Ethnic status is one of the most protected demographic groups in terms of test bias and discrimination in personnel selection, as such bias breaches many laws, morals, and ethical procedures. To date there has only been two published studies that have used New Zealand relevant ethnic groups when analysing whether personality measures used in pre-employment settings exhibit different mean scores. The present thesis performed a systematic evaluation on the impact of ethnicity on personality traits. The study examined the impact of ethnicity on NZ Army Officer applicant personality assessment scores measured by the EPQ-R and GPP-I. Four ethnic groups (NZ Europeans, Mâori, Pasifika, and Asian) were analysed for mean trait score differences among ethnic groups, the stability of these differences across different personality inventories and models, the variance of personality traits, the impact of ethnicity on age and gender relationships with personality traits, and how these differences were related to employment selection outcomes. The analyses revealed that ethnicity did have some impact on mean personality assessment scores used in the study. However, these were mostly small differences among ethnic groups. In addition, most of these differences found on the EPQ-R and the GPP-I were not consistent across inventories and models. There were no significant variance differences found on personality traits among ethnic groups. While initial analysis suggested that ethnicity did not influence the relationship between age and gender on personality assessment, further examination suggested that the relationship between gender and personality was impacted on by ethnicity. There were reassuring results found for New Zealand psychologists and HR specialists, as only two of the twelve traits analysed showed moderate differences on traits that were related to selection outcomes. However, for the NZ Army OSB selection process the findings in the present study indicate that the Lie scale on the EPQ-R and the Vigor trait on the GPP-I may need to be interpreted with caution for Asian and Pasifika groups. These results are discussed in terms of implications for personality theory, measurement, and the direction of future research.

Recent research has challenged the way in which personality and attitude constructs are measured. Alternatives have been offered as to how non-cognitive responses are modeled, the mode of delivery used when administrating such scales, and the impact of technology in measuring personality. Thus, the major purpose of the studies in this thesis concerns two interrelated issues of personality research, namely the way personality responses are best modeled, and the most optimal mode by which personality items are presented and associated modal issues. Three studies are presented. First, recent developments using an ideal point approach to scale construction are outlined, and an empirical study compares modeling personality items based on an ideal point approach (generalized graded unfolding model; GGUM) and a dominance approach (graded response model: GRM). Second, an extensive review of literature pertaining to the mode effect when transferring paper-and-pencil measures to screen was conducted, in addition to a review of the various types of computerized and innovative items and their associated psychometric information. Finally, nine innovative items were developed using various multimedia features (e.g., video, graphics, and audio) to ascertain the advantages of these methods to present items constructed to elicit response behavior underlying ideal point approaches, namely, typical response behavior. It was found that the dominance IRT model continued to produce superior model-data fit for most items, more attention needs to be placed on developing principles for constructing ideal point type items, the web-based version supplied 20% more construct information than the paper version, and innovative items seem to provide more data-model fit for students with lower personality attributes. While the innovative items may require more initial outlay in terms of time and development costs, they have the capacity to provide more information regarding test-takers’ personality levels, potentially using fewer items.

Recent research has challenged the way in which personality and attitude constructs are measured. Alternatives have been offered as to how non-cognitive responses are modeled, the mode of delivery used when administrating such scales, and the impact of technology in measuring personality. Thus, the major purpose of the studies in this thesis concerns two interrelated issues of personality research, namely the way personality responses are best modeled, and the most optimal mode by which personality items are presented and associated modal issues. Three studies are presented. First, recent developments using an ideal point approach to scale construction are outlined, and an empirical study compares modeling personality items based on an ideal point approach (generalized graded unfolding model; GGUM) and a dominance approach (graded response model: GRM). Second, an extensive review of literature pertaining to the mode effect when transferring paper-and-pencil measures to screen was conducted, in addition to a review of the various types of computerized and innovative items and their associated psychometric information. Finally, nine innovative items were developed using various multimedia features (e.g., video, graphics, and audio) to ascertain the advantages of these methods to present items constructed to elicit response behavior underlying ideal point approaches, namely, typical response behavior. It was found that the dominance IRT model continued to produce superior model-data fit for most items, more attention needs to be placed on developing principles for constructing ideal point type items, the web-based version supplied 20% more construct information than the paper version, and innovative items seem to provide more data-model fit for students with lower personality attributes. While the innovative items may require more initial outlay in terms of time and development costs, they have the capacity to provide more information regarding test-takers’ personality levels, potentially using fewer items.

