Paul Meehl as methodologist-philosopher of science: the formulation of metascience

Paul Meehl's contributions to methodology and the philosophy of science extend well beyond his widely known writings in such areas as construct validity and statistical significance testing. I describe one of Meehl's less well-known, but potentially most important, methodological undertakings: his work on metascience, or the science of science. Metascience could ultimately revolutionize our conceptualizations and understanding of science and provide considerable help to practicing scientists and scientific endeavors, including efforts to advance the development and appraisal of theories in psychology.     

Meehl and the ex post facto design

Paul E. Meehl's problem of systematic unmatching is discussed, and it is argued that his reasoning and conclusion are incorrect. Unmatching will probably occur seldom in empirical cases, and if it does take place, it will constitute some sort of undercorrection. Moreover, his attempt at interpretation of statistical control in terms of so-called "counterfactuals" is found to be misleading.     

Meehl's contribution to clinical versus statistical prediction

Paul E. Meehl's work on the clinical versus statistical prediction controversy is reviewed. His contributions included the following: putting the controversy center stage in applied psychology; clarifying concepts underpinning the debate (especially his crucial distinction between ways of gathering data and ways of combining them) as well as establishing that the controversy was real and not concocted, analyzing clinical inference from both theoretical and probabilistic points of view, and reviewing studies that compared the accuracy of these 2 methods of data combination. Meehl's (1954/1996) conclusion that statistical prediction consistently outperforms clinical judgment has stood up extremely well for half a century. His conceptual analyses have not been significantly improved since he published them in the 1950s and 1960s. His work in this area contains several citation classics, which are part of the working knowledge of all competent applied psychologists today.     

Schizotaxia, schizotypy, and schizophrenia: Paul E. Meehl's blueprint for the experimental psychopathology and genetics of schizophrenia

Paul E. Meehl proposed a model of the cause and pathogenesis of schizophrenia and related states in the early 1960s (Meehl, 1962), which he later revised in 1990 (Meehl, 1990). His model emphasized a genetically influenced aberration in neural transmission that could eventuate in clinical schizophrenia, nonpsychotic schizotypic states, or apparent normalcy depending on the coexistence of other factors. His model embodied the core ideas of the diathesis-stressor framework that would come to dominate experimental and developmental psychopathology for the next 40 years. The author reviews Meehl's model of schizotaxia, schizotypy, and schizophrenia and reviews and clarifies some frequent misunderstandings of the model.     

Categories and continua: a review of taxometric research

The taxometric procedures developed by Paul Meehl and his colleagues have been used in a large body of research seeking to distinguish between categorical and continuous models of latent variables. In this article the authors survey taxometric studies of psychopathological, personality, and other variables and draw conclusions about the taxonicity of latent variables in these domains. In addition, the authors review research on the validation and refinement of taxometric methods and make proposals for improving the application of taxometric research. The authors consider questions that remain about the conceptual status of taxometrics and raise new ones. They show that taxometric methodology has made an accelerating number of contributions to psychological research, has resolved several longstanding controversies, and has challenged some entrenched theoretical assumptions in differential psychology. Moreover, they contend that the research possibilities that it affords have yet to be fully exploited.     

Confirming the latent structure and base rate of schizotypy: a taxometric analysis.

Meehl's (1962, 1990) model of schizotypy and the development of schizophrenia implies that the structure of liability for schizophrenia is dichotomous and that a "schizogene" determines membership in a latent class, or taxon (Meehl & Golden, 1982). The authors sought to determine the latent structure and base rate of schizotypy. They applied Meehl's (1973; Meehl & Golden, 1982) MAXCOV-HITMAX taxometric analytic procedures to a subset of items from the Perceptual Aberration Scale (PAS; Chapman, Chapman, & Raulin, 1978), a prominent psychometric index of schizotypy, derived from a randomly ascertained nonclinical university sample (N = 1,093). The results, in accordance with Meehl's conjectures, strongly suggest that schizotypy, as assessed by the PAS, is taxonic at the latent level with a general population taxon base rate of approximately .10.     

Taxometrics in psychopathology research: an introduction to some of the procedures and related methodological issues

This article introduces a special section on the use of taxometrics to examine the categorical versus the dimensional structure of various forms of psychopathology. Paving the way into the special section, this introduction briefly describes 3 taxometric methods--mean above minus below a sliding cut (MAMBAC), maximum covariation (MAXCOV), and maximum eigenvalue (MAXEIG)--and discusses possible threats to statistical conclusion validity that often emerge when such techniques are applied in psychopathology research.     

What's in a taxon?

