feedback on the puzzle assignment

It’s about time that I start writing my own blog entries. As I have stated, this is all about timely feedback, and I understand that you are looking for my feedback as a teacher. I really like this assignment, as well as the book from which it originates± Constructing Social Theories by Arthur Stinchcombe (1968). This book presents, in very clear language, a view on the practice of social science research, which includes the building of theories. The exercise thus demystifies theory formation, and that is an important aim for me.

In addition, I see the exercise as a way to developing rigor in your thinking. And this is where I want to give you some practical tips.

In order to facilitate the discussion, let me translate the assignment into some more formal language. A good puzzle can be summarized as a surprising correlation between two variables (A, B). This correlation is something that is there, i.e. it is an empirical reality that has been found by people collecting data.

In explanations, we develop a plausible storyline which connects A and B, and this implies that one is the cause (or independent variable) and the other the consequence (or dependent variable). 'Independent' here means that we do not consider the causes of this variable. Without getting too philosophical about it, the causal link always remains a storyline: it can never be proven empirically. What we can do is to collect data in such a way that it provides strong evidence that the relationship is there (or: that we fail to find evidence that proves that the relationship is not there, which is considered to be a better test).

Using the language of dependent and independent variable allows me to summarize some of the things that can go wrong in your assignment (and this means that if you formalize your own puzzle along these lines, you can detect flaws yourself more easily!):

1. Your puzzle does not address two, but only one variable (such as: why is the CO2 emission level not rising in the EU). This means that you identify a dependent variable (B),  and your explanations then propose alternative independent variables  (A1, A2, A3). The assignment is to come up with three alternative ways in which the same A and B are connected. This is a common flaw, so don’t worry about it too much. You are certainly in the right direction, and when you have done this, half the work has already been done.

2. Your puzzle has an A and a B, but your explanations do not deal with the relationship between the two. A common mistake is to develop alternative explanations for the independent variable A, or to develop alternative independent variables (A’s). Again, this is simply a matter of forcing yourself to focus on the A and B you have chosen, and specifying the relationship between them.

Some of you come up with interesting puzzling situations (in class we discussed the one where ozone layer destruction and climate change have different time frames when it comes to developing solutions). But these are then not developed in a rigorous way to become a puzzle with the structure A à B. If you do not force yourself to fit in that structure, it becomes more difficult to be systematic, and develop good explanations about that exact relationship. So please use this feedback to develop your puzzle towards that goal! In the classroom we saw that it was not that difficult to translate the situation into a more formal structure.

Once you have clearly identified the variables, and thought up an explanation, it is important that you describe this explanation as concisely and precisely as possible. This means you provide a storyline of how variation in the independent variable leads to systematic variation in the dependent variable. When I ask you to develop storylines, it does not mean that I want you to present the three alternative explanations in a flowing piece of writing. So number the explanations (making clear where one stops and another one begins), and write them down concisely. This is important for the reader, but notice that following these tips also forces you to fix parts of the storyline that maybe you thought where there, but turn out to be missing when you need to write them down in this way.

Finally, a remark that is not so much about what should have been there, but what is the next step in developing explanations for a puzzle. This is to develop your puzzles and explanations in such a way that they can be tested in a research project. A first question here would be how you can observe your variables, which is sometimes easy, but in other cases might present difficulties. Second question is about the cases you will select for observation.