Paul Meehl is best known for his groundbreaking work on the etiology of schizophrenia, the clinical versus statistical prediction debate, and the development of taxometric methods. He proposed a genetic model for schizophrenia, introducing the concepts of schizotaxia and schizotypy, which became foundational to the diathesis-stress model. His 1954 book, 'Clinical Versus Statistical Prediction', demonstrated that statistical algorithms outperform clinical judgment in predicting behavioral outcomes. He also pioneered taxometrics to determine whether psychopathological conditions are categorical or dimensional. His work integrated philosophy of science, psychometrics, and clinical practice, making him a towering figure in 20th-century psychology.

Meehl's diathesis-stress model, first proposed in his 1962 APA presidential address, posits that schizophrenia arises from a genetic predisposition—a neural integrative defect he called 'schizotaxia'—that, when combined with environmental stressors, can lead to schizotypy and ultimately clinical schizophrenia. This model was revolutionary in emphasizing a specific, heritable biological vulnerability rather than a purely environmental or psychodynamic cause. Meehl argued that schizotaxia is a dominant gene effect that produces a pervasive neural deficit, which manifests behaviorally as schizotypy—a personality organization characterized by cognitive slippage, interpersonal aversiveness, and anhedonia. Only a subset of schizotypes go on to develop full-blown schizophrenia, depending on environmental factors. His theory spurred decades of genetic, neurobiological, and psychopathological research.

The clinical versus statistical prediction debate, ignited by Meehl's 1954 monograph, concerns whether human clinicians or statistical algorithms are better at predicting behavioral outcomes such as treatment response, violence risk, or academic success. Meehl reviewed the existing evidence and concluded that statistical formulas were consistently superior to or at least as accurate as clinical judgment, even when clinicians had access to more information. This finding challenged the intuitive confidence of clinicians and sparked ongoing research into decision-making biases, the limits of human cognition, and the conditions under which clinical expertise might add value. The debate remains central to evidence-based practice and has influenced the development of structured professional judgment tools in forensic and clinical settings.

Taxometrics is a family of statistical methods developed by Meehl and his colleagues to determine whether a latent variable is categorical (taxonic) or dimensional in nature. This is crucial for psychopathology because many psychiatric diagnoses assume categorical distinctions (e.g., having schizophrenia or not), while alternative models propose that symptoms exist on a continuum. Taxometric procedures—such as maximum covariance (MAXCOV), mean above minus below a cut (MAMBAC), and latent mode (L-Mode)—analyze the distribution of indicators to detect the presence of a taxon. Meehl hoped taxometrics would validate psychiatric nosology by identifying discrete disease entities. Although many disorders have been found to be dimensional, taxometrics remains a powerful tool for testing the structure of psychological constructs and has been applied to personality, mood, and anxiety disorders.

Paul Meehl was closely associated with the Minnesota Multiphasic Personality Inventory (MMPI), one of the most widely used psychological assessment instruments. As a graduate student at the University of Minnesota, he assisted Starke Hathaway in refining the MMPI and later co-authored the 'Atlas for the Clinical Interpretation of the MMPI' with Hathaway. Meehl also developed the K scale, a validity scale designed to correct for defensive responding, which improved the test's accuracy. His theoretical and empirical work on the MMPI contributed to its standardization and clinical utility. Beyond specific scale development, Meehl's broader contributions to construct validity and psychological assessment informed how the MMPI and other tests are interpreted in both research and practice.

Construct validity is the degree to which a test measures the theoretical construct it purports to measure, such as intelligence, depression, or anxiety. Meehl, along with Lee Cronbach, co-authored the seminal 1955 paper 'Construct Validity in Psychological Tests', which established the framework for validating psychological measures. They argued that construct validation is an ongoing, iterative process that involves integrating multiple lines of evidence—including convergent and discriminant validity, factor analysis, and experimental manipulation—to support the interpretation of test scores. Meehl emphasized that construct validity is not a property of the test itself but of the inferences drawn from test scores, a view that has become foundational to psychometric theory and has guided the development of virtually all modern psychological assessments.

Meehl was a vocal critic of the overreliance on null hypothesis significance testing (NHST) in psychology and other soft sciences. He argued that NHST is a weak and misleading method for theory verification because, with sufficiently large sample sizes, even trivial effects can achieve statistical significance. He contended that psychological theories should be evaluated by the precision of their predictions and the magnitude of effect sizes, not merely by p-values. Meehl advocated for stronger tests of theories, such as point predictions and cross-validation, and encouraged researchers to focus on replication and meta-analysis. His critiques contributed to the replication crisis and the subsequent movement toward open science, preregistration, and Bayesian methods in psychology.

Paul Meehl authored numerous seminal works. His most influential book is 'Clinical Versus Statistical Prediction: A Theoretical Analysis and Review of the Evidence' (1954), which sparked the clinical versus actuarial prediction debate. His 1962 paper 'Schizotaxia, Schizotypy, Schizophrenia' outlined his genetic model of schizophrenia. He co-authored the 'Atlas for the Clinical Interpretation of the MMPI' (1951) with Starke Hathaway. With Kenneth MacCorquodale, he wrote 'On a Distinction between Hypothetical Constructs and Intervening Variables' (1948), a classic in philosophy of science. His later works on taxometrics, including 'Factors and Taxa, Traits and Types, Differences of Degree and Differences in Kind' (1992), synthesized his statistical innovations. He also published extensively on metascience, philosophy, and the limitations of significance testing.