Discrepant Results in Mental Health Research : What They Mean, Why They Matter, and How They Inform Scientific Practices

Discrepant results are the most common results of mental health research, and may even occur multiple times in a single study. But what do discrepant results mean? How are researchers to make sense of them? Do they tell us something important about mental health, and if they do, how would we know? Discrepant Results in Mental Health Research challenges all that we think we know about mental health.

Discrepant Results in Mental Health Research probes the most common outcomes of mental health studies. Discrepant results appear in scores of studies traversing the globe--a literature dating back to the 1950s. This literature reveals that any two studies often differ in their estimates of anything from the prevalence of mental health conditions to the effects of mental health treatments. In fact, researchers often encounter discrepant results among findings made in a single study. Discrepant results factor into what we think we know about how often mental health conditions occur, what causes them, and how to treat them. Yet, researchers do not know what to do with discrepant results when they encounter them. The problem is not with their methods--discrepant results appear even when researchers use high-quality instruments to collect data. The problem lies with how researchers interpret their data, and the decisions they make with those data. To address this problem, the book reveals a conceptually grounded, evidence-based approach to discrepant results in mental health research. It describes the robust nature of these discrepant results, along with theoretical models for understanding and interpreting them. These models inform sound scientific practices. The book reviews work that has implemented these practices, and it also leverages illustrative case examples to facilitate content mastery. Additionally, it describes future directions in research on discrepant results across several areas of work, including measurement development, intervention science, data analysis, and clinical populations that have received relatively little attention on issues surrounding discrepant results, such as suicide risk and autism.