Revisiting the expectancy disconfirmation model: confirmation instead of disconfirmation?

The expectancy-disconfirmation model (EDM) has been widely tested and used as a way to explain citizen satisfaction in the public sector. Satisfaction is not only determined by absolute performance, but also by the difference... [ view full abstract ]

The expectancy-disconfirmation model (EDM) has been widely tested and used as a way to explain citizen satisfaction in the public sector. Satisfaction is not only determined by absolute performance, but also by the difference between initial expectations and actual performance. However, there has been little experimental work in this area. A study by Van Ryzin (2013) has been the only one to date to have fully tested the EDM in an experiment. Yet this study has not been able to confirm the heart of the model: the effect of disconfirmation on citizen satisfaction.

We carried three studies to replicate the original experiment: one close replication, one study with more extreme manipulation (expectations and performance cues) one study that reversed the order of the performance and expectations manipulation (expectations were provided ‘in hindsight’).

We found no support for the existence of a disconfirmation effect in these three studies. Instead, in study 3 - when expectations are prompted after a performance cue - we found that expectations may have an anchoring effect on satisfaction judgments. As such, a combination of high expectations and high performance cues results in relatively high satisfaction ratings, yet a combination of high performance and low expectations will lead citizens to anchor around the low expectations statement. This results in much lower satisfaction ratings, even though the objective performance remains the same.