Measuring higher education: comparing Danish survey and register data in the study of values

Abstract

Abstract This paper addresses both methodological and substantive issues. The methodological issue relates to the measurement of education, particularly tertiary education, in social science research, while the substantive issue concerns the relationship between higher education and social values. These issues are closely linked: if the measurement of tertiary education is biased, does this affect empirical findings regarding social values? Linking Danish data from the European Values Study 2017 with administrative registers enables a comparison of self-reported education with official records and an assessment of whether inconsistencies affect observed correlations between tertiary education and value orientations. The analysis reveals substantial divergence. The overall agreement between the two data sources is approximately 70%. Discrepancies are most pronounced for short-cycle tertiary education, where only 32% of cases align. Reporting a higher level of education than that recorded in the registers accounts for most inconsistencies, with 25% of respondents claiming tertiary education not found in the registers, whereas only 2.5% report a lower educational level than that recorded. The analysis identifies age, gender, income, age at immigration, and survey mode as significant predictors of discrepant reporting, suggesting influences such as social desirability and the non-recognition of foreign degrees. Despite these discrepancies, comparisons of value orientations across tertiary educational levels show striking consistency between register- and survey-based measures, with only one statistically significant difference in value orientations (out of 18 possible comparisons) between self-reported and registered measures of educational level. Despite the methodological challenges associated with survey data, these discrepancies do not appear to distort the main findings.