Exaggeration Bias and Article Citations in Agricultural Economics

Research credibility in agricultural economics is compromised by two interrelated factors: selective reporting and low statistical power. These factors contribute to exaggerated findings that appear more persuasive and garner more citations. This study analyzes 849 articles published in leading U.S. agricultural economics journals between 2018 and 2023, with 48,962 observations. Two empirical analyses are conducted. The first regresses citation counts on p-values reported in article tables, a proxy for statistical power, article topics, and journal and year fixed effects. The second predicts the time it takes a journal to be cited ‘10’ times, given p-values and statistical power. We hypothesized that citation counts would be negatively associated with p-values (i.e., lower p-value attract more citations), while no specific hypothesis was formed for statistical power, as it is unobservable to readers. The results show that citation counts are strongly influenced by topic novelty and journal prestige, with studies reporting lower p-values receiving more citations, whereas adequately powered studies receive fewer. The misalignment between research rigor and citation counts raises concerns that farmers may adopt recommendations based on less reliable findings, as agricultural extension services may rely on citation metrics when evaluating scientific research. Thus, aligning citation-based evaluations with empirical credibility is important not only for maintaining trust in science but also for informing decisions made by farmers and extension agents.