Mental health articles

McClelland et al. (1989) distinguished between implicit and explicit methods of personality assessment. Explicit measures, such as self-report instruments, assess psychological characteristics and needs individuals recognize about themselves and which they can articulate. In contrast, by analyzing a representative sample of an individual’s behavior during the assessment process, implicit or performance-based measures, such as the Rorschach or Thematic Apperception Test (Morray, 1943), tap into automatic patterns of behavior and motivations individuals may not be aware of or may not be willing or able to verbalize.

Although implicit or performance-based measures of personality may provide relatively limited information about the content of people’s thoughts or the specific symptoms they experience, performance-based measures, such as the Rorschach, may be particularly revealing of an individual’s underlying attitudes, coping styles, behavioral dispositions, and implicit motivations, even those about which they have little awareness. McClelland, Koestner, and Weinberger (1989) suggested that because implicit measures assess underlying patterns of thinking, reacting, and behaving that individuals may not recognize in themselves or which may not be consciously represented and articulated, implicit measures may be less susceptible to conscious control or fi ltering than explicit measures. In other words, implicit measures are less likely to be skewed by self-presentational concerns than explicit measures. For this reason, implicit measures may provide meaningful information about important psychological functioning that may not otherwise be acknowledged because an individual is not aware of certain aspects of their psychological operations, has not consciously formulated or labeled certain  aspects of their functioning, or is not willing to acknowledge what they do know (cf., Shedler, Mayman, & Manis, 1993). One implicit Rorschach measure of dependency needs is the number of responses coded for Texture. The link between Texture responses and a sense of basic security and trust in others that affects one’s comfort and willingness to become attached to others is not explicit.

In nonclinical populations, women who are coping with the effects of marital separation or divorce produced an elevated number of Texture responses, suggesting that their reaction to the loss of a marital relationship contributed to a sense of emotional deprivation and increased needs for closeness (Exner, 1993). Similarly, children placed in foster care for the fi rst time also produced an elevated number of Texture responses compared to their peers. This refl ects their reaction to separation from or loss of one or both parents. Other studies found that children in foster care who had frequent moves from one placement to another produced signifi cantly fewer Texture responses than their same-age peers who had lived with their parents since birth (Exner, 1993). These findings suggested the children in foster care did not have secure attachments and tended to back away from emotional involvement with others, most likely because they did not have consistent experiences of caretakers responding to their feelings and needs. Another example of an implicit measure related to social functioning is the number of cooperative movement responses (COP) produced on the Rorschach. COP signals an interest in being involved with others and an expectation that interactions have the potential to be friendly, harmonious, and mutually satisfying.

Note that even though the Rorschach does not directly inquire about social expectations or attitudes, the number of responses involving COP are related to social functioning. For example, Exner (1993) described a study of college freshman living in the same dormitory who rated one another on a number of dimensions. The college students who gave more than an average number of COP responses were fi ve times more likely than others to be rated as being “the most fun to be with” or “the easiest to be around.” As individuals who produce COP responses tend to have positive expectations concerning social interactions, their behavior is likely to be characterized as friendly, agreeable, and open to others and they are likely to be regarded positively by the people with whom they are involved.

Other studies investigating interpersonal dependency needs have found that scores obtained on explicit and implicit measures of dependency diverge in ways that are theoretically meaningful. Bornstein et al. (1993), for instance, compared male and female college students on a self-report measure of dependency and the Rorschach Oral Dependency Scale (ROD; Masling, Rabie, & Blondheim, 1967), an implicit measure of dependency based on responses to the Rorschach. Bornstein et al. (1993) hypothesized that since males are socialized to be strong, assertive, and independent, they are less likely to acknowledge dependent feelings and behaviors than women on objective, face valid measures. Consistent with this hypothesis, they found that although female college students consistently obtained higher scores than male college students when compared on self-report questionnaires assessing dependent traits, attitudes, and behaviors, male and female participants produced comparable scores on the ROD, an implicit measure of interpersonal dependency. They concluded that men and women have comparable underlying dependency needs, but that women are more willing than men to openly acknowledge these needs when asked directly about them using explicit measures of this construct. As discussed previously, McClelland et al. (1989) suggested that individuals can more easily control the impression they create on explicit than implicit psychological measures. Bornstein, Rossner, Hill, and Stepanian (1994) investigated this proposition in a series of studies examining whether individuals could consciously skew results on the ROD and a self-report measure, the Interpersonal Dependency Inventory (IDI; Blatt, D’Affl itti, & Quinlan, 1976). They found that scores on the IDI but not the ROD changed in response to experimental manipulations, such as instructions to respond to the measures as a dependent person would or when told that dependent characteristics were negative or positive. For instance, some participants were told that dependent characteristics interfered with interpersonal relationships and one’s ability to express emotions, while other participants were told that people who had more dependent characteristics had better relationships and were particularly good at reading other people.

