Slider versus Likert scales: Psychometric properties in ambulatory assessment

Slider scales, a type of visual analogue scale, are commonly used as a response format in smartphone-based ambulatory assessment studies. This may be due to several advantages of slider scales, including the use of a metric scale compared to an ordinal Likert scale and the intuitive use of slider scales on touchscreens. However, research on the comparability of Likert and slider scales remains scarce, especially in the context of ambulatory assessment (i.e., intensive longitudinal data). To address this gap, we conducted a 3-week ambulatory assessment study with four measurement occasions per day, in which we experimentally manipulated the response format (Likert vs. slider scales) between groups. Our final sample consisted of 21,730 measurement occasions nested in 406 participants. We tested for measurement invariance across response-format groups and compared their psychometric properties, including reliability, within-person variability, and validity. Results indicated measurement invariance across groups, with equal factor loadings at the within-person level and equal factor loadings, intercepts, and indicator-specific residual variances at the between-person level. In addition, we found no significant differences in reliability or validity and only minor differences in within-person variability. We discuss the implications of these findings for the design of ambulatory assessment studies and offer recommendations for future research.