-24 25-29 30-34 35-39 40-44 45-49

Total

Typical data quality issues: Universally, the sex ratio among live births is around 105 males per 100 females, typically ranging from 103 to 107 in sizeable populations (with the exception of populations where sex-selective abortions is widely practiced). The values in column D should be within these ranges. However, since surveys are influenced by chance fluctuations, one should be looking for systematically low or high ratios in all or most of the age groups (in several countries, very young daughters may not be reported, or deaths of males may not be reported). In most populations, death rates at early ages are higher for males than females - hence, the sex ratios among deceased children (Column L) should also be above 100.

Sampling Error Tables: Background

The sample selected in a survey is one of the many samples that could have been selected (with same design and size).

Sampling errors are measures of the variability between all possible samples, which can be estimated from survey results.


Calculation of sampling errors is very important;- Provides information on the reliability of your results- Tells you the ranges within which your estimates most

probably fall- Provides clues as to the sample sizes (and designs) to be

selected in forthcoming surveys


MICS4 sample designs are complex designs, usually based on stratified, multi-stage, cluster samples.

It is not possible to use straightforward formula for the calculation of sampling errors. Sophisticated approaches have to be used.

Versions 13 and above of SPSS are used for this purpose.

SPSS uses Taylor linearization method of variance estimation for survey estimates that are means or proportions.

This approach is used by most other package programs: Wesvar, Sudaan, Systat, EpiInfo, SAS


In MICS4, the objective is to calculate sampling errors for a selection of variables, for the national sample, as well as selected sub-populations, such as urban and rural areas, and regions.

Table SE.2: Sampling errors: Total sample

r - 2se r + 2se

Household availability of ITNs CH.10 0.048 0.011 0.221 2.371 1.540 1009 967 0.027 0.069Iodized salt consumption NU.5 0.821 0.022 0.027 3.130 1.769 1004 962 0.778 0.865Child discipline CP.4 0.725 0.020 0.028 1.570 1.253 792 757 0.684 0.766

Use of improved drinking water sources EN.1 0.753 0.057 0.075 16.594 4.074 6067 967 0.639 0.866Use of improved sanitation facilities EN.5 0.941 0.015 0.016 3.751 1.937 6067 967 0.912 0.970Net primary school attendance rate ED.3 0.601 0.015 0.025 0.798 0.893 859 819 0.571 0.632Net secondary school attendance rate ED.4 0.810 0.019 0.023 2.192 1.481 1013 968 0.772 0.847Primary completion rate ED.6 0.861 0.032 0.037 1.286 1.134 159 152 0.797 0.925Child labour CP.2 0.111 0.016 0.148 3.759 1.939 1443 1376 0.078 0.143Prevalence of orphans HA.10 0.036 0.006 0.169 2.542 1.594 2533 2417 0.024 0.048Prevalence of vulnerable children HA.11 0.555 0.023 0.041 5.181 2.276 2533 2417 0.509 0.601

Skilled attendant at delivery RH.5 0.958 0.024 0.025 3.515 1.875 256 244 0.910 1.000Antenatal care RH.3 0.926 0.026 0.028 2.363 1.537 256 244 0.874 0.977Contraceptive prevalence RH.1 0.464 0.015 0.032 0.876 0.936 1044 995 0.435 0.494Adult literacy ED.8 0.934 0.013 0.014 1.602 1.266 644 615 0.909 0.959Prevalence of female genital mutilation/cutting (FGM/C) CP.7 0.159 0.009 0.056 0.863 0.929 1541 1471 0.141 0.176Marriage before age 18 CP.5 0.140 0.023 0.165 1.296 1.138 308 294 0.094 0.186Polygyny CP.5 0.212 0.012 0.058 0.905 0.952 1044 995 0.187 0.237Comprehensive knowledge about HIV prevention among young people HA.3 0.037 0.005 0.124 0.866 0.930 1541 1471 0.028 0.046

Standard errors, coefficients of variation, design effects (deff ), square root of design effects (deft ) and confidence intervals for selected indicators, Country, Year

Standard error (se )

Value (r )Table

Confidence limits

Unweighted count

Weighted count

Square root of design

effect (deft )

Design effect (deff )

Coefficient of variation

(se/r )

HOUSEHOLD MEMBERS

WOMEN

HOUSEHOLDS

Standard error is the square root of the variance – a measure of the variability between all possible samples


r - 2se r + 2se






Value (r )Table

Confidence limits

Unweighted count

Weighted count


effect (deft )



(se/r )

HOUSEHOLD MEMBERS

WOMEN

HOUSEHOLDS

Coefficient of variation (relative error) is the ratio of SE to the estimate


r - 2se r + 2se






Value (r )Table

Confidence limits

Unweighted count

Weighted count


effect (deft )



(se/r )

HOUSEHOLD MEMBERS

WOMEN

HOUSEHOLDS

Design effect is the ratio between the SE using the current design and the SE that would result if a simple random sample was used. A DEFT value of 1.0 indicates that the sample is as efficient as a SRS


r - 2se r + 2se






Value (r )Table

Confidence limits

Unweighted count

Weighted count


effect (deft )



(se/r )

HOUSEHOLD MEMBERS

WOMEN

HOUSEHOLDS

Weighted and unweighted counts


r - 2se r + 2se






Value (r )Table

Confidence limits

Unweighted count

Weighted count


effect (deft )



(se/r )

HOUSEHOLD MEMBERS

WOMEN

HOUSEHOLDS

Upper and lower confidence limits are calculated as p +/- 2.SEIndicate the ranges within which the estimate would fall in 95 percent of all possible samples of identical design and size

Comprehensive knowledge about HIVprevention among young people

Thank you

Multiple Indicator Cluster Surveys Data Dissemination and Further Analysis Workshop

Documents

Transcript of Multiple Indicator Cluster Surveys Data Dissemination and Further Analysis Workshop