Narrative text biases attending the critical incidents technique

The critical incident technique (CIT) wherein informants describe their experiences and feelings attendant to a specific occasion is an attractive qualitative technique because it is easy to administer in a group setting, computer textual analysis is readily available, and it provides insight into marketplace phenomena. However, self‐disclosure theory predicts and studies have found that the informant’s gender, the target of the disclosure, and similarity between informant and target affect the quality of the informant’s narrative text. Using hypotheses from this knowledge base and adding the consumer construct of involvement, the authors conduct an experiment using CIT. Three hypotheses are supported: females disclose more than males, more is disclosed to friends than to strangers, and a high involvement topic produces more disclosure than does a low involvement one. Qualitative marketing researchers are warned that subtle biases exist in narrative text generated by the use of CIT, and precautions are suggested.