Q2 - Eating garlic reduces the levels

[doug.feng]

Premise:
Two studies with similar groups being studied:
1. Eating a garlic tablet for 4 months showed a 12% reduction in cholesterol and 12% reduction in tri.
2. Eating a medically inert tablet for 4 months showed a 2% reduction in cholesterol and 3% reduction in tri.

Conclusion: Eating garlic reduces levels of cholesterol and tri and reduces risks of cardio disease.

(A): Out of Scope. If it's available to the public doesn't matter.
(C): Out of Scope. Less than 4 months doesn't matter.
(D): Out of Scope. Large amounts of garlic doesn't matter.
(E): Out of Scope.

(B): Correct because it would help to eliminate alternate causes for the decrease in cholesterol and tri.

[ohthatpatrick]

Great (and efficient) explanation.

Hopefully when most students read this, they recognize the classic "correlation -> causation" fact pattern.

There was a correlation between taking the garlic tablet and having reduced levels. Does that prove that the garlic tablet was what caused the reduced levels?

No. We have to consider alternative explanations.

The first one to consider is reverse causality. Could having reduced levels of chol/tri have caused the subjects to take the garlic tablet?

That's crazy talk!

Okay, then we have to consider alternative causes for why one group's levels reduced a lot while the other's didn't.

Here we might consider such possibilities as
- diet
- exercise
- metabolism
- genetic differences
- environmental differences

As said before, (B) relates to a potential alternative explanation. What if the group taking the garlic tablet ended up also changing some of its diet to foods that were lower in cholesterol and triglycerides. In that case, the dietary change would be the real cause for the reduced levels.

This argument, like any science experiment, has to control for all other potential variables/differences between the two groups. If the two groups are IDENTICAL, except for whether or not they took the real garlic tablet, THEN we could prove that the garlic tablet was the cause.

Nice work.

[mkd000]

It was easy for me to eliminate the wrong answer choices, and I picked (B) because what if the group that was given medically inert tablets had an amount of garlic in their diet that was equivalent to that which was being consumed by the garlic tablet-taking group? In that case, the results of the study would be questionable.

Is this a flaw question? I had difficulty assessing what question type this question is. I guess if I had approached this question believing it was a flaw question, I would have taken the approach that ohthatpatrick took. Could you please provide me with some feedback?

Thank you.

[ohthatpatrick]

It's called an Evaluate question. They're relatively rare, fewer than 1 per test.

You can go several tests without seeing one. They're Assumption Family tasks, because the whole unifying principle of the Assumption Family is that we're evaluating arguments for
- Missing Logical Links
- Potential Objections
- Alternative Explanations

Evaluate sits in between Strengthen and Weaken, in a way. The answers are always phrased as questions or "whether or not" ideas. You're basically looking for an idea that could Strengthen if we went one way with it and Weaken if we went a different way with it.

For example
whether garlic tablets are available to the public or not has nothing to do with evaluating whether eating garlic has an effect on chol/triglyc.

And with our correct answer,
if the diets of the two groups were similar/identical, that STRENGTHENS the argument. This is a better conducted experiment. We can feel good knowing that the garlic tablet group definitely consumed a different level of garlic than the control group.

Whereas, if the diets of the two groups were random / different, that WEAKENS the argument. Poorly conducted experiment. Too much noise. How can we tell whether garlic made the difference when we don't really have any idea how much garlic each group consumed in its normal eating habits?

[kkate]

You can go several tests without seeing one. They're Assumption Family tasks, because the whole unifying principle of the Assumption Family is that we're evaluating arguments for
- Missing Logical Links
- Potential Objections
- Alternative Explanations

Don't Potential Objections and Alternative Explanations sort of go hand-in-hand? Can you please help me identify the difference between the two?

Thank you!