Dealing with Possible Baseline Inequalities Between Experimental Groups – The Case of Motor Learning. Issue 4 (3rd July 2020)

• Record Type: Journal Article
• Authors: Ziv, Gal; Lidor, Ronnie; Netz, Yael
• Is Part Of: Journal of motor behavior. Volume 52:Issue 4(2020)
• Page Start: 502
• Page End: 513
• Publication Date: 2020-07-03
• Subjects: Motor learning -- motor performance; Human mechanics -- Periodicals; Movement, Psychology of -- Periodicals; Motor Activity -- Periodicals; Motor Skills -- Periodicals; Bewegingsleer; Mécanique humaine -- Périodiques; Psychomotricité -- Périodiques; 612.705
• DOI: 10.1080/00222895.2019.1649996
• Languages: English
• ISSNs: 0022-2895

Abstract

Abstract: One important concept of experimental design is the random assignment of participants to experimental groups. This randomization process is used to prevent selection bias, as well as to provide a strong basis for a cause-and-effect relationship between the independent variable/s and the dependent variable/s. In small sample sizes, simple randomization may not provide equal groups at baseline for one or more of the variables, and therefore more restricted types of randomization, such as the stratified permuted-block randomization, can be used. A code was written to calculate the probability that simple randomization will not lead to equality between groups at baseline, and then an example of stratified permuted-block randomization was examined. The findings suggest that for certain variables that are commonly measured in experiments in motor learning, there is a relatively high probability that groups will not be equal at baseline after simple randomization. This observation reflects the small sample sizes usually found in the literature on motor learning. However, stratified permuted-block randomization does lead to greater equality among groups. Implications for researchers are discussed, and a flowchart is proposed that will allow researchers to decide whether to use simple or stratified randomization.