Using rapid cycle tests of change to develop the Carers Assistive Technology Experience Questionnaire: a cognitive interview study in the UK.
Sriram V., Jenkinson C., Peters M.
We describe the use of rapid cycle tests of change to pretest and develop a Carers Assistive Technology Experience Questionnaire for a survey of informal carers of persons with dementia. The Plan-Do-Study-Act (PDSA) cycle is a commonly used improvement process in healthcare settings. We used this method for conducting rapid cycle tests of change through cognitive interviews to pretest the questionnaire. The items for the questionnaire were developed based on an earlier systematic review and qualitative study. PDSA cycles were used incrementally with learning from each cycle used to inform subsequent changes to the questionnaire prior to testing on the next participant. DESIGN: Qualitative with use of cognitive interviews through rapid cycle tests of change. SETTING: UK. RESULTS: Nine participants were recruited based on eligibility criteria and purposive sampling. Cognitive interviewing using think aloud and concurrent verbal probing was used to test the comprehension, recall, decision and response choice of participants to the questionnaire. Seven PDSA cycles involving the participants helped identify problems with the questionnaire items, instructions, layout and grouping of items. Participants used a laptop, smartphone and/or tablet computer for testing the electronic version of the questionnaire and one participant also tested the paper version. A cumulative process of presenting items in the questionnaire, anticipating problems with specific items and learning from the unanticipated responses from participants through rapid cycle tests of change allowed rich learning and reflection to progressively improve the questionnaire. CONCLUSION: Using rapid cycle tests of change in the pretesting questionnaire phase of research provided a structure for conducting cognitive interviews. Learning and reflections from the rapid testing and revisions made to the questionnaire helped improve the process of reaching the final version of the questionnaire, that the authors were confident would measure what was intended, rapidly and with less respondent burden.