Do visual analogue scales perform better than Likert-type scales?

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14230%2F24%3A00139495" target="_blank" >RIV/00216224:14230/24:00139495 - isvavai.cz</a>

Result on the web

<a href="https://www.psychometricsociety.org/imps-2024" target="_blank" >https://www.psychometricsociety.org/imps-2024</a>

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Do visual analogue scales perform better than Likert-type scales?

Original language description

The Likert Scale (LS) and Visual Analogue Scale (VAS) are widely used response formats in social sciences research. Some scholars propose that VAS, which can be seen as a Likert scale with an infinite number of response options, should outperform LS by allowing for more nuanced responses, increasing systematic variance, and better approximating the continuous nature of data. However, existing research comparing the psychometric properties of LS and VAS yields only inconclusive results regarding descriptive statistics, internal consistency, and respondent preferences. Furthermore, evidence on other psychometric properties remains scarce. This study addresses these gaps by examining the differences between LS and VAS in terms of internal consistency, test-retest reliability, criterion validity, and response times utilizing an online within-subject experimental design. With a sample of 980 participants, we compared the five-point Likert Scale and the Visual Analogue Scale using two questionnaires: the Height Inventory (HI) and the Autonomy Subscale (AUT). We assessed both short-term dependability (test-retest reliability within a single session) and long-term stability (test-retest reliability over one week). Our findings indicate negligible or no differences in dependability, stability, or internal consistency between the LS and VAS. Criterion validity, as measured by the correlation of HI with self-reported height, was also equivalent for both scales. The only notable difference was in response time, with VAS taking approximately 16% longer to complete. We argue that VAS offers no significant advantages over LS, and due to the greater time demands of VAS, we conclude that using the LS is more worthwhile.

Czech name

—

Czech description

—

Type

O - Miscellaneous

CEP classification

—

OECD FORD branch

50101 - Psychology (including human - machine relations)

Project

<a href="/en/project/GA23-06924S" target="_blank" >GA23-06924S: Response-scale format effects on the psychometric parameters of items</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2024

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů