Some proponents of computer assisted psychological assessment have argued that computer administration of tests may reduce social desirability response bias and, consequently, increase the validity of test results. Although computer administration has been shown to decrease this response bias in nonclinical subjects, this effect has not been found in clinical populations. This author hypothesized that clinical subjects may experience repercussions because of their test responses (e.g., changes in treatment) and this could have a greater effect in determining responses than social desirability. To test this hypothesis, 76 adult psychiatric outpatients were assessed either by computer or clinician on measures of social desirability and psychopathology. One-half were tested under typical clinical conditions in which their therapists received feedback on their results and one-half were tested under standard research conditions assuring confidentiality. ANOVA tests revealed a significant difference between subjects tested under clinical (i.e., possible repercussions) and research (i.e., no repercussions) conditions on the measure of psychopathology and one of the two social desirability measures. Subjects presented themselves as having more traits which are considered psychologically healthy and fewer traits which are viewed as antisocial when their therapists would know their results. The type of assessor, either computer or clinician, had no significant effect on subjects' responses. Although both types of assessor were rated favorably, subjects rated the clinical interview as significantly more enjoyable than the computer administered version of the interview. Implications of these findings for future research and clinical practice were discussed. / Source: Dissertation Abstracts International, Volume: 54-03, Section: B, page: 1676. / Major Professor: Mark H. Licht. / Thesis (Ph.D.)--The Florida State University, 1993.

This dissertation focuses on the importance that status has on society and the difficulties in adequately addressing its effect on consumption. The motivation to consume for status has not been adequately conceptualized or operationalized in the social science literature. The author suggests a new construct to address this issue: status consumption. Status consumption is defined as the motivational process by which individuals strive to improve their social standing through the conspicuous consumption of consumer products that confer and symbolize status both to the individual and to surrounding significant others. / To illustrate this, the author presents a model of the proposed six antecedents, three dimensions, and six consequences of status consumption presented in this dissertation. Based on the three dimensions, a three factor Status Consumption Scale, SCS, is created. Hypotheses are presented along with a proposed methodology to test the reliability and validity of the SCS in a series of five studies utilizing both students and adults. The results of the five studies illustrates that the fourteen item SCS is a reliable and valid measure with good factor structure having three unidimensional subscales (Sociability, An Interest in Consuming for Status, and Buying for Nonfunctional Reasons). / Thus, this individual difference variable of status consumption is demonstrated in this dissertation to be an explanatory variable that may explain and predict some aspects of consumer behavior. / Source: Dissertation Abstracts International, Volume: 54-02, Section: A, page: 0600. / Major Professor: Ronald E. Goldsmith. / Thesis (Ph.D.)--The Florida State University, 1993.

The purpose of this study was to conduct a multimethod study of the dynamics of Learned Helplessness (LH) in an organization. Models of LH which exist in the literature were reviewed, and comparisons and contrasts among the models were made, leading to suggestions for a number of variables from these models which were examined. Interview and questionnaire data were gathered to examine a number of issues including the validity of LH as a construct in organizations, the occurrence, dimensionality and consistency of attributions made by organizational members, and the convergence of various measures of LH, with special attention given to the validity of the Organizational Attributional Style Questionnaire (OASQ). To analyze the data, content analysis, correlations, profile analysis and multivariate multiple regression were used. / Results indicated that LH is a construct deserving of further attention by organizational researchers. Specifically, results showed that there are differences between helpless and empowered respondents, that attributions are consistent in similar situations, that the results of using various measures to examine LH are convergent, and that the OASQ is a valid measure of LH in organizations. / Source: Dissertation Abstracts International, Volume: 53-10, Section: A, page: 3590. / Major Professor: Mark J. Martinko. / Thesis (Ph.D.)--The Florida State University, 1992.

