Cross-Sectional Data Uses
Kyle Fassett– At the 2019 Assessment Institute in Indianapolis, research analyst Allison BrckaLorenz and I presented on the challenges and benefits of cross-sectional data. We discussed with the audience several examples of research from the National Survey of Student Engagement (NSSE), the Beginning College Survey of Student Engagement (BCSSE), and the Faculty Survey of Student Engagement (FSSE).
While longitudinal data is advantageous for observing trends for a specific student (or group of students) over time, it can be difficult match or track datapoints. Often, this research is not conducted with engagement data as it varies greatly across student years; the assumption being, students will not necessarily “grow” in engagement but rather their experiences will change. For example, student faculty interactions in the first year will be frequent given advising support, and in the senior year as well as students partake in research experiences. That being said, BCSSE’s data regarding student expectations of engagement can be paired with their actual experiences captured by NSSE. As such, one may examine the same student to see how their experiences evolve.
Meanwhile, cross-sectional provides us a snapshot of students’ experiences at a certain moment in time. As it is the most common methods used with our data, in the session, we outlined four different examples through the lens of several different populations.
- First year students – Multi-Year Report
- Looking at a snapshot of the first-year experience over time can help us see trends as well as the implications of interventions.
- Senior students’ competencies – Nested Data
- Using tactics such as multilevel modelling allows us to examine variation at both the student and department/institutional level.
- Teaching practices over time – Descriptive
- Looking at pieces of data over years can help us look easily at patterns and demonstrate findings to a variety of constituents.
- Scholarship of Teaching & Learning –Content Summary
- Aggregating data creates an opportunity for drilling into a number of populations that may be underexplored