Does nature have joints worth carving? A discussion of taxometrics, model-based clustering and latent variable mixture modeling.
- Authors
- Lubke, G H; Miller, P J
- Year
- 2015
- Journal
- Psychological medicine
- PMID
- 25137654
- DOI
- 10.1017/S003329171400169X
- PMCID
- PMC4692716
Taxometric procedures, model-based clustering and latent variable mixture modeling (LVMM) are statistical methods that use the inter-relationships of observed symptoms or questionnaire items to investigate empirically whether the underlying psychiatric or psychological construct is dimensional or categorical. In this review we show why the results of such an investigation depend on the characteristics of the observed symptoms (e.g. symptom prevalence in the sample) and of the sample (e.g. clinical, population sample). Furthermore, the three methods differ with respect to their assumptions and therefore require different types of a priori knowledge about the observed symptoms and their inter-relationships. We argue that the choice of method should optimally match and make use of the existing knowledge about the data that are analyzed.
(a) Path diagram showing p observed variables, Y1−Yp, and an underlying latent variable (LV), which represents the construct. The LV can be either categorical (latent class variable) or continuous (latent factor or trait). (b) Path diagram for a hybrid model with a continuous factor (F) within each latent class of the latent class variable C.
(a) The three thin lines show the probability of endorsing three severe symptoms of an underlying continuous construct (thick curve). Only individuals with very high scores on the construct have a non-zero probability of endorsing the items, and a large proportion of the population will score zero on all items. (b) The thick lines show two latent classes, a taxon and a complement. The thin lines reflect again the probability of endorsing symptoms, which are here spread over the range of the construct.
(a) A smaller sample from a truly categorical construct consisting of a complement (black) and a taxon (red) might not suffice to detect taxonicity. There are too few red dots belonging to the taxon to clearly show a separate cluster. (b) In a larger sample, the power is more likely to be sufficient to detect the taxon.
(a) A skewed observed distribution (thick line) resulting from a population mixture distribution (dashed lines) consisting of three classes with equal variance. (b) The same skewed observed distribution can be due to a mixture with two components that have unequal variance.
| # | Section | Preview |
|---|---|---|
| 20 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | The objective of Meehl’s procedures was to address the question of continuous versus categorical… |
| 21 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | MAMBAC (mean above minus mean below a cut; Meehl & Yonce, 1994), MAXCOV-HITMAX (Meehl, 1995; Meehl &… |
| 22 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | The basic structure consists of first deriving theoretical expectations from premises, and then… |
| 23 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | The different procedures derive expectations from the same premises. In case of taxonicity the… |
| 24 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | The expectation concerning the pattern of covariances differs depending on the premise. Consider… |
| 25 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | the partition includes members of both groups then the covariance will be larger than zero. However,… |
| 26 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | The procedures are limited to two latent classes, a taxon and a complement class. Furthermore, the… |
| 27 | Three methods to decide between latent dimensionality and latent categories — Meehl’s taxometric procedures | As explained, the different procedures share the same premises regarding taxonicity and… |
| 28 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | This is a clustering approach based on a probability model. The key idea is to fit alternative… |
| 29 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | The starting point is that the observed data have a multivariate distribution, and that if there are… |
| 30 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | whether the clusters are meaningful; for instance, whether subjects in the tail of the skewed… |
| 31 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | In model-based clustering the mixture component distributions are most commonly assumed to be… |
| 32 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | The framework offers great flexibility for comparing alternative models that differ with respect to… |
| 33 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | To summarize, model-based clustering is based on the assumptions that (1) each component… |
| 34 | Three methods to decide between latent dimensionality and latent categories — Model-based clustering | Fraley & Raftery (2002) mention the possibility of modeling the within-cluster covariance matrices… |
| 35 | Three methods to decide between latent dimensionality and latent categories — LVMM | LVMM is similar to model-based clustering in that the user has to choose a distribution for the… |
| 36 | Three methods to decide between latent dimensionality and latent categories — LVMM | The general model framework permits fitting models with a specific factor structure that relates the… |
| 37 | Three methods to decide between latent dimensionality and latent categories — LVMM | Using LVMMs to distinguish between latent dimensions and categories requires great care when… |
| 38 | Three methods to decide between latent dimensionality and latent categories — LVMM | As in the case of model-based clustering, models can be estimated using the Expectation-Maximization… |
| 39 | Comparison of the three methods — Summary of assumptions | Taxometrics, model-based clustering and LVMM differ with respect to their assumptions and to the… |
| Name | Type |
|---|---|
| ADHD | phenotype |
| age | phenotype |
| attention problems | phenotype |
| cigarettes | phenotype |
| clinical population | cohort |
| depression | phenotype |
| dimensional disorder local | phenotype |
| dimensionality local | phenotype |
| disorder | phenotype |
| DSM categorizations local | phenotype |
| factor model local | phenotype |
| general population | cohort |
| latent class model local | phenotype |
| Latent construct local | phenotype |
| mood disorders | phenotype |
| neuroticism | phenotype |
| personality disorders | phenotype |
| psychiatric disorders | phenotype |
| race/ethnicity | phenotype |
| sex | phenotype |
| subjects | cohort |
| symptom | phenotype |
| symptoms | phenotype |
| taxonicity local | phenotype |
No uploaded files.
