A considerable amount of research has been done on Young’s Early Maladaptive Schemas. Most research conducted thus far has been done using the long form of the Young Schema Questionnaire (Young & Brown, 1990), although studies with the short form are in progress. The Young Schema Questionnaire has been translated into many languages, including French, Spanish, Dutch, Turkish, Japanese, Finnish, and Norwegian.
The first comprehensive investigation of its psychometric properties was conducted by Schmidt, Joiner, Young, and Telch (1995). Results from this study produced alpha coefficients for each Early Maladaptive Schema that ranged from .83 (Enmeshment/Undeveloped Self) to .96 (Defectiveness/Shame) and test-retest coefficients from .50 to .82 in a nonclinical population. The primary subscales demonstrated high test-retest reliability and internal consistency. The questionnaire also demonstrated good convergent and discriminant validity on measures of psychological distress, self-esteem, cognitive vulnerability to depression, and personality disorder symptomatology.
The investigators conducted a factor analysis using both clinical and nonclinical samples. The samples revealed similar sets of primary factors that closely matched Young’s clinically developed schemas and their hypothesized hierarchical relationships. Within one sample of undergraduate college students, 17 factors emerged, including 15 of the 16 originally proposed by Young (1990). One original schema, Social Undesirability, did not emerge, whereas two other unaccounted factors did. In an effort to cross-validate this factor structure, Schmidt et al. (1995) gave the Young Schema Questionnaire to a second sample of undergraduates taken from the same population. Using the same factor-analytic technique, the investigators found that, of the 17 factors produced in the first analysis, 13 were clearly replicated in the second sample. The investigators also found three distinct higher order factors. Within a sample of patients, 15 factors emerged, including 15 of the 16 originally proposed by Young (1990). These 15 factors accounted for 54$ of the total variance (Schmidt et al., 1995).
In this study, the Young Schema Questionnaire demonstrated convergent validity with a test of personality disorder symptomatology (Personality Diagnostic Questionnaire—Revised; Hyler, Rieder, Spitzer, & Williams, 1987). It also demonstrated discriminant validity with measures of depression (Beck Depression Inventory; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and self-esteem (Rosenberg Self-Esteem Questionnaire; Rosenberg, 1965) in a nonclinical undergraduate population.
This study was replicated by Lee, Taylor, and Dunn (1999) using an Australian clinical population. The investigators conducted a factor analysis. In accord with previous findings, 16 factors emerged as primary components, including 15 of the 16 originally proposed by Young. Only the Social Undesirability scale was not supported. (We have since eliminated Social Undesirability as a separate schema and merged it with Defectiveness.) In addition, a higher order factor analysis closely fit some of the schema domains proposed by Young. Overall, this study shows that the Young Schema Questionnaire possesses very good internal consistency and that its primary factor structure is stable across clinical samples from two different countries and for different diagnoses.
Lee and his colleagues (1999) discuss some reasons that the two studies produced somewhat different factor structures depending on whether a clinical or normal population was used. They conclude that the student samples probably had range effects, as it was unlikely that many of the students were suffering from extreme forms of psychopathology. The authors state that factor structure replication depends on the assumption that the schemas underlying psychopathology in clinical populations are also present in a random sample of college students. Young suggests that Early Maladaptive Schemas are indeed present in normal populations but that they become exaggerated and extreme in clinical populations.
Other studies have examined the validity of the individual schemas and how well they support Young’s model. Freeman (1999) explored the use of Young’s schema theory as an explanatory model for nonrational cognitive processing. Using normal participants, Freeman found that weaker endorsement of Early Maladaptive Schemas was predictive of greater interpersonal adjustment. This finding is consistent with Young’s tenet that Early Maladaptive Schemas are by definition negative and dysfunctional.
Rittenmeyer (1997) examined the convergent validity of Young’s schema domains with the Maslach Burnout Inventory (Maslach & Jackson, 1986), a self-report inventory designed to assess the negative impact of stressful life events. In a sample of California schoolteachers, Rittenmeyer (1997) found that two schema domains, Overconnection and Exaggerated Standards, correlated strongly with the Emotional Exhaustion scale of the Maslach Burnout Inventory. The Overconnection schema domain also correlated, although not as strongly, with two other inventory scales, Depersonalization and Personal Accomplishment.
Carine (1997) investigated the utility of Young’s schema theory in the treatment of personality disorders by using Early Maladaptive Schemas as predictor variables in a discriminant function analysis. Specifically, Carine looked at whether the presence of Young’s schemas discriminated patients with DSM-IV Axis II psychopathology from patients with other types of psychopathology. Carine found that group membership in the Axis II cluster was predicted correctly 83% of the time. In support of Young’s theory, Carine also found that affect appears to be an intrinsic part of schemas.
Although the Young Schema Questionnaire was not designed to measure specific DSM-IV personality disorders, significant associations appear between Early Maladaptive Schemas and personality disorder symptoms (Schmidt et al., 1995). The total score correlates highly with the total score on the Personality Diagnostic Questionnaire—Revised (Hyler et al., 1987), a self-report measure of DSM-III-R personality pathology. In this study, the schemas of Insufficient Self-Control/Self-Discipline and Defectiveness had the strongest associations with personality disorder symptoms. Individual schemas have been found to be significantly associated with theoretically relevant personality disorders. For example, Mistrust/Abuse is highly associated with paranoid personality disorder; Dependence is associated with dependent personality disorder; Insufficient Self-Control/Self-Discipline is associated with borderline personality disorder; and Unrelenting Standards is associated with obsessive-compulsive personality disorder (Schmidt et al., 1995).