An individual semi-structured interview was conducted with each family and formal caregiver. Each interview began with an explanation of its goal and procedure. During the interview, a guide was used to ensure no topic was overlooked. One version of the guide was designed for family caregivers and another for formal caregivers. These two versions included general questions on screams and on the key concepts derived from the theory of Adams and Gardiner (2005) in relation to this behavior: social context, inferences about the screams, and positioning within the triad. The following is an example of the questions on the social context: “Could you describe what for you is a typical visit to your family member?” The following, by contrast, is an example of the questions on positioning aimed at stimulating the reflection of caregivers in accordance with the critical aspect of the research: “Do you think there is something else you could do or change in your interaction with this older person who screams?” Finally, an example of the questions related to inferences is: “Could you describe your family member’s behaviors that help you understand why he/she screams?” General questions on screams allowed meanings of screams to be distinguished as understood by the family member, by the formal caregivers and by the older person living with dementia. Thus, triangulation of these data reflects the meanings of screams for the triad. In the course of data collection and analysis, questions were modified or added to verify with other participants what was learned previously. Interviews were conducted in the nursing home, lasting 75 minutes on average. They were recorded with the participants’ consent and transcribed verbatim.
During the research process, informal con versations also took place with staff members and families in the nursing home. They involved study participants but also other people who were interested in the study, such as head nurses, other nurses or nurses’ aids. Field notes were taken during data collection and analysis. These documented ideas, descriptions or elements that the ethnographer wished to investigate in greater depth. The notes guided the subsequent cycles of sampling, data collection and analysis. A journal was kept to document the ethnographer’s reflections on the phenomenon, the research process and relations with participants. All data were considered in the analysis.
Data analysis
Data analysis was performed following each data collection activity and guided the recruitment of new participants. First, data were reduced by assigning codes – 600 in all – to the data. The coded