Assessing Longitudinal Change: Adjustment for Regression to the Mean Effects

Abstract

Pascarella (J Coll Stud Dev 47:508-520, 2006) has called for an increase in use of longitudinal data with pretest-posttest design when studying effects on college students. However, such designs that use multiple measures to document change are vulnerable to an important threat to internal validity, regression to the mean. Herein, we discuss a brief history of regression to the mean and illustrate a straightforward procedure to make adjustments to initial pretest scores for regression to the mean effects utilizing a method developed by Roberts (in: G. Echternacht (Guest ed.) New directions for testing and measurement, 1980). Analyses are shown with both unadjusted and adjusted pretest scores, illustrating dramatic differences in conclusions about whether students change across time.Keywords Regression to the mean Á Pre-test/post-test design Á Methodology Á Longitudinal data Á College students Á Change Higher education researchers have been aided in their study of college and university faculty and students by longitudinal data, and the models that have guided most of the recent research on the effects of college on students are longitudinal in nature (Smart et al. 2000). Moreover, longitudinal studies have provided college and university administrators with a plethora of support in making data based decisions related to policies and practices applied to individual institutions. Pascarella (2006) has even called for an increase in use of longitudinal data with pretest-posttest design when studying effects on college students. While longitudinal data can be beneficial to scholars and administrators alike, it can also be very difficult to conduct, very time-consuming, and costly.

Sentence Intelligibility Before and After Voice Treatment in Speakers With Idiopathic Parkinson’s Disease

“…34 The pretreatment scores were then adjusted for RTM, following the method of Rocconi and Ethington. 35 A twoway ANOVA with repeated measures (treatment 3 speaker) was calculated using the mean pre-and post-treatment scores adjusted for RTM. Both the main effect of treatment and the treatment 3 speaker interaction remained statistically significant.…”

Perceptual Structure of Adductor Spasmodic Dysphonia and Its Acoustic Correlates

“…Regression to the mean is the tendency for extreme scores to move closer to the mean when measured the second time because of imperfect correlation between the two tests. 37 To account for the unreliability and remove unwanted variance, pretreatment scores were adjusted by the procedure described in the study by Rocconi and Ethington before statistical analysis.…”

Effects of modafinil and methylphenidate on visual attention capacity: a TVA-based study

“…A significant correlation of r=−0.59 (p<0.01) was revealed for methylphenidate, while the correlation for modafinil was not significant (r=−0.40; p>0.10). Regression to the mean effects can only occur when change and initial status are significantly correlated (Barnett et al 2004;Rocconi and Ethington 2009); this means it may have affected the difference between the placebo and methylphenidate condition but does not explain baseline performance-dependent effects of modafinil. In order to reduce a possible bias induced by regression to the mean, we made an adjustment to the baseline score according to the procedure introduced by Roberts (1980).…”