A taxometric analysis of type-D personality.
Summary (2 min read)
Introduction
- Published in Psychosomatic Medicine, Volume 71, Issue 9, pp. 981 - 986, by Lippincott, Williams & Wilkins / American Psychosomatic Society.
- Whether Type-D is dimensional (i.e., distributed as a continuous variable, with individuals varying quantitatively from each other) or taxonic (i.e., individuals are differentiated into non-arbitrary groups or categories) has important clinical and theoretical implications [8].
- While exact mechanisms may not be clear at present the underlying principle is that the existence of dimensional and taxonic models requires very different theoretical accounts.
Evidence for the Dimensionality or Taxonicity Type-D
- Recently, psychometric work using item response theory (IRT) was interpreted as support for a categorical interpretation of Type-D using a cutoff score of 10 on each of the two sub-scales (NA and SI) of the DS14 [11].
- While this evidence appears compelling, it does not demonstrate that Type-D is taxonic, just that there are between-group differences based on median splits.
- Given the very different implications for theory, research and clinical practice afforded by dimensional and taxonic conceptualizations, a direct test of the dimensionality of Type D is urgently required [8].
- Indeed in the case of Type-D the evidence shows that the prevalence of Type-D is significantly higher in clinical groups with cardiovascular disease compared to healthy controls [14].
Sample and Sampling
- Second, of the valid indictors those with high item-total correlations are retained as they represent the most valid indicators of the construct [26].
- With divisions represented on the x-axis and mean differences, covariance or eigenvalues on the y-axis, characteristic curves represent either a taxonic or dimensional solution.
- MAXEIG is used when there are more than 3 indicators, this is an extension of MAXCOV, where the 1st eigenvalue is calculated across successive windows on the remaining output variables.
- This means using a number of taxometric procedures to show consistency across Taxometrics of Type-D methods and showing the same results while successively increasing the number of cuts/windows [30, 32].
Prevalence of Type-D
- As Type-D is hypothesized to be an independent risk factor for cardiovascular disease, this prevalence rate appears high and in fact highlights how the simple taxonic model can be somewhat problematic as a prognostic risk factor in a healthy population.
- The indicator validities (expressed in standard units as Cohen’s d), item-total correlations and skew for each indicator are presented in Table 1.
- The second factor represents finding social contact inhibiting and uncomfortable and the authors refer to this as a social discomfort factor.
- Similarly mean nuisance covariance for the 3 indicators in the MAXCOV analyses for the taxon and complement at 50 windows was .23 and .26 respectively; at 150 windows they were .23 and .20 and for 200 windows they were .11 and .24 respectively.
Taxometric Analyses
- The pattern of results reported here indicates clearly that Type-D is better represented as a dimensional rather than a categorical construct.
- This finding has important theoretical and clinical implications for Type-D. Theoretical models should focus on additive multi-causal agents or risk factors [8].
- More research is now needed to explore the additive nature of these different mechanisms.
Type-D and a Dimensional Construct in Research and Practice
- Based on the above, theoretical models and clinical interventions should be examined using regression approaches with large unselected samples [8] which also have the advantage of increasing statistical power [33].
- The remaining MAMBAC and MAXCOV curves are available form the 1st author on request.
- It has been argued that it is appropriate to draw distinctions within a latent dimensional construct as long as these are systematic and empirically justifiable [8].
- An inflection point expresses the association between the continuous dimensional construct and other relevant clinical data (e.g., bio-markers, prognostic clinical outcomes or treatment responses).
- With respect to diagnosis it should also be noted that Type-D is a risk factor and not a clinical diagnosis in itself and should be Taxometrics of Type-D used in conjunction with other information.
Taxometrics and Psychosomatic Medicine: Functional Syndromes and Bio-
- It has been strongly argued that taxometric approaches have implications for many fields beyond their current application mainly to investigating psychopathology (e.g., depression, personality disorder) [15].
- Within the field of psychosomatic medicine these may be applied initially to the dimensionality of traits directly relevant to psychosomatic medicine.
Conclusions
- The present taxometric analysis indicates that Type-D is better represented as a dimensional construct.
- Denollet J, Pedersen SS, Vrints CJ, Conraads VM.
- Usefulness of type D personality in predicting five year cardiac events above ad beyond current symptoms of stress in patients with coronary heart disease.
- Detecting taxonicity with two quantitative indicators using means above and below a sliding cut (MAMBAC procedure), also known as Taxometric analysis I.
Did you find this useful? Give us your feedback
Citations
474 citations
270 citations
194 citations
Cites background from "A taxometric analysis of type-D per..."
...Even for the type D measurement, a dimensional operationalization rather than a categorical structure is supported [59]....
[...]
173 citations
161 citations
Cites background from "A taxometric analysis of type-D per..."
...…show the potential clinical applications of including entrapment as a key predictor of repeat suicide attempts; however, much more work is needed to show that these scales are indeed taxonic and that the cutoffs vary meaningfully with external criteria (see Ferguson, 2009; Ferguson et al., 2009)....
[...]
References
29,764 citations
6,270 citations
"A taxometric analysis of type-D per..." refers methods in this paper
...examined using regression approaches with large unselected samples [8] which also have the advantage of increasing statistical power [33]....
[...]
1,323 citations
849 citations
636 citations
Related Papers (5)
Frequently Asked Questions (10)
Q2. What are the future works in "A taxometric analysis of type-d personality" ?
Future theorizing and research examining the links between Type-D and cardiovascular disease should consider dimensional approaches in order to move this area of inquiry forward.
Q3. What is the basic principle of Meehl’s taxometric technique?
The basic principle of Meehl’s taxomeric technique istermed ‘coherent cut kinetics’ where indictor variables are split into input and output variables.
Q4. What is the idea that Type-D is a dimensional construct?
The idea that Type-D is a dimensional construct comes from the finding that anumber of individual differences theoretically and empirically related to the subcomponents of Type-D – worry, depression, anxiety – [14, 22-23] are dimensional [8].
Q5. What is the mean nuisance covariance for the taxon and complement?
Within the putative taxon and complement (using the recommended cutoffpoint of ≥ 10 on both NA and SI) the nuisance covariance for NA and SI for the taxon is .30 and for the complement it is -.09.
Q6. What is the question about the latent-dimensional structure of Type D?
the question addressed herein is about the latent-dimensional structure of Type D: if the construct is a general personality trait then its latent structure should be stable across populations.
Q7. What is the recent study to use item response theory?
psychometric work using item response theory (IRT) wasinterpreted as support for a categorical interpretation of Type-D using a cutoff score of 10 on each of the two sub-scales (NA and SI) of the DS14 [11].
Q8. What is the mean nuisance covariance for the 3 indicators in the MAXCOV analyses?
Similarly mean nuisance covariance for the 3 indicators in the MAXCOV analyses for the taxon and complement at 50 windows was .23 and .26 respectively; at 150 windows they were .23 and .20 and for 200 windows they were .11 and .24 respectively.
Q9. What is the way to test the effects of Type-D?
Given the dimensional nature of Type-D these potential synergistic effects of Type-D should be tested using both additive and multiplicative regression terms [see 18].
Q10. What is the main argument for using median splits to identify Type-D cases?
Using median splits to identify Type-D cases amounts essentially to using arbitrary cutoffs which others have argued against as a basis for suggesting taxonicity [cf. 9].The other type of evidence that could be offered to suggest that Type-D istaxonic is based on the following argument: Using these cutoffs Type-D demonstrates good prognostic outcomes.