Consistent with predictions, scores on the IDI shifted in response to experimental manipulations, while scores on the ROD did not. Bornstein et al. (1996) concluded that objective, explicit self-report measures of dependency are more susceptible to conscious manipulation than implicit dependency measures. This has the following implication: self-report measures can be quite effective when assessing psychological characteristics an individual recognizes and is willing to acknowledge. However, in other situations, such as when assessing traits or behaviors that are socially undesirable or which clash with an individual’s self-image, an implicit test which has low face validity may be preferable to a test with high face validity. In the latter situation, implicit measures, such as the Rorschach, may be preferred as they are less likely to be consciously biased or distorted by self-presentational concerns or deliberate efforts to create a desired impression than self-report methods. The Rorschach should be viewed as a valuable assessment tool because the empirically derived information it provides complements material obtained through self-report measures or clinical interview or provides a means to access data that otherwise may not be readily available (Ganellen, 1996). A study by Blais, Hilsenroth, Castelbury, Fowler, and Baity (2001) provides  another example of how the Rorschach functions as an implicit measure of psychological characteristics related to specifi c personality disorders. Blais et al. (2001) examined relationships among select Rorschach and MMPI-2 variables and DSM-IV-TR Cluster B personality disorder criteria. They entered theoretically relevant MMPI-2 and Rorschach variables into hierarchical regression analyses predicting DSM-IV-TR criteria for antisocial, borderline, histrionic, and narcissistic personality disorders. They found that both MMPI-2 and Rorschach variables added incrementally to the prediction of borderline and narcissistic personality disorder criteria scores while Rorschach variables independently predicted diagnostic criteria for histrionic personality disorder. For instance, the Rorschach variable that predicted histrionic personality disorder (FCCFC) relates to emotional reactivity and expressiveness, a key component of this disorder, whereas the Rorschach variable that predicted narcissistic personality disorder, the number of Refl ection responses, relates to an infl ated sense of self-worth, need for affi rmation from others, and a tendency to be more focused on oneself than others.  Consistent with the distinction drawn by McClelland et al. (1989), Blais et al. (2001) suggested that MMPI-2 scores tap material an individual is aware of and able to acknowledge, whereas Rorschach variables tap material that may be outside of his or her or awareness.

It is interesting to note that Rorschach variables, such as the number of Reflection responses and the number of responses involving color, were related to narcissistic and histrionic personality disorder criteria, respectively. As discussed previously, these Rorschach variables do not directly assess specific behaviors, mood states, or symptoms, such as narcissistic self-absorption, infl ated self-image, or emotional reactivity. Thus, the significant relationship between these Rorschach variables and diagnostic criteria was not found because these Rorschach variables are face valid, explicit measures of specific characteristics, experiences, or patterns of thinking and reacting an individual recognizes as being true for them. Instead, Rorschach variables presumably captured a dimension of personality functioning which the individual may not be aware applies to himself or herself or may not be able to articulate. These findings suggest that implicit measures can be used in a powerful manner to assess certain meaningful dimensions of psychological functioning. Meyer and Archer (2001) have raised some intriguing issues concerning the relationship between Rorschach variables, overt behavior and internal experiences, which may refi ne the distinction between explicit and implicit personality measures. They observed that no Rorschach variables are direct measures of specific psychological constructs or personality dimensions. For example, although some Rorschach variables are related to self-esteem issues and others are related to mood disorders, the Rorschach does not assess these constructs explicitly by asking directly how a person perceives himself or herself, as a self-report self-esteem questionnaire might, or whether they feel sad, have decreased levels of energy, and have trouble concentrating, as a self-report depression inventory might. In other words, the Rorschach “gets at” these constructs implicitly by relying on variables which are not face valid or obvious, but which have been shown empirically to be related to specifi c criterion variables, such as negative self-perceptions or aspects of a mood disorder or personality disorder.

Meyer and Archer (2001) noted that while Rorschach variables in general can be considered implicit measures of criterion variables, certain Rorschach variables may be more closely related to observable behavior than others. For instance, Meyer and Archer (2001) observed that Rorschach signs of disorganized thought processes, illogical thinking, and distorted, unrealistic perceptions are likely to be manifested in psychotic symptoms. In contrast, the link between other Rorschach variables, such as Refl ection responses, and overt behavior, such as a need to elicit affi rmation of self-worth from others, may be less direct and obvious. Meyer and Archer suggest that although both sets of variables (e.g., thought processes and perception, on the one hand, and Refl ection responses on the other) are implicit measures of psychological constructs, the strength of relationship between a Rorschach variable and an external criterion variable is likely to be stronger for those variables which are more directly linked to an external criterion variable than for those which are less clearly related to a criterion variable.

Tags: measures, personality