ReviewAppropriate Use of Bifactor Analysis in Psychopathology Research: Appreciating Benefits and Limitations
Section snippets
Problematic Interpretation of Bifactor Models: Reliance on Global Model Fit
The major criticism of the bifactor model is its potential for overfitting (29). A common approach to evaluating structural models is to compare several possible models and then retain the model showing the best overall (global) fit statistics such as χ2, comparative fit index (CFI), Tucker–Lewis index (TLI), root mean square error of approximation (RMSEA), unbiased standardized root mean square residual (USRMR), and Akaike information criterion (AIC) (10,11). This approach is problematic
Useful Applications of the Bifactor Model
When a latent variable model is fit to psychopathology data [see (47,48) for discussions on choice of latent variable models versus alternatives such as network models], bifactor models are useful for their ability to separate indicator variance associated with a general factor from variance associated with narrower group factors or specific indicators. This separation of general variance from unique variance can inform several questions.
Comparison With Alternative Models
We describe several useful applications of bifactor models. This is not to suggest that they are a panacea or appropriate for all research questions. Below, we compare the bifactor model with several common alternatives and consider when these alternatives may be more useful.
The Bifactor Model and Biological Substrates of Psychopathology
A growing area of research examines biological substrates of psychological constructs such as neurobiological and genetic correlates of individual differences in personality, cognition, or psychopathology (98, 99, 100, 101). For example, several studies have examined or proposed correlations of psychopathology general and group factors with genetic single nucleotide polymorphisms or neurobiological variables (e.g., gray matter volume, volume or activation of amygdala/prefrontal cortex circuits,
Modeling Cannot Fix Inadequate Research Design
To close, we reiterate that statistical modeling cannot make fundamental limitations of data disappear. The questions that data can address are a function of the research design, not the model chosen to analyze them. Cross-sectional relationships among indicators cannot speak to developmental processes regardless of the type of model (bifactor or network) or indicator (behavioral symptoms or biological variables) used. The appropriate level of analysis for psychopathology (e.g., symptoms,
Acknowledgments and Disclosures
This work was supported by the National Institute on Drug Abuse (Grant No. DA032582 [to MAB]).
The authors report no biomedical financial interests or potential conflicts of interest.
References (119)
- et al.
“P” and “DP”: Examining symptom-level bifactor models of psychopathology and dysregulation in clinically referred children and adolescents
J Am Acad Child Adolesc Psychiatry
(2018) - et al.
Criterion validity and utility of the general factor of psychopathology in childhood: Predictive associations with independently measured severe adverse mental health outcomes in adolescence
J Am Acad Child Adolesc Psychiatry
(2018) - et al.
Opportunities for the prevention of mental disorders by reducing general psychopathology in early childhood
Behav Res Ther
(2019) - et al.
The general factor of personality: The “Big One,” a self-evaluative trait, or a methodological gnat that won’t go away?
Pers Individ Differ
(2015) - et al.
The temporal stability of the bifactor model of comorbidity: An examination of moderated continuity pathways
Compr Psychiatry
(2017) - et al.
Correlates of a general psychopathology factor in a clinical sample of childhood sexual abuse survivors
J Affect Disord
(2018) The higher-order model imposes a proportionality constraint: That is why the bifactor model tends to fit better
Intelligence
(2016)On the distortion of model fit in comparing the bifactor model and the higher-order factor model
Intelligence
(2016)- et al.
The limitations of model fit in comparing the bi-factor versus higher-order models of human cognitive ability structure
Intelligence
(2013) - et al.
When and why the second-order and bifactor models are distinguishable
Intelligence
(2017)