A post edited by Kimmo Sorjonen
Both our own anecdotal perceptions and past research findings suggest that how we feel about our own worth, our self-esteem, affects how satisfied we are with our job. However, a new study has suggested that the findings may be over-estimated, and a different approach may be more suitable.
This may be because of the way past research has estimated the association between self-esteem and work experiences. This work used a method which accepts that different sources of data can have unique variations and then attempts to derive cause and effect. However, when a variable is measured twice, the second measurement tends to be less deviant than the first measurement (“regression to the mean”). When examining change in these variables, this can lead to an overestimation of the strength of the relationship between variables.
For example, let’s assume that there is a general positive relationship between self-esteem (perhaps caused by a confounding positive/negative personality trait) and job satisfaction. Even when people start with the same job satisfaction but different self-esteem levels, those with high or low self-esteem will tend to appear to experience a change in job satisfaction to the second measurement, even though this may not be the case. This occurs because the measurements tend to move towards a less deviant value between the two time points. As a result, we may erroneously conclude that job satisfaction really changed in response to initial self-esteem when, in fact, there may be no or a small effect.
From their analysis of the findings of 30 independent longitudinal samples, totaling 53,112 participants, Krauss and Orth (2022) concluded that self-esteem affects and is affected by work experiences. In their re-analysis of the data, Sorjonen, Melin and Melin (2023) identify “regression to the mean” as having significantly contributed to the reported bidirectional effects.
Given these findings, it is important that researchers are aware of the limitations of any statistical methods but specifically the statistical phenomenon of regression to the mean. The authors suggest a complementary method with a reversed treatment of time, in order to discriminate between “not yet disproven” and spurious increasing or decreasing prospective effects.
In summary, the potential relationship between self-esteem and job satisfaction remains an interesting and promising area of research. However, this constructive reanalysis shows that caution is needed in studying the relations between self-esteem and job satisfaction, and can help further our statistical understanding of personality psychology
Readers interested in the study can find the full paper available open access here.