ABSTRACT

There is a limited body of experimental research examining the comparability of completing self-report surveys using different computerized devices. Additionally, available literature has not used complete or optimal procedures for determining device equivalence. The current study examined the comparability of surveys completed using paper-and-pencil and three popular devices: smartphone, tablet, and desktop computer. Participants consisted of 211 college students randomly assigned to conditions who completed measures of personality, social desirability, and computer self-efficacy. Results showed evidence of qualitative equivalence (internal consistency and subscale intercorrelations) across conditions. For quantitative and auxiliary equivalence, both equivalence testing and Bayesian analyses were conducted. Equivalence testing indicated quantitative (mean score) equivalence, as well as comparability for one aspect of auxiliary equivalence (missing data). Other aspects of auxiliary equivalence (completion time and comfort completing questionnaires) suggested potentially meaningful differences. Bayesian analyses typically replicated these results, with some notable exceptions regarding auxiliary equivalence.

摘要

实验研究的机构很有限，无法使用不同的计算机设备来完成自我报告调查的可比性。此外，现有文献尚未使用完整或最佳的程序来确定设备的等效性。当前的研究检查了使用纸笔和三种流行设备完成的调查的可比性：智能手机，平板电脑和台式计算机。参与者包括211名大学生，他们被随机分配到完成个性，社会期望和计算机自我效能测量的条件下。结果表明，在不同条件下，定性对等的证据（内部一致性和子量表的相互关系）。对于定量和辅助等效性，进行了等效性测试和贝叶斯分析。等效性测试表明定量（均值）等效性，以及辅助等效性（缺少数据）一方面的可比性。辅助等效性的其他方面（完成时间和完成问卷的舒适度）表明可能存在有意义的差异。贝叶斯分析通常复制这些结果，但在辅助等价方面有一些明显的例外。