Barriball, K. Louise, and Alison While. "Collecting data using a semi-structured interview:a discussion paper." Journal of Advanced Nursing. 19. (1994): 328-335. Print.

This article discusses the steps made by a research team to increase validity and reliability within a large research project focused on continuing education for nurses in two health authorities.  The paper's intent is to highlight advantages, disadvantages, and specific pitfalls created when doing semi-structured interviews.  The authors concluded that despite difficulties the interview provided some real benefits including increased interest from the nurses and the best method for examining perceptions among a large and diverse population.

While this article doesn't present the research findings of the study it does grow out of an actual study using the techniques I am using, semi-structured interviews.  The authors, both highly educated nursing educators and researchers, were extremely thorough in their treatment of validity and reliability issues.  Medicine as a science is largely used to methods that are more quantitative while education lends itself to more qualitative examinations.  This study's research question and method straddles both fields.  By carefully addressing all of the possible problems in this separate paper the authors strengthen the results of their study.  Some studies I have read addressed reliability and validity, while others ignored it.  I have never seen it discussed as free standing supporting document.  The point by point considerations found here are useful for anyone conducting interviews in design considerations, conduct the interview and in offering a defense of their findings.

After briefly introducing their study the authors move directly to a discussion a problem affecting reliability recognized during the conception of the study, non-response.  If left unaddressed the studied will be skewed by a selective group of respondents that isn't representative of the actual population.  To address non-response and other problems with surveys the researchers, during the design phase, chose semi-structured interviews as a method to help with the following issues:

1. Overcome poor response rates of surveys.
2. Well suited to the exploration of attitudes, values, beliefs and motives.
3. Provides opportunity to evaluate validity of the respondent's answers by observing non-verbal indicators.
4. Facilitates comparability by ensuring that all questions are answered by each respondent.
5. Ensures that the respondent is unable to receive assistance while formulating a response.(329)

The researchers found "face to face contact with each potential respondent had three important outcomes;"(330) increased interest in the project, ability to reconsider by hesitant participants, and collection of information from non-participants.  The semi-structured interview, with standardized questions, advantaged the research by overcoming language barriers, clarifying tacit assumptions, providing the interview with latitude in probing responses, and increasing interactivity between interview and participants.  The authors also addressed ethical concerns revolving around the use respondents time and the need to include their input when creating an interview framework. Of special interest to me was a discussion of training of interviewers.  Because this was a large research project their was a need to use a team of interviewers.  To increase validity the research team instituted a very rigorous training regiment including reading of applicable texts about nurse continuing education, focused discussions on each question and dry runs.  They also instituted ongoing "maintenance"(334) of interviewer's skills.  The research team's treatment of the interview process was very complete and an inspiration for me on how to secure the best results.  I will adopt some of their strategies including time management, and a practice routine for myself before giving my first interview.