Whether taxometrics yields inferential knowledge to something latent is partly but not wholly a semantic question. Although the single variables are manifest indicator scores of individuals, the statistics computed from them via postulates of the formalism are not mere data summaries and will be incorrect or meaningless if the structural conjectures are false. The unidirectional derivability from postulates to data relations supplies the taxonic inferences' surplus meaning that constitutes conventional psychometric meaning of latency (e.g., latent class analysis). Surplus meaning beyond the purely mathematical may be provided by interpretive text attributing unobserved attributes or causal origin to taxon members.     

Pseudotaxonicity in MAMBAC and MAXCOV analyses of rating-scale data: turning continua into classes by manipulating observer's expectations

Taxometric procedures such as mean above minus below a cut and maximum covariance can determine whether a trait is distributed as a discrete latent class. These methods have been used to infer taxonic structure in several personality and psychopathology constructs, often from analyses of rating scale data. This is problematic given (a) well established biases in ratings, (b) the human tendency to think categorically, and (c) implicit typological models of personality and psychopathology among expert raters. Using an experimental method in which the cognitive sets of raters were manipulated as dimensional versus categorical, it is demonstrated that pseudotaxonicity can be created readily with rating scale measures. This suggests that researchers avoid an exclusive reliance on rating scales when conducting taxometrics investigations.     

Detecting latent taxa: Monte Carlo comparison of taxometric, mixture model, and clustering procedures

A Monte Carlo evaluation of four procedures for detecting taxonicity was conducted using artificial data sets that were either taxonic or nontaxonic. The data sets were analyzed using two of Meehl's taxometric procedures, MAXCOV and MAMBAC, Ward's method for cluster analysis in concert with the cubic clustering criterion and a latent variable mixture modeling technique. Performance of the taxometric procedures and latent variable mixture modeling were clearly superior to that of cluster analysis in detecting taxonicity. Applied researchers are urged to select from the better procedures and to perform consistency tests.     

The taxon concept: is it taxonic?

The question of whether the concept of a "taxon" (a nonarbitrary latent category) is itself categorical, or is a matter of degree, has lain dormant within taxometrics. I analyze the problem conceptually. Part of the meaning of "taxon," I hold, goes beyond the manifest statistical properties of admixed probability distributions; any of certain forms of causal nexus leading to admixed distributions are involved as well. I defend the thesis that the threshold question, "Is there a taxon?" has a yes-or-no answer, establishing that "taxonicity" is categorical (taxonic), but also that as some taxa are much more readily distinguished from their complement classes than are others, the taxon question is also quantitative. It is on this basis that Meehl and others held the view that taxonicity is a matter of degree; however, they were wrong to hold that it is only a matter of degree.     

Whither goest our redoubtable mentor, the MMPI/MMPI-2?

Four areas of MMPI use and development toward the year 2000 and beyond are discussed. First, although the MMPI-2 booklet is a clear improvement, we will continue to need to use both profiles, the MMPI-2 for normative purposes and the MMPI profile for pattern interpretation. Applying MMPI expectations to MMPI-2 profiles is a violation of Meehl's basic actuarial prediction concepts. Secondly, as psychologists begin prescribing medications, we may be able to substantially refine drug choices. Thirdly, the measurement of socioeconomic status levels appears to be the major, missing moderator variable in MMPI/MMPI-2 interpretation. Lastly, an effective and positive MMPI/MMPI-2 feedback paradigm is discussed that fits well within the managed care context.     

Eating disorder diagnoses: empirical approaches to classification

Decisions about the classification of eating disorders have significant scientific and clinical implications. The eating disorder diagnoses in the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) reflect the collective wisdom of experts in the field but are frequently not supported in empirical studies and do not capture the disorders of eating experienced by most people with an eating disorder. Statistical approaches to classification such as latent class analysis and taxometrics can help to create a classification system with greater scientific validity and clinical utility. The field would benefit from direct empirical comparisons of different classification schemes with various clinical and scientific validators. Such studies would enable the creators of the next DSM eating disorder classification to increase understanding of the advantages and disadvantages associated with choosing various diagnostic criteria sets for the eating disorders.     

Categorical versus dimensional status of borderline personality disorder

One of the many controversies concerning the borderline personality disorder (BPD) diagnosis is whether the construct refers to a categorical or dimensional variable. The current study used Meehl's (1973) maximum covariance analysis to investigate this issue. The charts of 409 psychiatric inpatients were systematically reviewed for the presence of BPD and dysthymic symptoms. Charts of 244 inpatients were also reviewed to assess the presence of indicators of male sex, a categorical variable. The results for BPD and dysthymia were consistent with a dimensional model, whereas those for male sex were consistent with a categorical model. A dimensional model of classification of BPD is recommended, and suggestions for future research are provided.     

