The logic of research inquiry and the use of the puzzle approach in research design

I like solving and putting together puzzles.

I like assembling evidence and theories to think about the problem, which can also be a metaphor for a puzzle (or as my dear friend Amber Wutich said, a jigsaw). I actually don’t like the type of physical puzzles shown in the photo (my brother loves assembling them but I hate not being able to complete the puzzle assembly). BUT I do like thinking about research questions and phenomena that leave me puzzled, baffled, mystified, perplexed, flummoxed.

I know that in social science scholarly circles, some people like the puzzle approach to developing research questions and others don’t particularly appreciate the model. So I asked on Twitter where we were with respect to the puzzle approach. I got several great responses (of which I reproduce a few below).

Dr. Mirya Holman, by the way, is amazing at research design.

I do this job because I love the “solve the puzzle” component of it. I see puzzles as both motivating the project overall (if X = A and Y = B, why does X*Y /= A*B) but also as how I often match data + methods to theory or design research

— Mirya R. Holman (@prof_mirya) August 27, 2022

As is Dr. Amber Wutich.

I use a jigsaw analogy a lot in teaching research design: here are the pieces that already fit, this piece is missing, what does it need to look like?

(But I also say RD has many correct solutions & that’s more a Jenga)@LibbyAWentz Diss Proposal book has a nice game exercise

— Amber Wutich (she/her) (@AWutich) August 26, 2022

Dr. Sheena Chestnut Greitens and I talk about research design all the time.

I like the concept of puzzles, in the sense of “perplexity over specific patterned outcomes.” Definitely OK/good, though not always necessary. I like the Day & Koivu subcategories of puzzles, too. That’s my two cents. 🙂

— Sheena Chestnut Greitens (@SheenaGreitens) August 26, 2022

There are, obviously, very valid critiques and concerns, as Dr. Ernesto Castañeda expresses here.

In my humble opinion. My comment is also reffering to the analogy of the jigsaw puzzle, where there is only one right outcome at the end. Research is not like that, and the biggest pay offs come from luck and finding the unexpected.

— Ernesto Castañeda (@DrErnestoCast) August 27, 2022

I use both my Twitter account and my blog as avenues to think out loud and get feedback on ideas I have been marinating. I’ve been thinking a lot about research puzzles, research questions, and research design in social science. This is normal for me, because I teach the foundational courses in our institution’s methodology sequence for the Masters and PhD programmes. I normally teach Research Design, Mixed Methods, Analysis and Interpretation of Qualitative Data, and Research Methods in Social Science. Thus, I am always thinking about better ways to teach how to craft good research questions and how to improve research design. The puzzle approach is popular in several disciplines within the social sciences and I thought I’d think about it more, this time in writing.

There are a lot of approaches to constructing research questions (and entire books focused on that very activity!), but one of the most popular is the development of a “puzzle” (or a “research puzzle”). I’ve been thinking about this particular approach for a very long time.

When you find something puzzling, you think “hmmm… this phenomenon is not operating the way I thought or I hypothesized it would, why would that be the case?” In this case, we’re not talking about puzzles like the ones you assemble, but what you find puzzling (or perplexing).

I’ve read a lot of articles and book chapters on this, and I find multiple definitions and classifications of puzzles (research puzzles, that is) somewhat idiosyncratic. But the mere notion of puzzle IS idiosyncratic: what you find puzzling, I may not find perplexing at all!

Let me give you an example: the scholarly literature on water conflicts indicates that one of the key factors to solving a dispute is to have all stakeholders agree and offer good faith solutions. HOWEVER, this did not hold in a case I studied (the Zapotillo aqueduct and dam). This conflict remained protracted for decades until the current President of Mexico came and unilaterally decided that one of the main stakeholders in the conflict (the state of Guanajuato) would not get water from the El Zapotillo dam and aqueduct.

An external actor TERMINATED he conflict, all of a sudden (not really as there are still some tensions and negotiations, but for practical purposes, that’s the case). So the puzzle here is: why (and how) did the President’s intervention change the dynamics of this water conflict? It really is puzzling!

To some water conflict experts (or specialists in conflict resolution), this external actor intervention intended to terminate a conflict may not feel puzzling at all. That’s why I find the research puzzle approach a bit shaky. The puzzle seems to be the “selling strategy”.

And by “the selling strategy” I mean the approach a scholar takes to convince the reader that the question they are asking merits being investigated, because it makes us scratch our collective heads.
“Look! This question is interesting, it left me puzzled!”

I DO teach my students and my thesis advisees how to construct a research puzzle because I find it a good “selling strategy”. Tell me why I (and the research community) should care about the question you are asking, and justify the investment in the research you will be doing.

I find that the review of the literature is a foundational step to design a good research puzzle and craft a solid research question. Puzzles specify the conditions and parameters under which the phenomenon under study contradicts the reality we are observing.

I also find @KarlGustafsson5 and @Prof_Hagstrom ‘s approach to constructing a research question through puzzles extremely helpful (I’m a political scientist, so I eschew towards those types of approaches).

Luckily their article is #Free2DownloadAndRead https://t.co/YVaO7Iz3ZJ

— Dr Raul Pacheco-Vega (@raulpacheco) August 26, 2022

One book chapter that I found extremely useful to help me think was this one, “From Questions and Puzzles to Research Project”, in the Sage Handbook of Research Methods in Political Science and International Relations by @adammccauley and @aruggeri_eu https://t.co/YsHqIa3iJ8

— Dr Raul Pacheco-Vega (@raulpacheco) August 26, 2022

As I specified in my thread, the puzzle approach (which can be seen as putting together a puzzle or as finding something puzzling, two different but complementary views) helps researchers set a potential direction for the study. That doesn’t mean serendipity doesn’t play a role.

The concept of research puzzles as thinking about something that a researcher find puzzling , strange, unexpected is in complete agreement with your point about luck and the serendipitous nature of research. So I don’t think we disagree (I also like the patchwork quilt metaphor shared by Dr. Louise Seamster)

My mentor called my current project a patchwork quilt and that made sense to me? for this reason–as compared to a quilt following a specific predetermined pattern

— Louise Seamster (@louise_seamster) August 27, 2022

I think Dr. Seamster summed it up well:

I think many of our game-puzzles provide the illusion of this process but with a fixed number of solutions–whereas a genuine research problem that puzzles you, as you’re saying, works differently. ie puzzles are not the same as puzzles

— Louise Seamster (@louise_seamster) August 27, 2022

I do hope this blog post is helpful for readers in choosing whether they want to use a puzzle approach, both to teach their students and to use in their own research projects.

You can share this blog post on the following social networks by clicking on their icon.