The main purpose of this study was to modify Dombrose and Slobin's Picture Titles Test so that it could be scored objectively. The test was used for measuring impulse, ego, and superego sex differences and for comparing age group differences of both sexes, on their impulse, ego, and superego responses. The possibility of bias in the subjects’ responses was also investigated. The modification of the Picture Titles Test involved the construction of a multiple-choice answer sheet which required the subjects simply to check off their responses. The test was then administered as a group test to a second year psychology class at the University of British Columbia, of which ninety females and one hundred and thirty-seven males, ranging in age from eighteen to twenty-two, remained for the statistical analysis. Twelve hypotheses were set up concerning sex differences and age group differences, and both the Chi-square and t test were used for testing the hypotheses. The hypotheses had to be supported at the .05 level of confidence in order to be confirmed. The results indicated that the objectively scored Picture Titles Test is a useful instrument for differentiating males and females and between age groups as there were four significant differences found, out of a possible nine, for the male and female comparisons and three, out of a possible six, for the age group comparisons. Only one of the significant differences for the age group comparisons was in the predicted direction, however, so the age difference hypotheses were not considered to be verified. There were three significant differences in the predicted direction for the male and female comparisons, but the results were so inconsistent for the different age groups of males and females compared that the hypotheses concerning the expected sex differences were also not considered to be confirmed. In addition, it was found that subjects within each age and sex group responded to the majority of pictures in the same way, and, consequently, it was concluded that there is bias probably both in the content of the pictures and the blocks of titles. In general, it was felt that the objectively scored Picture Titles Test is an improvement over the original subjectively scored version and that the modified Picture Titles Test has good potential as an instrument for measuring impulse, ego and superego sex and age differences. / Arts, Faculty of / Psychology, Department of / Graduate

Four studies were conducted with the aim of developing a measure of narcissism that, unlike previous measures, is not inherently pathological. In Study 1, the NPI—the closest approximation to such a measure—was administered to a large-scale sample. Two separate Principal Component Analyses (PCAs) failed to replicate its reported structure, and revealed further psychometric problems. Use of an alternative item format (Likert ratings) was shown to yield much higher reliabilities than the original forced-choice format. Using the Likert item-format, the Profile of Narcissistic Dispositions (POND) was developed in Study 2. The items were based on a comprehensive analysis of the literature on "normal" narcissism. Five reliable subscales emerged from an oblique factor analysis. All five loaded substantially on the first unrotated principal component. The relationship of the POND to established self-report measures was also explored. In Study 3, the POND's structure was replicated and empirical relations were expanded: In particular, the POND showed negative correlations with various self-reports of psychopathology. In Study 4, the POND was shown to predict peer ratings of narcissism. Further peer-ratings elaborated the character of normal narcissism, that is, an interpersonal style that is marked by a dominant and secure but disagreeable social presence. / Arts, Faculty of / Psychology, Department of / Graduate

