Does e-assessment always fit digital natives? A within-subject comparison between paper- and tablet-based gambling assessments in adolescents

Technological development has enabled the use of sophisticated methods for assessing multiple human behaviors. Despite the advantages of these new technologies, concerns exist regarding their equivalence with paper-based measures in epidemiological and health-related surveys. To date, literature on this topic in relation to adolescents is virtually nonexistent. This study compares respondents’ performance on the same survey using both paper- and electronic tablet-based assessment methods. A final sample of 135 adolescents (mean age 17.30 years, SD = 0.59; 56.3% males) consecutively completed two versions of the same survey on gambling behaviors and two questionnaires: The Gambling Motives Questionnaire (GMQ) and the South Oaks Gambling Screening-Revised for Adolescents (SOGS-RA). An ad-hoc questionnaire assessing participants’ satisfaction levels with each method was also used. The digital survey yielded a lifetime, past year, and past month gambling prevalence of 54.1%, 45.2%, and 27.4%, respectively. Paper-based prevalence rates were 3.7-5.2% lower (all p < .092) and there were discrepancies in gambling activities. Although the reliability of the questionnaires was high in both formats, total scores were consistently higher in the paper-based format. GMQ and SOGS-RA intraclass correlations between versions ranged from .856-.884. Unexpectedly, students preferred the paper-based survey to the e-assessment (51.5% vs. 48.5%) and also enjoyed it more (31.3% vs 26.1%). Paper- and tablet-based surveys yield different, albeit non-statistically significant, estimations of gambling behaviors even when the same participants were surveyed at one time. We recommend that consistency be routinely checked across assessment formats when adapting paper-and-pencil measures to digital formats