Elsevier

Journal of Criminal Justice

Volume 40, Issue 1, January–February 2012, Pages 10-20
Journal of Criminal Justice

Taxometrics and Criminal Justice: Assessing the Latent Structure of Crime-Related Constructs

https://doi.org/10.1016/j.jcrimjus.2011.11.003Get rights and content

Abstract

Purpose

The taxometric method is introduced as a way of investigating the latent structure of key criminal justice constructs.

Methods

Besides describing the theory, rationale, and procedures behind taxometrics, the taxometric method is applied to Wolfgang, Figlio, and Sellin’s (1972) delinquency cohort data to illustrate how the procedure works.

Results

After defining latent structure, its relevance to criminal justice is explored and the four core principles of taxometric analysis are described: coherent cut kinetics, consistency testing, comparison curve analysis, and evidence-based procedures. Sample and indicator preconditions for taxometric analysis are discussed and empirically-verified procedures for taxometric analysis are outlined: mean above minus below a cut (MAMBAC: Meehl & Yonce, 1996), maximum covariance (MAXCOV: Meehl & Yonce, 1994), maximum eigenvalue (MAXEIG: Waller & Meehl, 1998), maximum slope (MAXSLOPE: Grove & Meehl, 1993), and latent mode factor analysis (L-Mode: Waller & Meehl, 1998).

Conclusions

The taxometric method provides an avenue by which the latent structure of important criminal justice constructs can be explored and criminal justice research can be advanced.

Highlights

► Taxometric method capable of shedding light on key criminal justice concepts. ► Sample and indicator preconditions discussed. ► Principal procedures include MAMBAC, MAXCOV/MAXEIG, and L-Mode. ► Taxometrics illustrated with delinquency cohort data.

Introduction

It is has been repeatedly demonstrated that a majority of offenses are committed by a small minority of offenders. A general rule of thumb, based on years of research, is that more than half of all crime is committed by less than ten percent of the criminal population (Chaiken and Chaiken, 1982, Farrington and West, 1993, Shannon, 1982, Thornberry et al., 1995, Vaughn et al., 2011, Wolfgang et al., 1972). Variously referred to as life-course persistent delinquents (Moffitt, 1993), criminal psychopaths (Hare, 1993), and career criminals (Glueck & Glueck, 1930), these high-rate offenders account for over half the criminal offenses and close to three-quarters of the serious crimes committed in the United States each year (DeLisi, 2005). Habitual felon and sexual predator laws have been enacted and selective incapacitation policies have been advanced in the belief that targeting these high-rate offenders will significantly reduce the recidivism rate and help protect the public from this small homogeneous group of predatory individuals.

Another topic of concern to criminal justice scholars is the relationship between substance abuse and crime. Over two-thirds of the inmates in a national jail survey reported symptoms of alcohol/drug abuse or dependence (Karberg & James, 2005). Determining whether substance abuse/dependence is categorical or dimensional could help clarify this relationship. Cloninger has long maintained that a category of alcohol abuse, the male-limited pattern, is associated with high levels of criminality in both the father and child (Cloninger, Bohman, & Sigvardsson, 1981). This suggests that a category of behavior, Type II or male limited alcoholism, may explain, in part, the well-documented relationship between substance abuse and crime. These and other issues central to the criminology and criminal justice fields could be more thoroughly investigated and understood if researchers in these fields were more familiar with techniques designed to examine the latent structure of crime-related constructs like high-rate delinquency and Type II alcoholism.

Section snippets

What is latent structure and why is it important?

The manner in which a non-observable (psychological) construct distinguishes or differentiates between entities is known as its latent structure. The latent structure of a construct can be categorical, dimensional, or a mixture of both (Meehl, 1995). Categorical latent structure means that the construct divides entities into distinct groups or categories (e.g., Type I versus Type II alcoholism). Dimensional latent structure means that the construct orders entities along a continuum of scores or

Measuring latent structure

Traditionally, decisions about latent structure have been made on the basis of factor analytic or cluster analytic studies. Neither procedure, however, is particularly effective in differentiating between categorical and dimensional latent structure. To infer that a construct is dimensional based on the emergence of one or more factors constitutes an error in mathematical reasoning, says Meehl (1999), because statistical factors can be extracted from indicators of a categorical construct.

