HLM Behind the Curtain: Unveiling Decisions Behind the Use and Interpretation of HLM in Higher Education Research

Abstract

Hierarchical linear modeling (HLM) has become increasingly popular in the higher education literature, but there is significant variability in the current approaches to the conducting and reporting of HLM. The field currently lacks a general consensus around important issues such as the number of levels of analysis that are important to include and how much variance should be accounted for at each level in order for the HLM analysis to have practical significance (Dedrick et al., Rev Educ Res 79:69-102, 2009). The purpose of this research is to explore the use of a 3-level HLM model, appropriate contextualizing of results of HLM, and the interpretation of HLM results that resonates with practice. We used an example of a 3-level model from the National Study of Living Learning Programs to highlight the practical issues that arise in the interpretation of HLM within a higher education context.

Is work keeping us from acting healthy? How workplace barriers and facilitators impact nutrition and exercise behaviors

“…Hypotheses 2a and 2b were also fully supported at this level as the fixed effects for nutrition and exercise facilitators had significant effects on their respective behaviors. The t‐scores and mean differences for the analyses suggest a moderate to strong effect of barriers/facilitators to healthy behaviors, but due to the uncertain nature of the proper way to display effect sizes for MCRM (Niehaus, Campbell, & Kurotsuchi Inkelas, ), none are reported here. It should also be noted that there were very few days when three barriers or facilitators were reported (especially for exercise), so these means may not be as meaningful as those for the other day totals.…”

Optimizing Technical Education Pathways: Does Dual-Credit Course Completion Predict Students’ College and Labor Market Success?

“…HLM is an extension of linear regression (Raudenbush & Bryk, 2002;Snijders & Bosker, 2012). However, different from linear regression that requires independence among observations, HLM is an appropriate technique when the observations are nested (Niehaus, Campbell, & Inkelas, 2014). In the case of the current study, students are nested within high schools where they shared the learning environment before attending FVTC.…”

“…HLM is deemed necessary, therefore, to account for this dependency of learning environment and to further examine how high school characteristics are related to students' postsecondary learning outcomes, in addition to the individual-level characteristics. First, we examined whether the variance in the outcome variables is attributable to the school-level variables vis-à-vis individual-level variables by calculating the intraclass correlation (ICC; Niehaus et al, 2014) for the models with a continuous outcome variable. As a result, the low level of ICC indicates that the variance of the outcome variables was not strongly associated with these school-level variables (see Table 6).…”

Gender Differences in Learning Outcomes from the College Experiences of Engineering Students

“…By using the random variance component from an unconditional model (Niehaus, Campbell, & Inkelas, 2013;Raudenbush & Bryk, 2002), we found that the seven learning outcomes varied significantly across different engineering programs. The intraclass correlation indicated that 2% to 11% of the variability across the outcomes exists between programs (Table 4).…”