Video file

Chapter 4 : Looking Deeper - Are quality of interactions important in the evaluation of college life among first year’s?



The study analyses data from the 2023 iteration of This is an online survey offered to students of participating higher education institutions. There is a three-week completion window available for students at each institution, during the second semester. Participation is self-selecting and voluntary.

The analysis in this chapter includes only observations from first year undergraduate students.

Analytic sample, missing data, population, and weighting

Defining analytic sample

We employed the listwise deletion method for missing data in this analysis, so that only observations for which all item responses were completed were included in the study.

Figure 4.1 A illustrates the process by which the final analytic sample was defined. 8,415 observations were included in the study.

Development of analytic sample for the study
Figure 4.1: Development of analytic sample for the study

Sample weighting and population data

Following the definition of the analytical sample, sample data may be compared to population data for key variables to understand if any systematic differences appear between the sample and the population frame, in this case first year undergraduates in Ireland.

It is also noted that the is a voluntary survey and not designed as a representative stratified sample of the population frame, as such differences may arise between the total number of respondents and the population frame. However, analysis of response rates by multiple identified sub-groups of the population indicates that the sample may be regarded as broadly representative. A weighting variable is often employed in such cases to compensate for such differences, and a weighting variable has been provided by the data collection firm employed by

Figure 4.1 compares the total sample of first year undergraduate students, the defined analytical sample used in the study, the weighting of the analytic sample and the target population for the key variables of gender, age group, institute type and field of study.

thinking emoji

For gender, the application of the weighting variable corrects biases of an oversampling of females in both the raw data and the analytic sample and tracks relatively closely to the 2022 population data.

For age group, the comparative categories were not available at population level. However, it appears that the sample underestimates the number of older-aged students and that the application of the weighting data does not respond to fully correct this.

Some differences are apparent between sample and population for field of study, in particular an underestimate of students studying business and health, and an overestimate of those studying Arts. However, for other fields, the data is closely aligned after the application of the weighting variable.

All descriptive and inferential results are reported based on weighted data.

*Extracted from population data in 2022

**Age groups at population level reported in two categories only