Taxometric principles

The taxometric method is based on four core principles: coherent cut kinetics, consistency testing, comparison curve analysis, and evidence-based procedures. Meehl and his colleagues employed the principle of coherent cut kinetics to create the taxometric method (Grove and Meehl, 1993, Meehl and Yonce, 1994, Meehl and Yonce, 1996, Waller and Meehl, 1998). Moving a cut score through a distribution of indicators (cut kinetics) for the purpose of determining whether predictable results can be

Preconditions for taxometric analysis

A taxometric analysis is only as valid as the sample and indicators upon which it is based. Consequently certain sample and indicator preconditions need to be met before a taxometric analysis can be carried out.

Taxometric procedures

There are three general categories of non-redundant taxometric procedures: mean above minus below a cut (MAMBAC: Meehl & Yonce, 1994), maximum covariance/maximum eigenvalue (MAXCOV/MAXEIG: Meehl and Yonce, 1996, Waller and Meehl, 1998), and latent mode factor analysis (L-Mode: Waller & Meehl, 1998). Each procedure makes use of graphs in which the input indicator is aligned along the x- (longitudinal) axis and the output indicator is measured on the y- (vertical) axis.

Purpose

To provide an illustration of how the taxometric method can be applied to criminal justice data, the following set of analyses were conducted to determine whether a small group of delinquents from the 1945 Philadelphia birth cohort (Wolfgang, Figlio, & Sellin, 1972) known to be responsible for the majority of delinquent acts committed by the cohort formed a qualitatively distinct group or taxon.

Participants

The 1945 Philadelphia birth cohort is a group of 9,944 boys born in 1945 who lived in Philadelphia

Conclusion

The purpose of this paper was to describe the taxometric method and illustrate its potential utility to researchers, theorists, and practitioners in the fields of criminology and criminal justice. Readers interested in more information on how to apply the taxometric method to social and behavioral science data should consult a recently published paper by Ruscio, Ruscio, and Carney (2011). From the information presented in both papers, it would seem that taxometrics has a great deal to offer

References (69)

  • B.A. Green et al.

    The latent structure of alcohol use pathology in an epidemiological sample

    Journal of Psychiatric Research

    (2011)
  • P.E. Meehl

    Clarifications about taxometric method

    Journal of Applied and Preventive Psychology

    (1999)
  • M.G. Vaughn et al.

    The severe 5%: A latent class analysis of the externalizing spectrum in the United States

    Journal of Criminal Justice

    (2011)
  • M.S. Bartlett

    The statistical conception of mental factors

    British Journal of Psychology

    (1937)
  • D.J. Bauer et al.

    Distributional assumptions of growth mixture models: Implications for overextraction of latent trajectory classes

    Psychological Methods

    (2003)
  • J.G. Bernburg et al.

    Labeling: Life chances and adult crime. The direct and indirect effects of official intervention in adolescence on crime in early adulthood

    Criminology

    (2003)
  • J.M. Chaiken et al.

    Varieties of criminal behavior

    (1982)
  • C.M. Cleland et al.

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

    Psychological Reports

    (2000)
  • C.R. Cloninger et al.

    Inheritance of alcohol abuse: Cross-fostering analysis of adopted men

    Archives of General Psychiatry

    (1981)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • M. DeLisi

    Career criminals in society

    (2005)
  • M. DeLisi et al.

    All in the family: Gene x environment interaction between DRD2 and criminal father is associated with five antisocial phenotypes

    Criminal Justice and Behavior

    (2009)
  • J.F. Edens et al.

    Psychopathic, not psychopath: Taxometric evidence for the dimensional structure of psychopathy

    Journal of Abnormal Psychology

    (2006)
  • J.F. Edens et al.

