The first research objective of the present study was to examine the construct validity of MCMI with IN-EX model of psychopathology using the collected data. The result of ML in CFA indicated the MCMI was an adequate scale with a two-factor model, including the internalizing-externalizing model of psychopathology. Fit statistics of CFA (CFI, IFI and RMSEA) showed that MCMI with the internalizing and externalizing model of psychopathology fitted with the data collected from a developing country. Psychopathology is ideally classified in terms of the internalizing and externalizing models. The eighteen disorders; schizoid, avoidant, depressive, dependent, sadistic, negativistic, masochistic, schizotypal, borderline, paranoid, anxiety, somatoform, bipolar manic, dysthymia, posttraumatic stress disorder, thought disorder, major depression, and delusional disorder were ideally loaded on the internalizing form of psychopathology. In all, ten disorders were classified under Axis II (personality disorders), whereas the rest (eight disorders) were related to Axis I (clinical syndromes).The three disorders, that is, antisocial, alcohol, and drug were excellently loaded on the externalizing form of psychopathology. These findings of the internalizing and externalizing are consistent with the results of previous studies [4, 7–19, 21, 22, 24, 29, 37, 46, 64].
However, it should be noted that these previous studies had found a small number of disorders (four to six disorders) representing and manifesting the internalizing model of psychopathology. The structure of internalizing psychopathology in this research was extended to ten personality disorders and eight clinical syndromes. It captured the DSM-V model of psychopathology: Axis II (personality disorders) and Axis II (clinical syndromes). Uniquely, the internalizing model in the current study validated more mental disorders used in previous researches on psychopathology. This is congruent with Millon’s theory, which considers the importance and dominance of the broad general maladjustment factor with loadings from many scales/disorders [38].
One plausible interpretation is that the expanded scope of internalizing disorders is justifiable since they are attributable to a shared set of genetic factors [4, 9]. Røysamb et al. [65] found that the internalizing factor contained six indicators from Axis I (clinical disorders), namely anxiety disorder, major depressive disorder, dysthymia, anorexia nervosa, and post-traumatic stress, and pain disorder. In addition, the two indicators from Axis II (personality disorders): depressive and borderline, were loaded partly on this factor.
The current study extends existing disorders of the internalizing psychopathology structure by including various Axis I and Axis II psychopathology. Theoretically, this expanded the conceptualization of internalizing psychopathology which is consistent with the ideas of Krueger and Markon [23], suggesting that the internalizing-model can be tested with many mental disorders. Empirically, the result of the study is similar to the results of a previous study [16] that analyzed the correlation matrix of MCMI [38]. The set of internalizing and externalizing in the current research and previous study [16] are somewhat harmonious because of the two studies in terms of instrument and analyses procedures.
Although the current study expands the model of internalizing-externalizing psychopathology, there are notable differences between the current model and previous models. For example, previous studies presented/delineated internalizing psychopathology with two underlying factors: fear and distress [4, 10, 17, 21, 24], whereas the internalizing model in the current study did not bifurcate into distinct distress and fear components of internalizing disorders.
The externalizing psychopathology in this research included antisocial, alcoholic and drug disorders. The results of the present study analysis are strongly consistent with the results of previous studies [4, 9, 12, 16, 17, 64, 65], which demonstrated that antisocial, alcoholic and drug disorders all reflect an externalizing dimension. Substance abuse (e.g. alcohol and drug) and antisocial behaviour are genetically linked to an impulsive personality style [66]. In addition, these are both externalizing disorders indicators as the emotional release is directed outward without any restriction[67], sharing the fact that the expression of the detachment from social values maybe from childhood.
The second research objective was to evaluate the evidence of convergent validity and discriminant validity of MCMI with internalizing-externalizing model of psychopathology. The result of ML in CFA indicated the MCMI with the two-factor model (internalizing-externalizing model of psychopathology) was evidently characterized by convergent and discriminant validity. Given the ideal factor loadings of MCMI and the large proportion of variance, excellent reliability, and high variance extracted, it is evident that the disorders within each construct of psychopathology measure consistently and sufficiently their respective construct: internalizing-externalizing of psychopathology. It can be concluded that a convergent validity adequately characterized MCMI with the internalizing-externalizing model of psychopathology. These findings are consistent with the results of the previous studies [16, 40, 44, 45]. Similar findings across five discriminant validity evidences confirmed that MCMI with the internalizing-externalizing model of psychopathology were characterized by discriminant validity.
The value of the latent correlation was .73 (p ≤ 0.000), which was consistent with previous studies [4, 19]. The squared correlation between two constructs of psychopathology was less than their average variance-extracted. These techniques of discriminant validity predict that each construct of psychopathology assesses distinct concepts and meanings in the hypothesized model. This finding supports some previous studies which verified the distinction between internalizing and externalizing psychopathology [23]. Therefore, it can be concluded that the two-factor model of MCMI was differently designed to be assessed, indicating that the internalizing psychopathology construct was distinct from the externalizing psychopathology construct; this was good evidence of discriminant validity. Next, the two factors for MCMI exhibited discriminant validity, which strongly confirmed that the two-factor model of psychopathology for MCMI appeared to be the most favorable. Psychopathology with more than two-factor or single factors had unacceptable fit indices. These findings of the discriminant validity were consistent with a previous study that aimed to test the discriminant validity of MCMI regardless of the different methods of analysis [37].
4.1 Strengths and Limitations
The current research is the first study to verify psychometric evaluations of construct validity evidence of MCMI in a developing country using a sample of patients from outside Western culture.
This study is a novel attempt to offer evidence of the fit of the adult psychopathology model in terms of internalizing and externalizing forms using a psychiatric sample from a developing society.
The research expanded the structure of internalizing psychopathology to capture simultaneously ten personality disorders and eight clinical syndromes. It is consistent with the ideal vision that new research would expand the scope of the structure of internalizing and externalizing psychopathology (1-3). Subsequently, the organization of psychopathology in terms of internalizing and externalizing forms in the developing society is evaluated as a modern conceptual framework for adult psychopathology. The study provides original information on MCMI as a third clinical scale on clinical population using Maximum Likelihood (ML) in Exploratory Factor Analysis (EFA) and Structural Equation Modeling.
Although the results of this study are significant, certain limitations should be addressed for future research. The first limitation is that the important findings from this study were based on correlational analyses, relying on participants’ self-report of psychopathology measures. Therefore, developmentally longitudinal studies are needed to verify the findings reported here. Moreover, although the MCMI with the two-factor model: internalizing and externalizing psychopathology was evident by its construct validation, replication of this study is needed, using different methodologies such as Multitrait-Multimethod [68](Campbell & Fiske, 1959)(Campbell & Fiske, 1959)[68][68][68][68](Campbell and Fiske, 1959)[66]1(4) (e.g., self-reports, clinical observations, and interviews) and different samples (e.g., prisoners, patients, and community).
4.2 Implications and Recommendations
The results of the current study carry significant implications for clinical researchers and psychiatrists in hospitals. The MCMI is defined theoretically by internalizing and externalizing forms of psychopathology in clinical populations, which was significantly characterized by construct validity. Therefore, clinicians, clinical psychologists and researchers can assess psychopathology / mental disorders and personality disorders using MCMI in clinical and non-clinical populations with high confidence in terms of psychometric aspects (validity and reliability) in a developing society. Moreover, MCMI is a clinically accurate assessment for the diagnosis of mental disorders. It can be used in treatment to track changes over time by looking at patients’ scores on MCMI’s scales elevations.