Quantifying the Biasing Effect of Rapid Guessing on Estimates of Coefficient Alpha
2021
Loading...
View/Download File
Persistent link to this item
Statistics
View StatisticsJournal Title
Journal ISSN
Volume Title
Title
Quantifying the Biasing Effect of Rapid Guessing on Estimates of Coefficient Alpha
Authors
Published Date
2021
Publisher
Type
Presentation
Abstract
An underlying assumption of coefficient alpha is that random error is uncorrelated; however, this assumption is violated when examinees engage in similar construct-irrelevant behaviors across items. One construct-irrelevant behavior that has gained increased attention in the literature is rapid guessing (RG), which occurs when examinees answer quickly with intentional disregard for item content. To examine the extent that estimates of coefficient alpha are biased due to RG, a simulation study was conducted in which the ability characteristics of rapid responders and the percentage and pattern of RG were manipulated. After controlling for test length and difficulty, results indicated that RG characteristics had a practically negligible impact on estimates of coefficient alpha, with the average degree of bias found to range from -.05 to .02 for upwards of 30% of RG responses in the data. This negligible effect was supported in a meta-analytic investigation, which observed a difference in coefficient alpha of .07 when comparing filtered (i.e., RG responses removed) and unfiltered (i.e., RG responses left in the data) datasets across five studies and 12 effect sizes. These findings suggest that estimates of coefficient alpha are largely robust to violations of the assumption that random error is uncorrelated due to construct-irrelevant behaviors such as RG.
Description
Related to
Replaces
License
Collections
Series/Report Number
Funding information
Isbn identifier
Doi identifier
Previously Published Citation
Suggested citation
Deng, Jiayi; Rios, Joseph. (2021). Quantifying the Biasing Effect of Rapid Guessing on Estimates of Coefficient Alpha. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/219265.
Content distributed via the University Digital Conservancy may be subject to additional license and use restrictions applied by the depositor. By using these files, users agree to the Terms of Use. Materials in the UDC may contain content that is disturbing and/or harmful. For more information, please see our statement on harmful content in digital repositories.