Logical Reasoning

PT130 · S3 · Q19 Recent studies have demonstrated that smokers

A free, expert breakdown of this official LSAT Logical Reasoning question.

Save & drill this skill — build targeted practice sets from questions like this one
Video walkthroughs — watch every question solved step by step
81 official LSATs — as questions, timed sections & full-length tests

Recent studies have demonstrated that smokers are more likely than nonsmokers to develop heart disease.

Conclusion

There’s a positive correlation between drinking caffeinated beverages and developing heart disease (i.e. people who drink caff beverages are more likely than those who don’t to develop heart disease)

Evidence

People who smoke are more likely than those who don’t to develop heart disease.

People who smoke are more likely than those who don’t to drink caff beverages.

Evaluation

This is a weird argument, because it’s not going from correlation to causality. It’s just trying to chain together correlations.

If is A is correlated with B, and A is correlated with C, is it fair to say that B is correlated with C?

It seems reasonable, but it doesn’t have to be true. Consider this example:

People who are friends with Jenny are more likely than those who aren’t to like Taylor Swift. (Maybe 80% of her friends like T-Swift, while only 40% of her non-friends do)

People who are friends with Jenny are more likely than those who aren't to go to Palomar High School. (Maybe 80% of her friends go to Palomar, whereas an insanely tiny fraction of the 8 billion strangers who aren't her friends go to Palomar).

Does this mean that People who like Taylor Swift are more likely than those who don’t to go to Palomar High School? No, we couldn’t possibly get that sort of data from the first two facts. Just because A is correlated with B and with C doesn’t mean that B is correlated with C.

Goal

Look for an answer that says you can’t assume a correlation between two traits that are each separately correlated with a given trait, or something that allows us to argue that people who drink caff beverages are not more likely than those who don’t to develop heart disease.

19.

The argument's reasoning is most vulnerable to criticism on the grounds that the argument fails to take into account the possibility that

Correct
smokers who drink caffeinated beverages
Why this is right
This controls for smoking, and shows a negative correlation between drinking caff beverages and developing heart disease. This directly undermines the truth of the conclusion.
Skill tested: Flaw · how this choice captures the argument's function is the move to repeat next time.
57% picked this
something else, such as dietary
Out Of Scope Comparison: "most important"
The argument isn’t trying to assess which factor is most important in the development of heart disease. In fact, the author isn’t even concerned with causality, as her conclusion mentions. She’s only trying to establish a statistical truism: people who drink caffeinated beverages are more likely to develop heart disease.
7% picked this
drinking caffeinated beverages is more
Opposite
This strengthens the argument, by basically lending support to the truth of the conclusion that drinking caffeinated beverages is positively correlated with heart disease.
3% picked this
it is only among people
Too Weak / Unclear Impact
If most of the population has no correlation between A and B, but specific subset of the population does have a positive correlation between A and B, then when you average the whole population, there will still be a positive correlation between A and B. So this strengthens the conclusion. People may feel like it’s an objection, because it’s saying, “The conclusion is true, but only because it’s true for a subsection of the population”. That still mildly strengthens the argument, even more so if that subset (people with a hereditary predisposition to heart disease) is a large segment of the population, which it probably is. It will be watered down by the segment of the population for whom there is no correlation, but there will still be some skewing of the data towards a correlation. Say you had four buckets of Skittles. The five flavors are evenly distributed, so about 20% of the Skittles in each bucket are red. In a fifth bucket, you have like 90% red Skittles, 10% an even mix of the rest. If you combine those five buckets into one huge pool, the % of Skittles that are red will be higher than the normal 20% rate. It won't be anywhere near 90%, but it will still be elevated. There will still be a correlation of red appearing in the population. Similarly, the people with hereditary disease are the bucket of red skittles. A ton of them have a correlation between caffeine and heart disease, and when you combine these hereditary disease people with the rest of the general population (who have no correlation between caffeine and heart disease), you still end up with some statistical lumpiness that makes caffeine drinkers more likely to have heart disease than non-caffeine drinkers.
10% picked this
there is a common cause
Famous Flaw: Correlation vs. Causality
This argument isn’t causal, so it’s pretty out of scope what causes what. LSAT definitely primed us for this type of answer, but the author specifically says in her Conclusion that she is not assuming a causal connection between caffeine and heart disease. She’s only concluding a statistical correlation. If there’s a common cause underlying both drinking caff beverages and developing heart disease, then that strengthens the author’s conclusion that those two things will have a positive correlation.
23% picked this

Continue the review in LSAT Lab

Save this question, watch the video walkthrough, and drill similar questions in your LSAT Lab account.

Save and practice this skill

LSAT Lab

Turn this review into a targeted study plan.

Save this question, drill more like it, watch the video walkthrough, and track your progress in your LSAT Lab account.

Start practicing free