DELINEATING THE QUALITY OF INTERNET DATA: IMPLICATIONS FOR PERSONNEL ASSESSMENT
Yip, T. H., Hui, C. H., Cheng, H. C. and Yiu, V., The University of Hong Kong, Hong Kong
The Internet has recently become popular as a source of research data. However, procedures for evaluating the quality of such data have not been addressed.In the present study, data on a personality adjective checklist were collected over the Internet.Participants were recruited through the local broadcasting network and links from other sites. Demographic information such as the participant's present employment, previous employment, country of origin, gender, and percentage of customer-service related activities were also collected. We compared data collected from people who responded to the instrument more than once (those who were "experimentally-curious") against data collected from those who submitted responses once only (the "naive participants").We also examined if data from participants who had provided valid email addresses would differ from those who had supplied invalid email addresses.Test-retest reliability was examined with data from "experimentally-curious" participants. Present findings suggest that data collected over the Internet should be screened for contamination before further analyses. Suggestions are made for evaluating data collected via this media, and how such data collection method can be applied to personnel assessment.