D.Litt. et Phil. / The psychometric properties of the Comrey Personality Scales (CPS) for Afrikaans speaking university students were examined. The CPS and the 16 Personality Factor Questionnaire (16PF) were completed by 804 first year university students in Afrikaans. The CPS was translated into Afrikaans by the translation-back translation method. Scores for the 40 subscales of the CPS were subjected to an exploratory factor analysis. Nine factors were extracted and orthogonally rotated with the Tandem Criteria. Subsequently, the factors were compared with the eight factors obtained in the original normative study by means of coefficients of congruence. Six of the factors, namely Trust versus Defensiveness, Orderliness versus Lack of Compulsion, Emotional Stability versus Neuroticism, Extroversion versus introversion, Mental Toughness versus Emotional Sensitivity, and Empathy versus Egocentrism, showed congruence coefficients of 0,90 or higher. The Social Conformity versus Rebelliousness factor was clearly recognisable in the Afrikaans speaking sample, but the congruence coefficient was only 0,81. The Activity versus Lack of Energy factor split in two and one of these factors was very similar to the normative Activity factor, with a congruence coefficient of 0,89. The CPS subscales were also subjected to a multiple group confirmatory factor analysis wherein the measurement invariance of the CPS for Afrikaans speaking students and the normative American group was assessed. The results confirmed that the CPS measures the same constructs in the two groups, but the variance of the Afrikaans group for the Social Conformity versus Rebelliousness factor was much smaller than that of the American group. The measurement invariance of the CPS for Afrikaans speaking males and females was also assessed. The multiple groups confirmatory factor analysis supported the measurement invariance of the CPS for males and females. Subsequently, the mean scores of American male students and the Afrikaans speaking male students for the CPS were compared. The Afrikaans speaking students obtained higher scores for Social Conformity versus Rebelliousness and lower scores for Trust versus Defensiveness than their American counterparts. Similar results were obtained for the Afrikaans speaking and American females. The mean scores of the Afrikaans speaking males and females were also compared. The comparison of the males and females revealed that the males had higher Mental Toughness versus Emotional Sensitivity scores. The females had higher scores for Empathy versus Egocentrism, Trust versus Defensiveness, Orderliness versus Lack of Compulsion, and Social Conformity versus Rebelliousness. In the last place the CPS and the 16PF scales were subjected to an interbattery factor analysis. Six factors were extracted and orthogonally rotated. The rotated factors supported the construct validity of the CPS and 16PF scales and also provided support for four of the so-called Big Five factors of personality.

No description available.

Following a discussion of the relationship between abnormal personality and psychopathology, and the concept of validity in psychometrics, the development and present form of the GPPT are described. Several important criticisms of the test are then detailed together with a review of all published research relating to it. Considering the HSRC's interest in standardising this instrument for use in South Africa, and since it is being extensively used for individual assessments within Stikland Hospital, four studies have been undertaken to clarify its validity: Study 1 concerns normative data and shows, from the test-results of 100 Afrikaans working adults of Bellville, that South African norms can be expected to differ radically from those obtained in the USA. Study 2 concerns the construct validity of the GPPT scales. Based on data from 168 Stikland patients, correlations between GPPT scores and a large number of other personality measures indicate that some of the interpretations offered by the test-developers are invalid for this population. Study 3 concerns the validity of the test as a measure of "mental health", and shows that, where significant, score deviations are related to neurotic rather than to sociopathic or psychotic maladjustment. Age and sex were not systematically related to score variance except for the "Withdrawal" scale, where a slight tendency was found for females to score higher. No single GPPT scale, including the composite "total score", was effective for reliable individual assessment of mental health. Study 4 concerns the test's validity as a measure of adjustment by investigating the scores before and after treatment of certain patients who improved dramatically. Results indicate that the TRQ and Total scores are effective in this respect.

Within the past decade the field of clinical psychology has become increasingly concerned with projective tests as a means of facilitating the understanding of the dynamic makeup of the individual personality. The enthusiasm over these techniques has resulted in the development of a barrage of projective techniques, purporting to uncover dynamics and "complexes" to a sometimes fantastic and often dubious extent. The subjective nature of these tests appears to render them most difficult to quantify and validate experimentally. Although the Mosaic test dates its origin twenty-five years back, there have been few, if any, successful validation studies. Those that have been attempted, have been poorly executed and their results inconclusive, or incongruous with other findings. The bulk of the literature on the test seems to consist of enthusiastic affirmation and ways and means of interpretation but little objective data to substantiate the findings. The test is considered by many to be most successful in its sensitivity to various entities of psychiatric diagnosis. That is, the mosaic design of psychotics, narcotics, and "normal" individuals, etc., are alleged to fall into distinct types. Those who a:re confident in this have offered certain criteria as indicative of mosaics of these various clinical groupings. The purpose of this study is to determine how successfully mosaic patterns constructed by persons with psychiatric diagnoses of schizophrenia can be differentiated from those constructed by non-schizophrenic (i.e., "normal'') individuals, and which criteria are found to be most valid in making such a distinction.

No description available.

No description available.

No description available.

Digitized by Kansas Correctional Industries