Qualitative Data: An Introduction to Coding and Analysis

9. Designing Your Next Study Using Theoretical Sampling

9 ■ ■ ■ ■ ■ ■ ■ ■ ■ Designing Your Next Study Using Theoretical Sampling The Promise Keeper Fathers IN THE PREVIOUS CHAPTER, we proposed that theory developed in the context of a specific sample could extend beyond that sample to apply to other sample populations. After you have completed your first research study, you will have to deal with the following question. The theory that you developed in your first research study is applicable only to the research sample from that study. How do you further develop your theory so that it is useful for understanding other research samples? This is what we call the transferability question. As you will recall from chapter 8, the qualitative concept of transferability roughly corresponds to the quantitative concept of generalizability. In doing qualitative research, you try to develop transferable theories. This means that the abstract pat­ terns described by your theoretical constructs will be applicable to other research samples, even though their specific content will not be. Here is how the transferability question can be stated for the role strain construct that we developed in the Haitian Fathers Study. 92 ❙ Designing Your Next Study Using Theoretical Sampling How does the role strain construct that emerged from our particular re­ search sample apply to other groups of Haitian fathers, and to other subcul­ tures of fathers? The question, then, is how to develop the theory from your first re­ search study into a more transferable theory. There are two distinct issues. 1. How do you design your next study (or sequence of studies) in order to develop your theory further? 2. How do you analyze the data from your next study (or sequence of studies) in order to develop your theory further? Designing your next study will be the topic of this chapter, and analyz­ ing the data from your next study will be the topic of the following chap­ ter. Our discussion of the design question will focus on strategies for se­ lecting a new research sample, using the method of theoretical sampling. After we explain theoretical sampling, we will present five steps for de­ signing your next study. As in previous chapters, we will illustrate the methodology with mater­ ial drawn from the Yeshiva University Fatherhood Project. In particular we will describe the study that we published after the Haitian Fathers Study, which investigated fathers involved in the Promise Keepers, a fundamen­ talist Christian movement. The original paper describing this work is in­ cluded as Appendix C (Silverstein, Auerbach, Grieco, & Dunkel, 1999). You should read through it at this point. Theoretical Sampling What Is Theoretical Sampling? The method for selecting a research sample for your next study is called theoretical sampling. Theoretical sampling is the process of choosing a research sample in order to extend and refine a theory. The Haitian Fathers Study provides an illustration of the need for the­ oretical sampling. In that study, we described Haitian fathers’ subjective Designing Your Next Study Using Theoretical Sampling ❙ 93 experience of fatherhood in terms of two theoretical constructs for which we had extensive textual evidence: BICULTURAL GENDER ROLE STRAIN, and CONSTRUCTING A MORE GRATIFYING DEFINITION OF FATHERHOOD. In addition, we developed two other theoretical constructs: RELIGIOUS FACILITATING IDEOLOGY, and SOCIAL SUPPORTS. We had less solid tex­ tual evidence for these two constructs. The goals of our next study, the Promise Keepers Study, were to dis­ cover whether these theoretical constructs could be extended beyond the Haitian fathers sample, and if so, to gather more textual evidence to sup­ port them. We used theoretical sampling to achieve these goals. Differences Between Random Sampling and Theoretical Sampling Table 9.1 summarizes the differences between theoretical sampling and random sampling. The differences reflect the contrast between qualitative and quantitative approaches to scientific knowledge. To uncover the differences, consider this question about our Haitian Fathers Study: What useful knowledge about fathers can come from a sin­ gle study that uses a convenience sample of 22 self­selected fathers from a particular religious community in Brooklyn, New York? One very natural answer is: “Not much useful knowledge at all. Your study only describes those 22 men.” Both the random sampling perspective and the theoretical sampling perspective agree that more research needs to be done before one can Table 9.1 Differences Between Random Sampling and Theoretical Sampling Random Sampling Theoretical Sampling Goals Generalizable propositions Transferable constructs about a population about a research concern Methods for Representative...