This study explored the usefulness of the warp factor, a measure of the extent to which a multidimensional scaling solution is non-Euclidean in Galileo theory, as a measure of cognitive dissonance. A direct measure of cognitive dissonance has been lacking from dissonance theory, provoking criticism in the literature. Meanwhile, proponents of Galileo theory have claimed that dissonance is among the theoretic explanations of warp. By testing the usefulness of warp as a measure of dissonance, this study contributes to both these theoretical bodies of literature within the communication discipline. / The methodology consisted of a randomized, post test only design where dissonance was manipulated by asking each subject in the experimental conditions to write a counter-attitudinal essay. Subjects were students in large-enrollment undergraduate communication classes at a major southeastern university. Because many subjects refused to cooperate with the manipulation, data were collected twice and the results combined for analysis. A "bootstrap" technique for computing error around the warp factor was used to allow hypothesis testing. / The hypothesis relating levels of dissonance to size of the warp factor was not supported, as no significant warp was measured for any group. The results are thought to be due to subjects' success in reducing the induced dissonance before it could be measured using the Galileo technique. Tentative conclusions include the possibility that the warp factor is not useful as a measure of the kind of cognitive dissonance which can ordinarily be manipulated in an experiment because in practice this type of dissonance can be reduced by subject self-persuasion or other means before it can be measured using warp. / Source: Dissertation Abstracts International, Volume: 49-08, Section: A, page: 2008. / Major Professor: C. Edward Wotring. / Thesis (Ph.D.)--The Florida State University, 1988.

The purpose of this study was to develop an instrument to evaluate organizational attributional style (OAS). OAS is defined as an individual's characteristic attributional tendencies concerning work related issues. Based on a review of the existing attributional literature, a scale was developed to assess OAS. The scale was given to two groups of subjects along with measures of other constructs of interest. Both an exploratory and confirmatory factor analysis were performed and relationships between OAS and other variables (e.g. performance, turnover) were assessed. Reported reliabilities of the scale was acceptable and some evidence for the construct validity of the scale was found. The results of this study should serve as a guide to future attributional research in organizational behavior. / Source: Dissertation Abstracts International, Volume: 52-10, Section: A, page: 3656. / Major Professor: Mark J. Martinko. / Thesis (Ph.D.)--The Florida State University, 1991.

The present study compared two groups of community-bound persons, over 55 years old, who had sustained deterioration of cognitive and behavioral skills for over 6 months. These changes had been medically diagnosed as either Senile Dementia of the Alzheimer's Type (SDAT) or Multi-Infarct Dementia (MID). There were 15 subjects in each group. All participants were relatively free of any known psychiatric or medical disorder, outside the presumptive dementia condition. / A profile of these groups' performance on physiological, behavioral, and psychometric measures was obtained. Physiological measures, EEG(P300) and CT Scan, confirmed the presence of neurological changes in 28 subjects; abnormal patterns in the EEG(P300) were also positively related to the subjects' age. Functional deficits, rated by the caregivers, increased proportional to the length of time since onset of the illness. Descriptively, within the first three years of the illness, the SDAT group preserved self-care skills longer than the MID group, while the latter maintained their communication skills at the pre-morbid functional levels. Neuropsychological data were obtained in six areas, including orientation, memory, abstraction, judgment, visual-motor coordination, and language. Significant differences were noted in recent memory, concentration, comparative judgment, and processing of specified words. / A linear discriminant analysis identified the psychometric and behavioral scales most sensitive to the differentiation of SDAT and MID conditions. The resulting brief but thorough screening scale showed a predictive and reliability coefficient above chance. Data also indicated a physician's bias in rating the level of cognitive functioning of the SDAT group. The SDAT subjects, who were older, were rated as more cognitively impaired than the MID subjects, who were younger. There were no significant differences between these groups in the length of time since onset of dementia. / Source: Dissertation Abstracts International, Volume: 49-03, Section: B, page: 0910. / Major Professor: Lloyd F. Elfner. / Thesis (Ph.D.)--The Florida State University, 1987.

Poorly written police reports interfere with the efficient functioning of the criminal justice system. Accounts from several police jurisdictions have indicated that many officers lack the necessary writing skills to complete reports which require documenting chronological events in an incident, reporting witness statements, and describing evidence. This study attempted to identify the important dimensions of police report writing. On this foundation, a job-related video test was developed and its psychometric properties investigated. The test, the New Orleans Police Department Written Communication Test (NWCT), was designed to assess the report writing skills of New Orleans Police Department Recruits and Officers. Two versions of the test were developed, each having direct and indirect subtests. Subjects were 110 police officers who had been on the job one to two years. Internal consistency, alternate forms reliability, interscorer reliability, intraclass reliability, and test-retest reliability were estimated. Reliability was also estimated within a generalizability framework with persons and raters as facets. Both forms of the NWCT were administered to the officers as well as Hammill and Larsen's (1988) Test of Written Language-2 (TOWL-2). The TOWL-2 was administered to the officers to estimate the convergent validity of the NWCT. Convergent validity was also assessed by determining the relationship between NWCT scores and other measures of writing skills. Subjects' police reports were rated by Assistant District Attorneys and police personnel whose ratings served as criteria. Internal consistency (Cronbach's alpha) estimates differed by subtest, and the overall test lacked internal consistency (both Forms A and B). Alternate forms reliability coefficient was extremely low (.22) for the alternate direct subtests. The test-retest coefficient was.73 ($p < .001$) for the indirect portions of the test which were identical for both forms of the test. Analysis of Variance (ANOVA) was used to estimate the variance components for the effects of raters, persons, and the interaction between raters and persons. A large amount of variance was attributed to differences in how raters scored subjects for the direct subtests. With the indirect subtests, however, much of the variance was attributed to differences in the test performance of the subjects. Convergent validity evidence was weak, although the NWCT did correlate significantly with some of the variables of interest. Criterion-related validity evidence, using ratings of work samples as criteria, was insufficient / acase@tulane.edu

