| The purpose of this study was to investigate the impact of Internet-based test administration on the responses to a self-estimate of ability. The Ability Explorer (Harrington & Harrington, 1996) is the instrument used to conduct this investigation. Previous literature over the past 30 years has reported a number of studies comparing computer-based to traditional paper-pencil administration of various tests and inventories. However, few studies were uncovered which examined Internet-based administrations.; The sample consisted of 73 volunteers drawn from three sections of Educational Psychology 102, Career Development, at Texas A&M University between Spring 1999 and Spring 2000. The majority of the participants were freshmen and sophomores who had not yet declared a major. A four-group test-retest design (Paper-Paper, Paper-Internet, Internet-Paper, Internet-Internet) was used to compare the two versions of the instrument. Additional analyses were performed to determine if there were relationships or interactions between the scales and the moderator variables of Gender, SAT, GPA, and SES. Participants' satisfaction with either or both versions of the Ability Explorer was also ascertained via a survey administered at the completion of the study.; By examining both correlational and mean difference data, the two versions of the instrument were found to be generally equivalent. Some scales in some analyses did yield mean differences, but they could be chance occurrences given the many comparisons examined. On the average, respondents rated themselves higher on some scales in the second administration, but this phenomenon was not related to mode of administration. There were relationships found between Gender and some ability scales. Males rated themselves higher on scales that are stereotypically male-oriented (Manual, Numerical/Mathematical, Scientific, and Technical/Mechanical), while females rated themselves higher on the Social scale. Also, Gender was found to interact with mode (paper-pencil versus Internet) on several scales, but no easily identifiable patterns emerged.; Responses to the satisfaction surveys were quite positive for both versions of the Ability Explorer. However, when asked directly which version the respondent preferred, the overwhelming majority selected the Internet version.; A discussion of the findings, as they relate to the literature, as well as recommendations for further research and practice are presented. |
