SNAPS 1999 Sampling Procedures

Contact Person

The sampling procedures across the system for the Student Needs and Priorities Survey (SNAPS) will be coordinated by Dr. Lawrence A. Jordan, Director, Analytical Studies at California State University, Los Angeles. Dr. Jordan has written a document, entitled "Campus Sampling Procedures-SNAPS 1999" that explains in precise detail how to execute each of three possible sampling designs. This document represents the standard for campus decisions. A copy of it was passed out at the North and South SNAPS training workshops (or mailed to absentee SNAPS coordinators). SNAPS coordinators should contact Dr. Jordan as soon as possible to finalize their sampling strategies. He can be contacted by telephone at (213) 343-2730 or by e-mail: ljordan@calstatela.edu.

Overview of Sampling Procedures

SNAPS should be administered to students in samples of classrooms at each CSU campus. In other words, course sections should be randomly selected and students within targeted classrooms shall constitute the sample of respondents. Thus a list of eligible course sections for a designated term should represent the first stage of the SNAPS sampling frame.

Campuses can construct the sampling frame from either their printed schedule of classes or from a specially constructed, machine-readable list of sections generated from their Student Information System. The latter source is preferred because records for individual course sections can be augmented with classifying information, such as level of instruction.

Course sections should be drawn from three basic sampling strategies:

  1. Unstratified: Select eligible classroom sessions with equal probability from a list of eligible sections.
  2. Proportionate stratification: Classify the eligible sections into selected strata, and determine the target number of sections so that the stratum sizes will be proportional to campus population sizes.
  3. Disproportionate stratification: Classify the eligible sections into selected strata, and determine the target number of sections so small but important strata can be over-sampled.

If the third strategy is elected, sampling weights will be created so over- and under-sampled sections will represent their true proportion in the population.

The 22 campus samples will be united to create a systemwide sample. Records will be re-weighted so that each campus sub-sample is proportionate in size to its systemwide proportion. Campus files containing SNAPS 1999 responses will have sampling weights appended to each record.

 

To create all the necessary sampling weights necessary to create campus files and a systemwide file, a Sampling Frame Report must be built by every campus coordinator and submitted to Dr. Jordan. Everyone should review their reports with Dr. Jordan before they are submitted. The basic format for the Sampling Frame Report is displayed below.
Mnemonic Field Description FORMAT First
Column		 Last
Column

COURSE

Course Number

8X

1

8

SEC

Section Number

3X

9

11

ENR

Scheduled Enrollment

3N

12

14

LEV

Course Level

1X

15

15

CSN

CS Number

3X

16

18

HEGIS

Discipline Code

5X

19

23

CCU

Course Credit Units

3.1N

24

26

STR

Stratum Code

2X

27

28

STRN

Stratum Number of Sections

4N

29

32

STRENR

Stratum Enrollment

5N

33

37

CC

Campus Code

2X

38

39

SC

Section Code

2X

40

41

SAMPN

Obtained Sample Size

4N

42

45

The mode of collecting the surveys is optional. For example: 1) coordinators may administer the questionnaires in class, or 2) hand them out in class but either have them return the completed answer sheets at a later date, or mail the survey to local addresses. But regardless of the mode selected, a Sampling Frame Report must be submitted to Dr. Jordan.