The effect of skew on the standardized item alpha, an estimate of internal consistency closely related to Cronbach's coefficient alpha, was examined through Monte Carlo techniques. First, situations in which skew may exist were described, then attempts to address problems created by skew were discussed. Next, the logical relations among split-half reliability, Cronbach's coefficient alpha, and the standardized item alpha were examined and shown to be equivalent when all item variances are equal. Finally, results from Monte Carlo simulations were presented which compared alpha computed from standard normal variables to alpha computed from the same samples but where skew was induced by the lognormal transformation and finally to alpha computed from the ranks of the lognormal values. Both the extent and direction of skew were systematically varied as was the size of the population interitem correlation--rho. Since factorial complexity and the average interitem correlation have been shown to be related to alpha, the effect of skew on both of these was also examined. Results indicated that skew increases factorial complexity, decreases the average interitem correlations, and decreases alpha. These results were especially pronounced when rho was small. Transforming the values with a natural log transformation, which produces results equivalent to bivariate normal values, and ranking the lognormal values eliminated the skew as well as the liabilities associated with skew. However, ranking resulted in a slight decrease in the standardized item alpha compared to standard normal values / acase@tulane.edu

This study investigated (1) the tenability of the assumptions underlying human resource utility analysis models, (2) methods of operationalizing these models, and (3) manager's perceptions of the accuracy of results from these models. The investigation was conducted using several behavioral and sales performance measures from 434 sales-representatives who attended a product information and sales strategy training program. Low positive relationships were observed between behavioral performance and the value of performance, demonstrating that it is tenable to assume a linear (but not a perfect) relationship between behavioral performance and the value of performance. Such linear relationships, however, were not uniformly observed, which suggests that the tenability of this assumption may be moderated by the latent behavioral and economic domains as well as the operationalizations of the utilized measures. Alternative methods of operationalizing the utility model's components, including the effect size (i.e., d) as well as alternative methods of calculating and summarizing variability in the value of employee performance (i.e., $\sigma\sb{y}$ (sometimes referred to as SDy), A$\nu y,$ and A$\sigma\sb{\rm R}),$ yielded results with different levels of accuracy: some components were quite accurate (overestimating the parameter by 2.3 percent) and others were quite inaccurate (overestimating the parameter by 107.2 percent). These levels of accuracy were perceived to be significantly different by human resource and training department managers in a Fortune 500 company. These managers' ratings, on average, indicated that some methods of estimating utility components are accurate enough for decision making purposes. Implications for utility analysis research, performance measurement research, and the use of utility models for applied human resource decision making are discussed / acase@tulane.edu

Trimming and Winsorization are methods for dealing with the problem of outliers. This study investigated the robustness of these two procedures when there was skew in the underlying distributions. Sample data were first generated from a population with a normal distribution, and were then transformed into two skewed distributions, the lognormal and the power distributions. Trimming and Winsorization were performed on both the normal samples and transformed samples. Both Type I and Type II errors of two sample t tests and confidence intervals were investigated for evidence of robustness and power. The study found that trimmed and Winsorized means, as defined by Tukey and McLaughlin (1963) and Dixon and Tukey (1968), were not very efficient estimates of population means when the underlying distributions were skewed. While student's t test was too conservative for skewed distributions, moderately Winsorized t, as defined by Fung and Rahman (1980), was found to bring the Type I error closer to their stated levels. Furthermore, Winsorized t was found more powerful than student's t when the underlying distributions were skewed, especially when the skew was heavy, and it performed better than trimming in most cases in this study. The study suggested that these two procedures should only be used with great caution, and Winsorizing only one data point from each side of a sample was recommended with heavily skewed distributions / acase@tulane.edu