Biostatistics Computing Project by Samy Uguy fall 2003.

9
Biostatistics Computing Project by Samy Uguy fall 2003

Transcript of Biostatistics Computing Project by Samy Uguy fall 2003.

Page 1: Biostatistics Computing Project by Samy Uguy fall 2003.

Biostatistics Computing Project

by Samy Uguy

fall 2003

Page 2: Biostatistics Computing Project by Samy Uguy fall 2003.

Purpose of the program:

• to read and access SPSS data files in SAS.• to create a SAS program that will link a SAS

data set from time one to a SAS data set at time two.

• to produce a basic descriptive tables (means and percentages).

Page 3: Biostatistics Computing Project by Samy Uguy fall 2003.

Features use in the program

• Store SPSS data files in .por (portable file)• Convert SPSS data files to SAS data files• Create temporary SAS data • Proc contents to check what data we have• Develop a macro for linking subjects based on

selected identifiers• Create new ID (play with numbers)• Select certain choices over various possible

answers for presentation purposes

Page 4: Biostatistics Computing Project by Samy Uguy fall 2003.

Features continue…

• Use .first and .last for data elimination to produce certain data sets

• Create unique, duplicates, and linking only data sets

• Proc tabulate with ods statement and background style for maneuvering table presentations

Page 5: Biostatistics Computing Project by Samy Uguy fall 2003.

List of external files used

• E:\Pre.por

• E:\Post.por

Page 6: Biostatistics Computing Project by Samy Uguy fall 2003.

ID CODING SYSTEM

• region: 9 values (1 digit)

• agency: coded 101 to 199 (3 digits)• program: coded 201 to 299 (3 digits)• info1: number of letters in last name. (1 digit)• info3: day of birth (2 digits)• q1_1 : gender (1 digit)• other possible identifiers: age, grade, eye color,

parents education (not use in this presentation)

Page 7: Biostatistics Computing Project by Samy Uguy fall 2003.

Brief Info on the project

• Youth Tobacco Evaluation Project, which is conducted by VCU, is a State-wide project funded by Virginia Tobacco Settlement Foundation.

• The evaluation examines whether the intervention programs affect participant knowledge, attitudes, skills, and behaviors.

• Questionnaires include more than 150 questions in Pre (time 1) and Post (time 2).

Page 8: Biostatistics Computing Project by Samy Uguy fall 2003.

The project cont.

• Survey is conducted by the grantees/ agencies

• Time lapse from Pre (time 1) to Post (time 2) on average is about 6 months

• Programs are varied (from single session to more than 50 sessions)

Page 9: Biostatistics Computing Project by Samy Uguy fall 2003.

Here is the SAS program itself