Comparing visual analogue and Likert-type scales

Psychometrics
Surveys
My PhD project: Do visual analogue and Likert-type scales yield comparable data? Three studies on psychometric properties and response styles.

Paper in Behavior Research Methods Interactive app

Surveys often use Likert-type scales with a few ordered categories, but visual analogue scales (VAS), often presented as sliders, are becoming common, especially on smartphones. My dissertation, Comparing Visual Analogue and Likert-Type Scales: Psychometric Properties and Response Styles (submitted, defence in February 2027), asks whether the two formats can be used interchangeably.

Three studies

  1. Psychometric properties in daily life. In an experience sampling study, participants were randomly assigned to one of the two formats. I analysed the data with multilevel and dynamic structural equation models (MSEM, DSEM). The formats showed strong measurement invariance and largely comparable reliability and validity, but Likert-type responses had more error variance and varied more over time. Published in Behavior Research Methods.
  2. Detecting response styles in VAS data. A new model for extreme, midpoint and acquiescent responding in visual analogue scale data, see BIRM-RS.
  3. Response styles across formats. When the same people answered personality items in both formats, individual differences in extreme responding generalised across formats to a considerable extent. Visual analogue scales do not eliminate response styles.

Take-away

The formats yield largely comparable results but are not simply interchangeable. The response format deserves to be treated as an explicit decision in questionnaire and study design.