The Effects of a Computer-Assisted Interview Tool on Data Quality

Although computer-assisted interview tools have much potential, little empirical evidence on the quality and quantity of data generated by these tools has been collected. In this study we compared the effects of using Virre, a computer-assisted self-interview tool, with the effects of using other data collection methods, such as written responding and face-to-face interviewing, on data quantity and quality. An intra-sample, counter-balanced, repeated-measures experiment was used with 20 students enrolled in a computer science education program. It was found that there were significantly more words and unique clauses per response in the Virre condition than in the written response condition; however, the odds of avoiding in-depth responding were much greater in the Virre condition than in the written response condition. The effect sizes observed in this study indicated that there was a greater quantity of data and a higher rate of response avoidance during Virre data collection than what would have been expected, based on previous literature, during face-to-face data collection. Although these results statistically generalize to our sample only, these results indicate that audio-visual, computer-assisted self-interviewing can yield as high, or higher, quality of data as other commonly-used data collection methods.