In this knowledge base
| Title | Year | PMID |
|---|---|---|
| Using Patterns of Genetic Association to Elucidate Shared Genetic Etiologies Across Psychiatric Disorders. | 2017 | 28343281 |
External
| Title | Authors | Journal | Year | Link |
|---|---|---|---|---|
| Can MMPI-2-RF confirm the structural validity of the somatoform spectrum of the Hierarchical Taxonomy of Psychopathology (HiTOP) in patients with chronic low back pain. | Sadeghi S et al. | — | 2025 | → |
| Heterogeneity in high hypnotic suggestibility: Models, evidence, and synthesis. | Terhune DB | — | 2025 | → |
| On the structure of psychoeducational constructs: taxometric analysis and epistemological implications. | Stamovlasis D et al. | — | 2025 | → |
| Clinical subtyping using community detection: Limited utility? | Agelink van Rentergem JA et al. | — | 2023 | → |
| Parent-Child Discrepancies in Perceived Parental Emotion Socialization: Associations with Children's Internalizing and Externalizing Problems in Chinese Families. | Fan J et al. | — | 2023 | → |
| Probing the overarching continuum theory: data-driven phenotypic clustering of children with ASD or ADHD. | Deserno MK et al. | — | 2023 | → |
| The Somatoform Spectrum Within the Hierarchical Taxonomy of Psychopathology System: A Taxometric Test of the Latent Structure. | Hartmann J et al. | — | 2022 | → |
| Are body dysmorphic symptoms dimensional or categorical in nature? A taxometric investigation in adolescents. | Bala M et al. | — | 2021 | → |
| Balancing typological and dimensional approaches: Assessment of adult attachment styles with Factor Mixture Analysis. | Morales-Vives F et al. | — | 2021 | → |
| Evidence for the taxonic latent structure for DSM-5 intermittent explosive disorder in adults. | Fanning JR et al. | — | 2021 | → |
| Heterogeneity and Factorial Structure in Alzheimer's Disease: A Cognitive Perspective. | Zangrossi A et al. | — | 2021 | → |
| Les progrès dans la réalisation de la classification quantitative de la psychopathologie. | Krueger RF et al. | — | 2021 | → |
| A valid and reliable measure of nothing: disentangling the "Gavagai effect" in survey data. | Arias VB et al. | — | 2020 | → |
| Assessing the heterogeneity of autism spectrum symptoms in a school population. | Morales-Hidalgo P et al. | — | 2018 | → |
| Illustration of Step-Wise Latent Class Modeling With Covariates and Taxometric Analysis in Research Probing Children's Mental Models in Learning Sciences. | Stamovlasis D et al. | — | 2018 | → |
| Longitudinal Model Building Using Latent Transition Analysis: An Example Using School Bullying Data. | Ryoo JH et al. | — | 2018 | → |
| Progress in achieving quantitative classification of psychopathology. | Krueger RF et al. | — | 2018 | → |
| An Empirical Assessment of the Sensitivity of Mixture Models to Changes in Measurement. | Cole VT et al. | — | 2017 | → |
| Discrete category of mother-to-infant bonding disorder and its identification by the Mother-to-Infant Bonding Scale: A study in Japanese mothers of a 1-month-old. | Matsunaga A et al. | — | 2017 | → |
| Disentangling Heterogeneity of Childhood Disruptive Behavior Problems Into Dimensions and Subgroups. | Bolhuis K et al. | — | 2017 | → |
| Fitting latent variable mixture models. | Lubke GH et al. | — | 2017 | → |
| Person-centered approaches in the study of personality disorders. | Hicks BM et al. | — | 2017 | → |
| The Social Aptitudes Scale: looking at both "ends" of the social functioning dimension. | Axelrud LK et al. | — | 2017 | → |
| Using a hybrid subtyping model to capture patterns and dimensionality of depressive and anxiety symptomatology in the general population. | Wardenaar KJ et al. | — | 2017 | → |
| Using Patterns of Genetic Association to Elucidate Shared Genetic Etiologies Across Psychiatric Disorders. | Cho SB et al. | — | 2017 | → |
| Getting less of what you want: reductions in statistical power and increased bias when categorizing medication adherence data. | Tueller SJ et al. | — | 2016 | → |
| Kinds versus continua: a review of psychometric approaches to uncover the structure of psychiatric constructs. | Borsboom D et al. | — | 2016 | → |
| Latent variable mixture modeling in psychiatric research--a review and application. | Miettunen J et al. | — | 2016 | → |
| Psychometric properties and a latent class analysis of the 12-item World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) in a pooled dataset of community samples. | MacLeod MA et al. | — | 2016 | → |
| Disorders without borders: current and future directions in the meta-structure of mental disorders. | Carragher N et al. | — | 2015 | → |