Recent research has challenged the way in which personality and attitude constructs are measured. Alternatives have been offered as to how non-cognitive responses are modeled, the mode of delivery used when administrating such scales, and the impact of technology in measuring personality. Thus, the major purpose of the studies in this thesis concerns two interrelated issues of personality research, namely the way personality responses are best modeled, and the most optimal mode by which personality items are presented and associated modal issues. Three studies are presented. First, recent developments using an ideal point approach to scale construction are outlined, and an empirical study compares modeling personality items based on an ideal point approach (generalized graded unfolding model; GGUM) and a dominance approach (graded response model: GRM). Second, an extensive review of literature pertaining to the mode effect when transferring paper-and-pencil measures to screen was conducted, in addition to a review of the various types of computerized and innovative items and their associated psychometric information. Finally, nine innovative items were developed using various multimedia features (e.g., video, graphics, and audio) to ascertain the advantages of these methods to present items constructed to elicit response behavior underlying ideal point approaches, namely, typical response behavior. It was found that the dominance IRT model continued to produce superior model-data fit for most items, more attention needs to be placed on developing principles for constructing ideal point type items, the web-based version supplied 20% more construct information than the paper version, and innovative items seem to provide more data-model fit for students with lower personality attributes. While the innovative items may require more initial outlay in terms of time and development costs, they have the capacity to provide more information regarding test-takers’ personality levels, potentially using fewer items.

Recent research has challenged the way in which personality and attitude constructs are measured. Alternatives have been offered as to how non-cognitive responses are modeled, the mode of delivery used when administrating such scales, and the impact of technology in measuring personality. Thus, the major purpose of the studies in this thesis concerns two interrelated issues of personality research, namely the way personality responses are best modeled, and the most optimal mode by which personality items are presented and associated modal issues. Three studies are presented. First, recent developments using an ideal point approach to scale construction are outlined, and an empirical study compares modeling personality items based on an ideal point approach (generalized graded unfolding model; GGUM) and a dominance approach (graded response model: GRM). Second, an extensive review of literature pertaining to the mode effect when transferring paper-and-pencil measures to screen was conducted, in addition to a review of the various types of computerized and innovative items and their associated psychometric information. Finally, nine innovative items were developed using various multimedia features (e.g., video, graphics, and audio) to ascertain the advantages of these methods to present items constructed to elicit response behavior underlying ideal point approaches, namely, typical response behavior. It was found that the dominance IRT model continued to produce superior model-data fit for most items, more attention needs to be placed on developing principles for constructing ideal point type items, the web-based version supplied 20% more construct information than the paper version, and innovative items seem to provide more data-model fit for students with lower personality attributes. While the innovative items may require more initial outlay in terms of time and development costs, they have the capacity to provide more information regarding test-takers’ personality levels, potentially using fewer items.

<p></p><p>A large body of literature suggests that cognitive abilities are important determinants for training and job performance, including flight performance. The associations between measures of ability tests and job performance have been the focus of many empirical studies, resulting in an overall conclusion that general mental ability, <i>g</i>, is the main source of prediction, while other narrower abilities have limited power for predicting job performance. Despite the attention given to cognitive ability-flight performance relationships, their associations have not been fully understood at the broad construct level, and most extant literature focused on the relations at the observed scores level. Thus, the present dissertation study was designed to contribute to the progression of this understanding by examining the relations between cognitive abilities and flight training performance, using data from four U.S. Air Force (USAF) pilot samples. For comparison, one navigator and one air battle manager sample were also analyzed. The data were obtained from correlation matrices of prior investigations and analyzed via structural equation modeling (SEM) procedures. </p> <p> Four studies are reported in the thesis: (1) preliminary study, (2) primary validation study, (3) cross-validation study, and (4) cross-occupation validation study. The preliminary study assessed the test battery used in the subsequent predictive studies. The primary validation study introduced a bifactor predictive SEM model for testing the influence of cognitive abilities in predicting pilot performance. The cross-validation study assessed the consistency of the predictive model suggested in the primary validation study, using three additional pilots’ samples. The cross-occupation validation study compared the predictive model using data from three aviation-related occupations (flying, navigation, air battle management). Ability factors were extracted from scores of pilot applicants on the Air Force Officer Qualifying Test (AFOQT), the USAF officers’ primary selection test battery, whereas the flight performance scores were obtained from pilot records during the flight training program.</p> <p> In addition to the <i>g</i> factor, <i>verbal ability, quantitative ability, spatial ability, perceptual speed ability, and aviation-related acquired knowledge </i>are the six latent cognitive ability factors investigated in the reported studies. Pilot performance measures were modeled either as observed or latent variables covering ratings of academic and hands-on flying performance in different phases of the training program. The studies of this thesis established that (1) general ability contributes substantially to the prediction models; however, it is not the only important predictor, (2) aviation-related acquired knowledge is the most robust predictor of pilot performance among the abilities examined, with a role even exceeding that of <i>g</i>, (3) perceptual speed predicted pilot performance uniquely in several occasions, while verbal, spatial, and quantitative abilities demonstrated trivial incremental validity for hands-on pilot performance beyond that provided by the <i>g</i> measure, and (4) the relative importance of cognitive abilities tends to vary across aviation occupations.</p><br><p></p>