PBA Home > Institutional Research & Analysis > Surveys > Senior Survey 2004 > Method

Senior Survey, Spring 2004

Method

Population and Sample

The population for the 2004 Senior Survey consisted of CU-Boulder seniors who were enrolled in Spring 2004 and who had not requested that their personal directory information be kept private (n = 6,676). A sample of 3,955 seniors was randomly selected and invited to participate in the survey.

In order to provide major/department- and college-level information, a stratified random sampling design was used in which strata consisted of a combination of major, department, and college/school/cluster. For statistical reasons, a minimum of six respondents were required in order to produce major/department-level results. To ensure that we could provide results for all medium-to-large majors/departments, all seniors in majors/departments containing between 60 and 119 seniors were included in the sample, whereas 10% of seniors from relatively small majors/departments (those containing fewer than 60 seniors) were randomly sampled. A maximum of 120 seniors were randomly sampled from all very large majors/departments (those with 120 or more seniors). This procedure is illustrated in the table below.

Even though small majors/departments were less likely to have enough respondents for us to provide results, it was important to sample seniors from them in order to characterize the campus as a whole. Additional information about majors, departments, and colleges/schools/clusters is contained in a table of n-counts for the population, sample, and respondents.

It was possible for a student to be included in the population more than once. For example, a student with two majors would be included twice (once for each major), and a student with three majors would be included three times. Within a particular stratum, however, a student with more than one major would appear only once. The total number of students in the population was 7,852. Accounting for students with multiple majors reduces this number to 6,676, which is the unduplicated population size.

Simple random sampling was performed within each of the population strata. At that stage of the sampling process, it was possible for a student to be included in more than one stratum, depending on whether he or she had more than one major. Therefore, students with more than one major underwent a secondary random selection process, in which only one of their majors was chosen for the final sample. This process ensured that the final sample included only one record for each student and that, for those students with multiple majors, the major to which they would refer when responding to survey questions was randomly chosen.

Reporting

Statistics in this report for individual majors/departments are unweighted. In contrast, statistics for the entire campus and colleges are weighted to adjust statistically for different sampling fractions and response rates over majors/departments.

Last revision 01/10/05