Download the full-sized PDF of Adjustment for the Regression to the Mean Effects in Studies with Repeated MeasuresDownload the full-sized PDF



Permanent link (DOI):


Export to: EndNote  |  Zotero  |  Mendeley


This file is in the following communities:

Graduate Studies and Research, Faculty of


This file is in the following collections:

Theses and Dissertations

Adjustment for the Regression to the Mean Effects in Studies with Repeated Measures Open Access


Other title
Type of item
Degree grantor
University of Alberta
Author or creator
Ibrahim, Quazi I
Supervisor and department
Maximova, Katerina (School of Public Health)
Senthilselvan, Ambikaipakan (Department of Public Health Sciences)
Examining committee member and department
Wild, Cameron (Department of Public Health Sciences)
Dinu, Irina (Department of Public Health Sciences)
Department of Public Health Sciences
Date accepted
Graduation date
Master of Science
Degree level
In repeated measures data, large or small values at the initial measurement tend to be followed by values that are closer to the mean at the follow-ups measurements. This tendency is called regression to the mean (RTM). The presence of the RTM effect is inevitable in repeated measures data because of less than perfect correlation (correlation coefficient < 1) between the repeated measurements. Despite the growing evidence of the presence of RTM effects in clinical and public health studies based on repeated measures data, very few studies have evaluated and considered them when interpreting observed changes over time. In intervention studies, an RTM effect is mixed with an intervention effect in observed changes. It is extremely important to separate the RTM effect from the observed change in order to isolate any intervention effect and thus to make valid inferences about the effect of the intervention. In studying changes in outcome variables in repeated measures studies, RTM effects should always be adjusted for the valid interpretation of the changes and unbiased assessment of the intervention effects. The choice of methods to control for the RTM effect should be based on the type (continuous, count) and shape (normal, non-normal distribution) of the outcome variables of interest. A new method of estimating RTM effects for non-normal data using simulation is proposed. The method is a combination of bootstrap sampling from the standardized outcome variable and matrix decomposition of the correlation matrix between the repeated measurements. The method is applied to adjust for the RTM effects in studying changes in mean drinks in a typical week in a study evaluating the impact of a brief alcohol intervention on youth. In the study, mean drinks followed a positively skewed distribution. The proposed method estimated the RTM effects considering the true distribution (positively skewed) of the outcome and in doing so, provided more accurate estimation of the intervention effects compared to other methods considered in the thesis. The method ensured valid interpretation of the observed changes in the outcome by providing the most accurate estimation of the RTM effect and then removing it from the data. The proposed method could be applied to adjust for the RTM effect in non-normal repeated measures studies.
Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.
Citation for previous publication

File Details

Date Uploaded
Date Modified
Audit Status
Audits have not yet been run on this file.
File format: pdf (PDF/A)
Mime type: application/pdf
File size: 744121
Last modified: 2016:06:24 17:40:24-06:00
Filename: Ibrahim_Quazi_I_201509_MSc.pdf
Original checksum: 4341fee85c54bf3ffe331a279156e96c
Activity of users you follow
User Activity Date