Page 97 - THE DUTCH TALKING TOUCH SCREEN QUESTIONNAIRE

Evoluation of the Dutch and Turkish version
own words how he interpreted the question and why he chose the answer options he chose. When a respondent encountered problems in responding to a question, he was asked what he thought the exact nature of the problem was and why he behaved as he did in response to the question. He also was asked for suggestions for improvement of the question in terms of wording, layout, instructions, etc. Additionally, the respondent was asked to describe his health problem(s) and treatment goal(s) in his own words. Comparing these descriptions to the respondent’s responses to the questionnaire during step 1 of the TSTI provided useful information as indicators of the validity of the data collected by the DTTSQ. Finally, the respondent was asked if he recognized himself in the outcomes of the questionnaire that were shown at the end of the questionnaire (see Appendix 1: 15. “Overview all outcomes of the questionnaire” ). After the TSTI was finished, researcher IT collected the demographic data through a brief structured interview.
Data Analysis
Data were analyzed using a thematic content analysis approach [27]. Four types of data were analyzed: (1) video recordings of the first two steps of the interview, (2) Dutch transcriptions of the third step of the interview, (3) observed respondent behavior in field notes, and (4) background information regarding the educational level, age, gender, and occupation of each respondent. Researcher MW started with open coding, coding all fragments of the twenty-four transcripts of step three of each interview using MAXQDA 10 of VERBI Software GmbH, Berlin. The codes and fragments of seven randomly selected transcripts were validated by two peer researchers by independently coding each transcript with the coding scheme developed by researcher MW. Differences in fragmentation or coding were discussed during consensus meetings.
To get more familiar with the data and to create an overview, researcher MW made a descriptive summary of each case on the basis of all four types of generated data after she finished open coding. Each summary contained all emerging themes regarding problems that occurred during the four phases of the response process as described by Tourangeau: (1) comprehension: (a) comprehension of text and wording and (b) interpretation of the meaning of the text, (2) retrieval: gathering relevant information, (3) judgment: assessing the
91
4.1