A Primer on Intensive Longitudinal Design and Analysis for Classroom Research

• In Event: 30.030 - Intensive Data Approaches to Students' Day-to-Day, Moment-to-Moment Emotions, Motivation, and Engagement

Abstract

Purpose. The purpose of the proposed presentation is to highlight important design and analysis features of intensive longitudinal studies and encourage applied researchers to consider using such designs when appropriate, while also informing them of the associated challenges. In intensive longitudinal studies, researchers collect a large set of repeated measures (e.g., daily) over a relatively short time span. A typical study involves collecting repeated measures on an outcome and one or more predictors to learn how individuals respond to influences or events occurring in their daily lives.

Data Source. We recently conducted such a study with 218 high school students in over 40 high school science classes to examine the influence of specific daily teacher practices as perceived by students during instruction on subsequent student motivation as reported after each class day for a 6-week period. The examples we use in this presentation will draw from our experiences in conducting this study and analyzing data.

Results and Scholarly Significance.Our presentation will focus on the following features of intensive longitudinal studies. First, we will provide a brief overview of the main features of intensive longitudinal studies. In particular, we will contrast intensive longitudinal studies with studies that do not collect repeated measures (between-subject designs) or collect a limited number of repeated measures (typical longitudinal and growth curve modeling studies). These latter types of studies are more widely used. Yet, intensive longitudinal studies provide important advantages.

Second, we will highlight the main advantages of intensive longitudinal studies. These include studying phenomena (e.g., use of teaching practices and student motivation) as they occur in context and unfold over time, resulting in greater generalizability of study conclusions. Also, the use of intensive repeated measures allows researchers to learn how participants respond on a given day as an event occurs. For example, on days when teachers use certain teaching practices, what happens to student motivation? This question is best addressed by the collection of daily measures.

Third, we will highlight the types of research questions that may be addressed in intensive longitudinal studies, which are typically focused on the main effect of one or more daily predictors on a daily outcome. A key message we wish to convey is that, unlike growth curve modeling, the effect of time (or change across time) may not be of interest or may not be present. We will present data which shows that an outcome measured daily may have no particular pattern of change over time and yet is strongly related to changes in a daily predictor.

Finally, challenges for such studies include selecting good brief measures, encouraging participants to provide data, use of technology in collecting data, data handling and management, and data analysis. Analyzing data involves use of sophisticated methods (multilevel modeling, multilevel EFA and CFA), as well as inclusion of temporal carryover. Further, analysis methods suitable when a smaller number of repeated measures are collected are generally not feasible. Recommendations for addressing these issues will be provided.

Authors

• Keenan A. Pituch, The University of Texas - Austin

• Erika Alisha Patall, University of Southern California