Dissertation
ALLEGIANCE ASSOCIATION OR BIAS? A QUASI-EXPERIMENTAL TEST OF THE ALLEGIANCE BIAS HYPOTHESIS
Washington State University
Doctor of Philosophy (PhD), Washington State University
01/2021
DOI:
https://doi.org/10.7273/000002471
Handle:
https://hdl.handle.net/2376/120473
Abstract
Researcher allegiance is associated with treatment outcomes in psychotherapy clinical trials. What causes these positive associations is unclear and debated. The true efficacy hypothesis asserts they are a byproduct of fundamental treatment differences. By contrast, the allegiance bias hypothesis asserts that allegiance affects the outcomes of comparative trials, with results favoring researcher-preferred treatments. This quasi-experimental study sought to test whether allegiance affects results, by assembling pairs of research reports on randomized trials of the same psychotherapies for the same disorder conducted by researchers with opposing allegiances. The pairwise results were then assessed for distinctive patterns that would suggest the operation of allegiance bias. In addition, this study aimed to overcome the standard reprint method’s limitations by using a novel web-search method to rate allegiance. A total of 19 final pairs were identified (k = 38). Seven pairs were rated “strong” matches, and 12 pairs were considered “weak” matches. The main outcome for each study was the original author conclusions about the relative efficacy of the treatments. The two additional outcomes involved the magnitude and significance of posttreatment effect sizes. For 10 of the 19 pairs, the outcome patterns suggested the operation of allegiance bias on the main outcome (53%). Allegiance bias appeared evident for five of the seven pairs considered “strong” matches (71%). Thus, there was broad evidence consistent with the view that researcher allegiance affects the outcomes of some comparative trials. However, a minority of pairwise outcomes appeared best explained by true efficacy differences, providing some mixed support for the allegiance bias and true efficacy hypotheses. Closer examination of the mixed findings suggests that allegiance bias may be more prevalent in specific areas of clinical research. Implications for psychotherapy research and future directions are discussed.
Metrics
97 File views/ downloads
68 Record Views
Details
- Title
- ALLEGIANCE ASSOCIATION OR BIAS? A QUASI-EXPERIMENTAL TEST OF THE ALLEGIANCE BIAS HYPOTHESIS
- Creators
- Jacob A Zimmerman
- Contributors
- David K. Marcus (Advisor)David K Marcus (Committee Member)Walter D. Scott (Committee Member)Benjamin O. Ladd (Committee Member)
- Awarding Institution
- Washington State University
- Academic Unit
- Psychology, Department of
- Theses and Dissertations
- Doctor of Philosophy (PhD), Washington State University
- Publisher
- Washington State University
- Number of pages
- 113
- Identifiers
- 99900606955601842
- Language
- English
- Resource Type
- Dissertation