Thread by @raulpacheco

I have been thinking about writing a thread on how to link theory with research, which probably fits with the question that was asked by @PhDForum earlier today - how do we choose a theoretical framework.

I'm going to try to formulate this discussion as clearly as possible.

This discussion about how to link theory with research (and with the method) is one I have had with @salazar_elena and @gcaleman for a while now. How do we link all the theories we read into what we see in the empirical work?

I believe that there are three elements at play.

1) There are various types and levels of theory (grand theory, meso-level theory, micro-level theory), etc.

2) We (scholars, students, practitioners) need to read very broadly to be able to discern across theories.

3) We need to learn how to establish THEORETICAL EXPECTATIONS

One example I really enjoy and use frequently when teaching about theory and empirics is this paper by @ProfTortuga on forbearance:

cambridge.org/core/journals/…

In this paper, Holland develops a new theoretical approach based on her empirical research across countries/cities.

Note how in her abstract, Holland quickly summarizes theoretical expectations AND her contribution. There are at least 2 theoretical approaches that help explain the gap between what's encoded as law & what's actually occurring on the ground: resource constraints and bureaucracy

Holland establishes two theoretical expectations:

1) if there is poor state capacity to monitor and enforce, we may expect that there will be limited enforcement activity.

(she cites Levitsky and @VickyMurilloNYC on this)

2) if there's inadequate bureaucratic control...

... there's a higher likelihood that there will be limited enforcement activity.

These are two observations that Holland makes from absorbing, summarizing, integrating and presenting the various theories surrounding poor regulatory enforcement.

She then introduces her own...

Holland makes it clear how her framework borrows from other theories (including price theory) and in doing so, these borrowed theoretical concepts help her explain how states choose not to enforce regulation.

This is an excellent example of how to apply theory to explain things

Theories help establish an expectation of how the world should work. We need theory to establish exactly what we expect to see

Empirical research then tests those theories and asserts whether the theories being used actually do help explain the phenomenon we are observing.

So, if we reverse-engineer Holland's paper, we can see the empirical phenomenon she is looking to understand (limited, constrained regulatory non-compliance).

She then establishes the various theories that could potentially help her explain this non-compliance/non-enforcement.

We choose the theory depending on the empirical phenomenon we are examining and the research question we are trying to understand, and our prior experience (and reading/understanding) of how the phenomenon will operate.

Thus selecting a theoretical framework does not happen "a priori".

I never decided that "oh I am going to study the governance of river basin councils using the Ostroms' frameworks".

I looked at the phenomenon, and reviewed the literature to see how others have looked at it.

For example, in my work on water conflicts, I look at the different theories on which factors could combine to make a water dispute happen.

There are theories that indicate that under resource scarcity conditions, actors will want to hoard resources and thus engage in conflict.

Theories help the researcher set theoretical expectations tha can then be tested on the ground.

For example, in this paper, Snow et al investigate the factors that drive protests by individuals experiencing homelessness academic.oup.com/sf/article-abs…

Snow et al find empirical support for resource mobilization theory (one of the most popular among scholars of social movements).

Thus, instead of arriving with a theoretical framework in hand, we need to establish which phenomenon we want to study/explain and the theories...

... that have been previously used to explain this phenomenon.

Thus, in closing:

a) selecting a theoretical framework for a study usually happens after reading and synthesizing a lot of literature on how the phenomenon has been analyzed before.

I wouldn't do it "a priori"

b) linking theory with research is particularly important because it helps us establish theoretical expectations (and develop alternative explanations, something that apparently has been forgotten when teaching research design).

Alternative explanations are based on theory.

I strongly believe it is fundamental that we teach our students both elements, how to link research with theory and how to select a theoretical framework, and if I were to add a third element, how to establish alternative explanations for the same phenomenon and discern...

... which elements/theories/evidence best explain what we are trying to understand.

</end thread>

NEW BLOG POST: Linking theory with research, choosing a theoretical framework and developing alternative explanations raulpacheco.org/2020/12/linkin…

This thread in handy blog post format.