The purpose of this study was to compare the achievement and test anxiety level of students taking a conventional paper-and-pencil science test comprising multiple-choice questions, and a computer-based version of the same test. The study assessed the equivalence of the computer-based and paper-and-pencil tests in terms of achievement scores and item characteristics, explored the relationship between computer anxiety and previous computer experience, and investigated the affective impact of computerized testing on the students. A 2 X 2 (mode of test administration by gender) factorial design was used. A sample of 54 male and 51 female Grade 10 students participated in the study. Subjects were blocked by gender and their scores on a previous school-based science exam. They were then randomly assigned to take either the computer-based test or the paper-and-pencil test, both versions of which were identical in length, item content and sequence. Three days before the test, all students were given the "Attitude questionnaire" which included pre-measures of test and computer anxiety. Immediately after taking the test, students in the computer-based group completed the "Survey of attitudes towards testing by computers" questionnaire which assessed their previous computer experience, their test anxiety and computer anxiety level while taking the test, and their reactions towards computer-based testing. Students in the paper-and-pencil test group answered the "Survey of attitudes towards testing" questionnaire which measured their test anxiety level while they were taking the paper-and-pencil test. The results indicate that the mean achievement score on the science test was significantly higher for the group taking the computer-based test. No significant difference in mean scores between sexes was observed; there was also no interaction effect between mode of test administration and gender. The test anxiety level was not significantly different between the groups taking the two versions of the test. A significant relationship existed between students' prior computer experience and their computer anxiety before taking the test. However, there was no significant relationship between previous computer experience and the computer anxiety evoked as a result of taking the test on the computer. Hence, the change in computer anxiety due to taking the test was not explained by computer experience. Of the students who took the computer-based test, 71.2 % said that if given a choice, they would prefer to take the test on a computer. Students indicated that they found the test easier, more convenient to answer because they did not have to write, erase mistakes or fill in bubbles on a scannable sheet, and faster to take when compared to a paper-and-pencil test. Negative responses to the computer-based test included the difficulty involved in reviewing and changing answers, having to type and use a keyboard, fear of the computer making mistakes, and a feeling of uneasiness because the medium of test presentation was unconventional. Students taking the computer-based test were more willing to guess on an item, and tended to avoid the option "I don't know." It is concluded that the computer-based and the paper-and-pencil tests were not equivalent in terms of achievement scores. Modifications in the way test items are presented on a computer-based test may change the strategies with which students approach the items. Extraneous variables incidental to the computer administration such as the inclination to guess on a question, the ease of getting cues from other questions, differences in test-taking flexibility, familiarity with computers, and attitudes towards computers may change the test-taking behaviour to the extent that a student's performance on a computer-based test and paper-and-pencil test may not be the same. Also, if the tasks involved in taking a test on a computer are kept simple enough, prior computer experience has little impact on the anxiety evoked in a student taking the test, and even test-takers with minimal computer experience will not be disadvantaged by having to use an unfamiliar machine. / Education, Faculty of / Curriculum and Pedagogy (EDCP), Department of / Graduate

The purpose of this research was to determine if personality type and system response time have any effect on state anxiety and user response time. The sample for this study consisted of senior and graduate level college students who possessed basic know 1 edge of a text editor. Each test subject was administered the Jenkins Activity Survey to determine scores for Type A versus Type B, speed and impatience, involvement, and competitiveness. The test subjects were randomly assigned to one of three treatment groups (good, variable, and poor system response time). They were required to edit a text file which contained multiple errors. The test subjects were provided hard copies of the file with errors (errors highlighted) and the file as should appear without the errors. The test situation for each test subject was identical, except for changes in system response time. The A-state scale of the State-Trait Anxiety Inventory (STAI) was administered to the test subjects immediately prior to the edit task in order to determine pre-task state anxiety levels. The A-state scale of the STAI was again administered immediately after the edit task in order to determine post-task state anxiety levels. Analysis of variance, analysis of covariance, regression, and two sample t-tests were used to analyze the data collected. All hypotheses were tested at the alpha .05 level. The most significant finding of this study was the positive relationship between state anxiety and system response time. It was originally predicted that the Type A personality would experience a greater increase in state anxiety than the Type B personality. However, that was not found to be true. Both Type A and Type B individuals experience an increase in state anxiety during periods of poor or variable system response time. This study also confirms prior research regarding user and system response time. There is a significant positive relationship between user response time and system response time. Personality type, specifically the Type A personality, contributes toward this relationship.

The purpose of this investigation was to determine the relationship between computer experience and selected demographic variables to knowledge about computers; knowledge of computers, experience with computers and selected demographic variables and attitudes towards computers to perceived computer helplessness. The study utilized two samples. Questionnaires were hand delivered to 68 out of 81 students enrolled in recreation classes at North Texas State University and questionnaires were sent to 130 Dallas Park and Recreation professionals via their inter-department mail (115 surveys returned). The data analysis indicated that the knowledge, attitude, and helplessness scales were highly reliable. Regression analysis indicated that knowledge, experience, attitudes, sex, age, and occupation were significant predictors of helplessness.

A group of 125 female college students were given Spielberger's State Anxiety Inventory and Bem's Sex Role Inventory. Those females with minimal typing skills and minimal computer experience interacted with a computer ten minutes and then given a second State Anxiety Inventory. Pretest and posttest data were obtained from forty-eight subjects selected on the basis of Bem's Sex Role Inventory to test the hypotheses: females that score high on the Androgynous scale of the Bem Sex Role Inventory (BSRI) will experience relatively low levels of anxiety while interacting with computer, females that score high on the Feminine and Undifferentiated scales of BSRI will experience relatively high levels of anxiety while interacting with a computer, and females that score high Masculine on the BSRI will experience relatively intermediate levels of anxiety. Androgynous females will produce high performance scores on the computer typing task while Feminine or Undifferentiated subjects will have the lowest performance scores. No significant differences among the four groups were found F(3,44)=2.14, p > .05 for anxiety change. No significant differences were found for performance scores among the four groups F(3,44)=.773, p > .05. Implications for alternative predictors of anxiety are discussed.

Lee Wing-Sze Wincy. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2001. / Includes bibliographical references (leaves 39-47). / Abstracts in English and Chinese. / ABSTRACT --- p.2 / ACKNOWLEDEGEMENT --- p.4 / TABLE OF CONTENTS --- p.5 / INTRODUCTION OF RESEARCH TOPIC --- p.7 / THEORETICAL OVERVIEW --- p.9 / To be or not to be: Cognitive appraisal of stressful event --- p.9 / Psychological Vulnerability: Attitude towards using computer and computer anxiety --- p.13 / Burnout as the manifestation of occupational strain --- p.15 / Social Support --- p.17 / "The consequences of ""burning out"": Intention to Leave" --- p.19 / PRESENT INVESTIGATION --- p.19 / Confirmatory Analysis --- p.19 / Model specification and model testing --- p.20 / METHOD --- p.22 / Sample --- p.22 / Measurement of theoretical constructs --- p.23 / Pilot Study --- p.26 / RESULTS --- p.27 / Item level confirmatory analysis --- p.27 / Testing of measurement model --- p.30 / Evaluating the structural model --- p.32 / DISCUSSION --- p.34 / Implication of findings --- p.34 / Limitations and future studies --- p.37 / REFERENCES --- p.40 / APPENDIX A --- p.49 / APPENDIX B --- p.50 / APPENDIX C --- p.51 / TABLE 1 TO TABLE4 --- p.58 / FIGURE CAPTION --- p.61 / FIGURE1 --- p.62 / FIGURE2 --- p.63