**8. Data collection**

A standardized formal interview using a structured questionnaire was used in this research, as interviews tend to have a high response rate and offer the opportunity to correct misunderstanding with regard to questions. The questionnaire enabled the interviewer to carry out ratings while controlling for incompleteness. Interviews are more successful with respondents who have a reading or language difficulty and thereby can aid response rate [67].

#### **8.1. Recruitment and response rate**

It is important that the sample is representative of the population, and if a large percentage do not respond, the sample may become biased. The accepted cut-off response rate for surveys is 65% [67, 68]. In this research, an advance participation invite notification was sent to respondents. The letter included an explanation about the research; how respondents were chosen; confidentiality and anonymity; and the time that it would take to complete the questionnaire. Other strategies implied to deal with response rate were careful pilot work in order to ensure that the design of the questionnaire was delivering similar answers from respondents. Additionally, a cover letter was sent to all respondents and an incentive such as a copy of the results of the survey, if so requested, was assured by the administrators of the questionnaire. In the case of non-response, where people refused, or had moved away or were on holidays, a substitute with matching characteristics was used. The administrators also endeavoured to find out whether the reasons for non-response were somehow connected with the topic of the research. Both research administrators were afforded interview skills training as interview skills were considered an important part of the interview response rate. This training enabled the administrators together with the researcher to check responses and related problems by monitoring the completed questionnaires by comparing the results obtained by various interviews for patterned differences.
