Comparability of the BWVr across 9 nations

H. van Herk, Julie Anne Lee

Research output: Contribution to ConferencePaperProfessional

Abstract

A recent development in measuring Schwartz values is best - worst scaling (BWS; Lee et al., 2017). In this approach, respondents assess small sets in which several items are shown at the same time and the task is to choose the best and the least important value in their lives. After answering all sets the researcher can deduct the respondent’s value profile. Main questions are (1) to what extent are these BWS measured value profiles comparable to value profiles using a traditional rating scale to measure the value items and (2) to what extent is the best-worst measure comparable across nations. To answer the first question we use student samples that answered the items in the Best–Worst Refined Values scale (BWVr) using both Best-Worst scaling and rating scales. The results show that the average intra-individual correlation between BWS and rating is about .50. Assessment of predictive validity and stability indicate that BWS and rating are comparable, but not similar. To answer the question on comparability of BWVr across nations, we use large data sets from the adult population in 9 nations in Europe, Asia and the Americas. In these nations, we assess response styles in best-worst measured BWVr and link the resulting Schwartz values profiles to data on attitudes and behaviour. We find that comparability of response tendencies and predictive validity is present in subsets of the 9 nations. Implications for cross-cultural research are discussed.
Original languageEnglish
Publication statusPublished - Jul 2018

Fingerprint

Values
rating scale
scaling
rating
student

Keywords

  • Schwartz values
  • Best-Worst Scaling
  • Response bias

Cite this

@conference{a91258517f6f48f181b8f76e9a18307c,
title = "Comparability of the BWVr across 9 nations",
abstract = "A recent development in measuring Schwartz values is best - worst scaling (BWS; Lee et al., 2017). In this approach, respondents assess small sets in which several items are shown at the same time and the task is to choose the best and the least important value in their lives. After answering all sets the researcher can deduct the respondent’s value profile. Main questions are (1) to what extent are these BWS measured value profiles comparable to value profiles using a traditional rating scale to measure the value items and (2) to what extent is the best-worst measure comparable across nations. To answer the first question we use student samples that answered the items in the Best–Worst Refined Values scale (BWVr) using both Best-Worst scaling and rating scales. The results show that the average intra-individual correlation between BWS and rating is about .50. Assessment of predictive validity and stability indicate that BWS and rating are comparable, but not similar. To answer the question on comparability of BWVr across nations, we use large data sets from the adult population in 9 nations in Europe, Asia and the Americas. In these nations, we assess response styles in best-worst measured BWVr and link the resulting Schwartz values profiles to data on attitudes and behaviour. We find that comparability of response tendencies and predictive validity is present in subsets of the 9 nations. Implications for cross-cultural research are discussed.",
keywords = "Schwartz values, Best-Worst Scaling, Response bias",
author = "{van Herk}, H. and Lee, {Julie Anne}",
year = "2018",
month = "7",
language = "English",

}

Comparability of the BWVr across 9 nations. / van Herk, H.; Lee, Julie Anne.

2018.

Research output: Contribution to ConferencePaperProfessional

TY - CONF

T1 - Comparability of the BWVr across 9 nations

AU - van Herk, H.

AU - Lee, Julie Anne

PY - 2018/7

Y1 - 2018/7

N2 - A recent development in measuring Schwartz values is best - worst scaling (BWS; Lee et al., 2017). In this approach, respondents assess small sets in which several items are shown at the same time and the task is to choose the best and the least important value in their lives. After answering all sets the researcher can deduct the respondent’s value profile. Main questions are (1) to what extent are these BWS measured value profiles comparable to value profiles using a traditional rating scale to measure the value items and (2) to what extent is the best-worst measure comparable across nations. To answer the first question we use student samples that answered the items in the Best–Worst Refined Values scale (BWVr) using both Best-Worst scaling and rating scales. The results show that the average intra-individual correlation between BWS and rating is about .50. Assessment of predictive validity and stability indicate that BWS and rating are comparable, but not similar. To answer the question on comparability of BWVr across nations, we use large data sets from the adult population in 9 nations in Europe, Asia and the Americas. In these nations, we assess response styles in best-worst measured BWVr and link the resulting Schwartz values profiles to data on attitudes and behaviour. We find that comparability of response tendencies and predictive validity is present in subsets of the 9 nations. Implications for cross-cultural research are discussed.

AB - A recent development in measuring Schwartz values is best - worst scaling (BWS; Lee et al., 2017). In this approach, respondents assess small sets in which several items are shown at the same time and the task is to choose the best and the least important value in their lives. After answering all sets the researcher can deduct the respondent’s value profile. Main questions are (1) to what extent are these BWS measured value profiles comparable to value profiles using a traditional rating scale to measure the value items and (2) to what extent is the best-worst measure comparable across nations. To answer the first question we use student samples that answered the items in the Best–Worst Refined Values scale (BWVr) using both Best-Worst scaling and rating scales. The results show that the average intra-individual correlation between BWS and rating is about .50. Assessment of predictive validity and stability indicate that BWS and rating are comparable, but not similar. To answer the question on comparability of BWVr across nations, we use large data sets from the adult population in 9 nations in Europe, Asia and the Americas. In these nations, we assess response styles in best-worst measured BWVr and link the resulting Schwartz values profiles to data on attitudes and behaviour. We find that comparability of response tendencies and predictive validity is present in subsets of the 9 nations. Implications for cross-cultural research are discussed.

KW - Schwartz values

KW - Best-Worst Scaling

KW - Response bias

M3 - Paper

ER -