Cross-National Comparison of Equivalence and Measurement Quality of Response Scales in Denmark and Taiwan

Liao, Pei-shan ^[2] ; Saris, Willem E. ^[3] ; Zavala-Rojas, Diana ^[1]
1. [1] Universitat Pompeu Fabra
  
  Universitat Pompeu Fabra
  
  Barcelona, España
2. [2] Center for Survey Research, RCHSS
3. [3] Sociometric Research Foundation
Mostrar afiliaciones +
Localización: Journal of official statistics, ISSN 0282-423X, Vol. 35, Nº. 1, 2019, págs. 117-135
Idioma: inglés
Enlaces
- Texto completo

Resumen
- The split-ballot multitrait-multimethod (SB-MTMM) approach has been used to evaluate the measurement quality of questions in survey research. It aims to reduce the response burden of the classic MTMM design, which requires repeating alternative formulations of a survey measure to the same respondent at least three times, by using combinations of two methods in multiple groups. The SB-MTMM approach has been applied to the European Social Survey (ESS) to examine the quality of questions across countries, including the differences in response design and measurement errors. Despite wide application of the SB-MTMM design in Europe, it is yet unknown whether the same quality of survey instruments can be achieved in both a different cultural context and in a logographic writing system, like the one in Taiwan. This study tests for measurement invariance and compares measurement quality in Taiwan and Denmark, by estimating the reliability and validity of different response scales using the SB-MTMM approach. By using the same questions as in the ESS, a cross-cultural comparison is made, in order to understand whether the studied response scales perform equally well in Taiwan, compared to a European country. Results show that quality estimates are comparable across countries.