Session: Publication Bias Assessment Methods and Interpretation in Meta-Analysis: A Friendly Introduction (Society for Social Work and Research 26th Annual Conference - Social Work Science for Racial, Social, and Political Justice)

313 Publication Bias Assessment Methods and Interpretation in Meta-Analysis: A Friendly Introduction

Sunday, January 16, 2022: 9:45 AM-11:15 AM
Congress, ML 4 (Marriott Marquis Washington, DC)
Cluster: Research Design and Measurement
Anao Zhang, Ph.D., University of Michigan-Ann Arbor, Joyce Lee, MSW, MS, The Ohio State University and Sunghyun Hong, MSW, University of Michigan-Ann Arbor
Systematic reviews and meta-analysis have become important scientific methods to inform evidence-based social work practice. An essential yet often overlooked component of research synthesis is the assessment of publication bias. Publication bias refers to the phenomenon that systematic bias occurs in published academic research (i.e., only studies with statistically significant findings are being published). If not effectively detected, literature of a scientific field would consist of substantial biases resulting from inflated research evidence, and in worst case scenarios, leading to false conclusions. Over the years, numerous methods have been developed in evaluating publication bias in systematic review and meta-analysis studies. These, for example, include the funnel plot, Egger's regression, and trim-and-fill methods, among others. Yet, conclusions drawn from these publication methods are not always consistent with each other, and interpretations of these methods are often different. Therefore, it is important for social work research synthesists to develop a better understanding of publication bias methods, as well as their corresponding interpretations, so that rigorous conclusions can be drawn given a body of scientific evidence. Workshop presenters will start by introducing a set of common methods evaluating publication bias, including: 1) Funnel plot; 2) Kendalla's Tau Test (i.e., Begg and Mazumdara's rank correlation test); 3) Egger's regression; 4) Fail-Safe Numbers; 5) (Iterative) Trim-and -Fill; and 6) Vevea and Woods sensitivity weighting function. A conceptual and technical review of all these methods will be presented, and each method's latest development will be discussed. Comparisons across different methods will be presented next, and the presenters will conclude by providing a list of recommended practice guidelines relevant to publication bias methodologies in social work research.

Career Level and Prerequisites: Fundamental knowledge publication is preferred but not required.

Pedagogical Approach: The pedagogical approach will include a PowerPoint presentation, questions and answers, hands-on demonstration, and small-group activities. All data, codes, and materials will be shared with attendees.

See more of: Workshops