    Exploring the taxometric status of psychopathy among youthful offenders: Is there a juvenile psychopath taxon

    Law and Human Behavior

    (2011)
  • D.P. Farrington

    Predictors, causes, and correlates of male youth violence

    Crime and Justice: A Review of Research

    (1998)
  • D.P. Farrington et al.

    Criminal penal and life histories of chronic offenders: Risk and protective factors and early identification

    Criminal Behaviour and Mental Health

    (1993)
  • Frazier, T. W., Ruscio, J., & Youngstrom, E. A. (submitted for publication). Comparing latent variable models and...
  • S. Glueck et al.

    500 criminal careers

    (1930)
  • M.R. Gottfredson et al.

    A general theory of crime

    (1990)
  • H.G. Grasmick et al.

    Testing the core implications of Gottfreson and Hirschi's general theory of crime

    Journal of Research in Crime and Delinquency

    (1993)
  • W.M. Grove

    The validity of cluster analysis stopping rules as detectors of taxa

  • W.M. Grove et al.

    Simple regression based procedures for taxometric investigation

    Psychological Reports

    (1993)
  • J.-P. Guay et al.

    A taxometric analysis of the latent structure of psychopathy: Evidence for dimensionality

    Journal of Abnormal Psychology

    (2007)
  • R.D. Hare

    Without conscience: The disturbing world of the psychopaths among us

    (1993)
  • R.D. Hare

    The Hare Psychopathy Checklist─Revised manual

    (2003)
  • G.T. Harris et al.

    Psychopathy as a taxon: Evidence that psychopaths are a discrete class

    Journal of Consulting and Clinical Psychology

    (1994)
  • N. Haslam

    The latent structure of personality and psychopathology: A review of trends in taxometric research

    Scientific Review of Mental Health Practice

    (2011)
  • N. Haslam et al.

    Taxometric analysis of fuzzy categories: A Monte Carlo study

    Psychological Reports

    (2002)
  • J.C. Karberg et al.

    Substance dependence, abuse, and treatment of jail inmates, 2002

    (2005)
  • M.F. Lenzenweger

    Consideration of the challenges, complications, and pitfalls of taxometric analysis

    Journal of Abnormal Psychology

    (2004)
  • G. Lubke et al.

    Performance of factor mixture models as a function of model size, covariate effects, and class-specific parameters

    Structural Equation Modeling

    (2007)
  • G. Lubke et al.

    Latent class detection and class assignment: A comparison of the MAXEIG taxometric procedure and factor mixture modeling approaches

    Structural Equation Modeling

    (2010)
  • D.K. Markus et al.

    Adjudicative competence: Evidence that impairment in “rational understanding” is taxonic

    Psychological Assessment

    (2010)
  • McDonald, R. P. (1967). Nonlinear factor analysis. Psychometric Monograph, No....
  • Cited by (14)

    • The latent structure of depression symptoms and suicidal thoughts in Brazilian youths

      2019, Journal of Affective Disorders
      Citation Excerpt :

      Even though a previous taxometric study in Brazil found evidence for dimensionality (Eulálio et al., 2015), the sample included only adults aged 60+ years. As previously suggested (Walters, 2012b), the latent structure of a disorder might be different across the developmental course. Furthermore, the latent structure of depression might feed upon the covariation patterns of a set of variables that are specific to a given culture.

    • Reconciling questions about dichotomizing variables in criminal justice research

      2013, Journal of Criminal Justice
      Citation Excerpt :

      Our findings also have implications for investigations of the taxometric properties of criminal justice constructs. While the debate as to whether a construct is truly continuous or truly categorical may be important for understanding the nature and origin of the construct (e.g., Walters, 2012), its resolution is not necessary for researchers to know how to best handle their data. Our results offer a way to “cut the Gordian knot” provided by the taxometric debate, since researchers can use continuous indicators and know that their results will be valid in all but a few specific circumstances.

    View all citing articles on Scopus
    View full text