Speaker 1: Validity and reliability are probably among the most confusing and frustrating terms when it comes to qualitative research. There are so many definitions and so many discussions and so many alternative terms have been put forward, so it doesn't really help to understand what validity is and how we can ensure that our findings are valid or how we can increase these findings' validity. So in this video, I'll take you through six steps to increase the validity of your qualitative findings. In quantitative research, validity and reliability are quite straightforward terms. So reliability refers to replicability and consistency of certain measurements and validity to whether this measurement is measuring what it's supposed to measure. So it's quite straightforward. But think about qualitative research. Can we really talk about consistency of our instruments? Imagine that you're interviewing the same person twice and asking the same questions. Even though you're asking the same questions, this person is not likely to give you exactly the same answers. So for this reason, reliability doesn't really refer to qualitative research. It's not that relevant. And usually, people discuss validity rather than reliability of qualitative studies. And validity of qualitative research is usually discussed in terms of three common threads to validity, which are three different types of bias. Respondent bias, researcher bias, and reactivity. So respondent bias refers to a situation where your participants are not giving you honest responses for any reason. They may feel that the topic is threatening to their self-esteem, for example, or they may simply try to please you and give you the answers they think you are looking for. Researcher bias refers to the influence of your previous knowledge and assumptions on your study, which may be a very dangerous and a very risky factor in your study. I've talked about the role of assumptions quite a lot in my other videos and in my blog. And finally, reactivity refers to the role of you as a researcher and your influence, your physical presence in the research situation, and its possible influence on the data, on what the participants say, and so on and so forth. And in order to minimize the potential influence of these three types of bias on your study, Robson suggests the following six strategies to deal with threats to validity. Prolonged involvement refers to you as a researcher being involved in the research situation in your participants' environment, which is likely to result in the increase in the level of trust between you and your participants. This in turn is likely to reduce the risk of respondent bias and reactivity as you generate this common trust. However, it is likely to increase the risk of researcher bias because you and your participants are likely to generate some set of common assumptions. And as I said, assumptions may be a very dangerous thing for your research. Triangulation is such a broad topic and I'm sure that you've at least heard about it before, if not read about it. Triangulation may refer to many things, including triangulation of data, so when you collect different kinds of data, triangulation of methodology, when you have, for example, mixed methods research, or triangulation of theory, where you're comparing what's emerging from your data to previous existing theories. In any case, triangulation is likely to reduce all kinds of threats to validity, so just remember that it's always good to consider triangulating these different aspects of your study. Peer debriefing refers to any input or feedback from other people. This may happen during internal events, such as seminars or workshops in your university, or external, such as conferences. In any case, the feedback and quite likely criticism that you'll receive from other people helps you become more objective and helps you see and become aware of certain limitations of your study. And this is likely to reduce researcher's bias, so again, researcher's bias which was about your previous assumptions and your previous knowledge. So you're becoming more objective and more aware of how your study may be improved. Member checking may mean a couple of things, but in essence it refers to the practice of seeking clarification with your participants. So asking them to clarify certain things before you actually jump into conclusions and describe your interpretation of that data. So it may be simply keeping in touch with your participants, sending them a text message or an email, and asking them whether what you think they meant when they said something in the interview is actually what they meant. Another practice is to send them interview transcripts. So to send them the whole transcript and ask them to delete or change things or add things to that transcript. And finally, you have a method called validation interview, which is all about member checking. So it's basically a whole interview which serves the purpose of this clarification that I discussed. So after you've conducted the first run of analysis after the interview, you conduct another interview and you just ask your participants about your interpretations and about anything that was not clear to you. Negative case analysis is one of my favorite things to do. And I talk extensively about it in my self-study course on how to analyze qualitative data. But basically what it involves is analyzing these cases or data sets that do not match the rest of the data, do not match the trends or patterns that emerge in the rest of the data. And although you may feel tempted to ignore these cases, you may fear that they will ruin your data or your findings, quite often they tell you more about the rest of the data than these actual other cases themselves. So negative cases highlight not just how this one case is different from the rest of the data, but they actually highlight the similarities between the rest of the data. So this is a very, very valuable and important thing to do. And finally, keeping an audit trail means that you keep a record of all the activities involved in your research. So all the audio recordings, your methodological decisions, your researcher diary, your coding book, just having all of this available so you can, for example, demonstrate it to somebody. So again, this way you become really transparent and the validity of your findings cannot really be argued. Importantly, don't worry about having to apply all these strategies in your study. Firstly, some of them are almost natural, like peer debriefing. So as a student, it's very likely that you will receive feedback, you will talk to other people about your study, you will receive feedback and criticism. So you don't really have to worry about consciously applying it as a strategy. And secondly, you can choose some of these strategies, a combination of these strategies. You don't really have to apply every single one on the list. However, it is important to think about validity and it's very important to talk about it in your study. So if you demonstrate that you are thinking about validity and you demonstrate what exactly you did to increase this validity, it will be a major, major advantage to you and to your study.