Inductive Circuits
Inductive Circuits
How to apply this framework to our visual model
The Bradford Hill Criteria are certainly comprehensive, but we still need something practical. How can we visualize these ideas in a way that will aid analysis?
The circuit model that we used for deduction can still be our model for causation. We’re just going to need to map certain elements to make it all make sense.
Let's start with our standard circuit, ignoring for a moment the Rebuttal as its inclusion doesn't affect this conversation. We can always add it back in if necessary.
Our original circuit had a Conclusion Subject and Predicate. In a causal argument, the Conclusion Subject is the Proposed Cause itself, and the Predicate is the verb "to cause" in some form followed by the Proposed Effect. So, we are going to replace our Purple Claim at the top with something only slightly different to account for this shift:
Now comes the tougher task: how do we place the rest of our casual elements?
Well, Minor Premises are supposed to be facts about the Conclusion Subject. So, which of our elements is closest to that? Isn’t that what the mechanism is? A “presentation of facts related to the cause”? So that should definitely go on the left under the proposed cause.
But there’s another element that also acts a bit like a minor premise: Temporality. That too is a fact about the Cause, that it occurs before the Effect. So that should ALSO go on the left, and since Temporality is a prerequisite for the Mechanism to even make sense, we structure it this way:
Ok, now for the Major Premise. Do we have an element that acts as a rule or principle?
…
Uh oh.
…
But wait! Isn’t that exactly what we said about inductive reasoning? We are generalizing a rule from the evidence, so that means there isn’t one yet. And that’s OK, because we DO have some Backing…
What is it that makes the Mechanism likely? Well, there are two categories of things: evidence that the mechanism is reasonable and evidence that the mechanism is likely.
What from our Bradford Hill analysis suggests that the mechanism is reasonable? Coherence and Analogy, ideas that suggest that what we’re proposing makes sense given the other things we know or observe in the world.
What then suggests that it is likely? The presence of examples that demonstrate that mechanism. That is our data and all those associated Bradford Hill elements: Strength, Gradient, Experimental Evidence, Consistency, and Specificity. These aren’t PROOF, they simply suggest that it is more likely, just as our Backing did in the deductive circuits. And so, we can make a circuit that looks like this:
So in the end, we have a circuit that looks very much like our deductive circuit: it has the same shape and the same concepts. The biggest difference is that there will never be a hard and fast Major Premise in causal reasoning, leaving us dealing in probabilities and uncertainty.
This also limits the types of questions that can be asked about a causal stimulus. For instance, the LSAT would never ask for a sufficient assumption in these questions, because there is no way to reach sufficiency. Similarly, you will not be asked what must be true/false or what can be inferred. On occasion there will be a necessary assumption question, and that will be in a case where the conclusion is about a certain cause-effect relationship, but the data address a correlation between different phenomena (we will address this in a different lesson in greater detail!)
Another thing to note:
The Bradford Hill criteria are not hard rules but rather guidelines to assess the likelihood of a causal relationship, designed to allow for reasoned judgment in the face of complex and multifactorial health issues. For the purposes of the LSAT, these criteria can often be simplified into the three-pronged model we introduced. However, correct answers will address the more detailed parts of this model frequently enough that knowing it is vital to confident and efficient analysis.