How Cognitive Biases Influence Problematic Research Methods Practices

Mon, 19 Feb 2024 00:00:00 +0000

A growing body of academic research addresses issues related to questionable choices and errors in the use of research methods in published business research. These problematic research method practices (PRMPs) may be purposeful or unconscious, but they reduce the rigor of academic research and can harm the accumulation of scientific knowledge. Yet, absent from much of this literature is a theoretically grounded approach to understanding why these problematic practices occur. Prior scholars have summarized specific types of PRMPs, but attributions about their causes are primarily limited to research lack of motivation or poor doctoral education. While these may certainly be at play, the current manuscript proposes that the deeper psychological phenomenon of cognitive bias is a likely explanation. Cognitive biases occur when human cognition produces an outcome that is systematically distorted from objective reality (Haselton, Nettle, and Murray, 2016). More colloquially, cognitive biases are systematic errors that humans make when they are faced with perceiving, remembering, and understanding information. These unintentional biases are particularly likely when that information is voluminous and ambiguous. Cognitive biases are explained by two theories—heuristic theory and fuzzy trace theory. Heuristic theory suggests that humans default to using mental shortcuts as a means to make decisions more efficiently (Chaiken and Ledgerwood, 2012). Further, fuzzy trace theory explains how memory and reasoning can be flawed (Reyna and Brainerd, 1995). Because of the limitations of the human mind, heuristic theory and fuzzy trace theory act to create unintentional cognitive biases. The current manuscript argues that the cognitive biases of source confusion, gist memory, repetition effects, bandwagon effects, and confirmation bias are mostly subconscious means by which researchers make errors in research methods use. We argue that these biases are not a useful part of the didactic approach to research, but are rather mental shortcuts that can limit researcher effectiveness. Next, specific PRMPs are addressed: reliance on methodological myths and urban legends, errors in citations, use of questionable research practices, and inappropriate use of artificial intelligence (AI) tools and technology in research. Finally, there are a number of insights and recommendations derived from research on cognitive biases to assist scholars in promoting research methods best practices. In particular, researchers can combat cognitive biases by recognizing what they are and by providing more transparency about research methods use in their articles. Incentives for authors and reviewers may reduce the impact of cognitive biases on PRMPs. Editors should create and share clear guidelines on the use of AI in research. In summary, this manuscript addresses those critical issues, fills a gap in current research regarding why PRMPs occur, and provides researchers with key insights to effectively combat cognitive biases.

Statistically Validating a Theory Represented by a Venn Diagram

Crystal Evans, Greg Evans, Lorin Mayo, Tammy Corcoran — Mon, 18 Mar 2024 00:00:00 +0000

To date, there has been no proposed method to statistically validate Venn diagrams. We seek to correct this shortcoming. This paper is a review of a proposed method that offers the possibility of statistically validating Venn diagrams through the lens of the management vs. leadership debate in business. Through this research, we demonstrate a way to statistically validate Venn diagrams by using a modified method of exploratory factor analysis (EFA). First, when performing EFA to validate a Venn, we suggest the scree plot of eigenvalues will indicate how many circles should be in the diagram. Additionally, when normally conducting EFA, cross-loaded items are removed. However, when using EFA to validate a Venn, we propose items that cross load should be retained and placed in the corresponding intersection of the two (or more) circles of the diagram. Applying this method to a sample of 431 (n=431) employees aged 25 years or older, we created a statistically validated Venn diagram that identifies those skills that are uniquely management, uniquely leadership, and the overlap as reported by employees. As a result, this research provides scholars with the opportunity to classify actions as leadership or management based on their placement within the statistically validated Venn diagram of management skills and leadership skills. Importantly, through the application of this new research method, we bring the possibility of statistical confirmation to many of our social science theories that are represented by Venn diagrams. In the Discussion section, we offer a critique of possible limitations of the method and mistakes that researchers can make when applying this method.

Electronic Journal of Business Research Methods

How Cognitive Biases Influence Problematic Research Methods Practices

Statistically Validating a Theory Represented by a Venn Diagram