The maxslope taxometric procedure: mathematical derivation, parameter estimation, consistency tests

This article first explains concepts in taxometrics, including the meaning of "taxon" in relation to taxometric procedures. It then mathematically develops the MAXSLOPE procedure of Grove and Meehl which relies on nonlinear regression of one taxometric indicator variable on another. Sufficient conditions for MAXSLOPE's validity are set forth. The relationship between the point of maximum regression slope (MAXSLOPE point) and the HITMAX cut, i.e., the point on a variable which, if used as a diagnostic cut-off score, yields maximum classification accuracy, is analyzed. A sufficient condition is given for the MAXSLOPE point to equal the HITMAX cut; however, most distributions have different MAXSLOPE and HITMAX points. Equations and an algorithm are spelled out for making a graphical test for the existence of a taxon, estimating taxometric parameters, and conducting consistency tests; the latter serve as stringent checks on the validity of a taxonic conjecture. The plausibility of assumptions made, in deriving MAXSLOPE equations, is discussed, and the qualitative effects of violations of these assumptions are explained.     

Are infant attachment patterns continuously or categorically distributed? A taxometric analysis of strange situation behavior

Contemporary attachment research is based on the assumption that at least three types of infant attachment patterns exist: secure, avoidant, and resistant. It is not known, however, whether individual differences in attachment organization are more consistent with a continuous or a categorical model. The authors addressed this issue by applying P. E. Meehl's (1973, 1992) taxometric techniques for distinguishing latent types (i.e., classes, natural kinds) from latent continua (i.e., dimensions) to Strange Situation data on 1,139 fifteen-month-old children from the NICHD Study of Early Child Care. The results indicate that variation in attachment patterns is largely continuous, not categorical. The discussion focuses on the implications of dimensional models of individual differences for attachment theory and research.     

Psychosis proneness and clinical psychopathology: examination of the correlates of schizotypy

The present report examined the associations between the Perceptual Aberration Scale (PAS), a prominent psychometric index of hypothetical psychosis proneness, and several measures of clinical psychopathology in a nonpsychotic psychiatric sample (N = 101). Patients were examined by experienced clinicians using structured psychiatric interviews to assess DSM-III-R Axis I and II conditions and rated for anxiety, depression, severity of illness, and current adult social competence. Elevated scores on the PAS were most closely associated with anxiety and depression as well as schizotypal, schizoid, avoidant, and obsessive-compulsive personality disorder symptomatology. Hierarchical regression analysis identified schizotypal symptoms and anxiety as the two underlying psychopathological processes most useful in explaining variance in PAS scores. Results are interpreted as supporting both the clinical relevance and research utility of the PAS and enhancing the construct validity of Meehl's model of schizotypy.     

Deconstructing, reconstructing, preserving Paul E. Meehl's legacy of construct validity

The question of the status of cause-and-effect explanations of human behavior that posit physically existing causative factors and those that, on the other hand, posit hypothetical entities in the form of "useful fictions" has a long history. The influence of the works of Jeremy Bentham and Hans Vaihinger, as well as the later influence of Francis Galton, is described. Issues of the validity of hypothetical constructs and related problems of measurement and definition as found in psychoanalytic theory construction and in trait theory are examined. The significant and continuing interest generated by the landmark studies of K. MacCorquodale and P. E. Meehl (1948) and L. J. Cronbach and P. E. Meehl (1955) as well as the central importance of P. E. Meehl's thinking are described.     

Primary and secondary hypohedonia

Having shown taxometrically that there exists a hypohedonic schizotypal taxon in a college population, J. J. Blanchard, S. W. Gangestad, S. A. Brown, and W. P. Horan (2000) suggested that P. E. Meehl erred in revising his 1962 theory by postulating a normal-range individual differences variable of hedonic capacity that potentiates schizotypy into schizophrenia. The aversive drift and secondary anhedonia of Meehl's theory imply that the schizotypal taxon will generate hypohedonic taxonicity in an adult population. Psychometrically measurable hedonic disposition (as distinguished from genetic primary hedonic capacity) is "dragged along" by the schizogene, especially in the social domain. To choose between causal interpretations, it could be ascertained whether the schizotypal anhedonic taxon is composed of individuals who are schizotaxic on the basis of psychophysiological, cognitive, and soft neurologic indicators.     

Paul E. Meehl's contributions to personality assessment

Many of Meehl's interests converged in his work on personality assessment. In empirical research, development of the K scale of the Minnesota Multiphasic Personality Inventory and various scales of other tests are noteworthy, but his most profound contributions were conceptual and philosophical. Essays on the role of base rates in clinical decision and on construct validity in psychological tests provided vital insights on core issues in assessment. The concept of nomological nets allowed systematic investigation of unobservable theoretical entities, with liberating effects on all natural sciences. An active clinician, Meehl understood deeply the complexities of professional service, but his insistence on empirical validation as the ultimate arbiter of the value of any procedure required all attainable rigor in the scientific foundations of clinical practice.