Social Desirability and Central Tendency Bias in the Survey and Variation of Life Satisfaction

Abstract

Even though life satisfaction is an empirically extensively researched well-being concept, there are still open questions regarding the variation of life satisfaction and the influence of variables that are not collected in data sets such as the German Socio-Economic Panel (SOEP). In this article, I explore the question to what extent the variation of life satisfaction is influenced by biases resulting from the setting of the life satisfaction survey. Using different methodological approaches, including distributional analysis, descriptive statistics, Markov transition probabilities and logistic regression, I investigate to what extent a possible central tendency or social desirability bias influences the variation of life satisfaction. This brings me to the following conclusions: 1. People who are interviewed in the presence of an interviewer are more satisfied with their life than people who are interviewed in writing without an interviewer. This suggests that people express themselves more positively about their well-being in the presence of others. 2. On the 11-point integer life satisfaction scale defined between 0 and 10, the interviewees prefer to choose points 5 and 8, which form the midpoints of the scale, and choose these also repeatedly more often than other points. This is important for adaptation research: it suggests that people prefer to report life satisfaction that is balanced between good and bad, leaving room for improvement even in satisfied states. The social desirability and central tendency bias have a stronger influence on the variation of life satisfaction than many happiness correlates examined. Regarding the growth of life satisfaction, point 8 seems to be the maximum on average. A higher value is implausible on the basis of the study results.