Second, the effect of low variability in service ratings on the size of the correlation between tip
size and service quality cannot legitimately be called a bias if the ratings accurately reflect consumers’
perceptions of service quality. To sustain their argument that restricted variability in service ratings bias
tests of the tipping–service relationship, Bodvarsson and Gibson would need to provide some evidence
that the restricted variability in service ratings is due to response or sampling biases. They provided no
such evidence.Readers who doubt that restaurant service is as uniformly good as suggested by the data