Testing for comparability of human values across countries and time with the third round of the European Social Survey

Abstract

This study tests the compatibility and comparability of the human values measurements from the third round of the European Social Survey (ESS) to measure the 10 values from Schwartz’ (1992) value theory in 25 countries. Furthermore, it explains the dangers associated with ignoring non-invariance before comparing the values across nations or over time, and specifically describes how invariance may be tested. After initially determining how many values can be identified for each country separately, the comparability of value measurements across countries is assessed using multigroup confirmatory factor analysis (MGCFA). This is necessary to allow later comparisons of values’ correlates and means across countries. Finally, invariance of values over time (2002-2007) is tested. Such invariance allows estimating aggregate value change and comparing it across countries meaningfully. In line with past results, only four to seven values can be identified in each country. Analyses reveal that the ESS value measurements are not suitable for measuring the 10 values; therefore, some adjacent values are unified. Furthermore, a subset of eight countries displays metric invariance for seven values, and metric invariance for 6 values is found for 21 countries. This finding indicates that values in these countries have similar meanings, and their correlates may be compared but not their means. Finally, temporal scalar invariance is evidenced within countries and over time thus allowing longitudinal value change to be studied in all the participating countries.