#14 - Editorial: A proposed new law would limit elementary

[Administrator]

Complete Question Explanation

Assumption—CE. The correct answer choice is (D)

The editorial outlines two separate arguments. One is the argument held by “most parents” and the other is that of the “researchers”. The parents’ conclusion is that smaller class sizes cause students to become more engaged in the learning process. The researchers disagree, because their research shows that smaller classes do not cause any change to students’ average grades. The argument and its counterargument have the following structure:

• Argument (parents)
  
  Premise: Small class size allows teachers to devote more time to each student.
  
  Sub. Conclusion: Students become more engaged in the learning process.
  
  Conclusion: Limiting class size is a good idea.
  
  
  Counterargument (researchers):
  
  Premise: Increasing the amount of time teachers spent individually with students has left the students’ average grades were unchanged.
  
  Conclusion: The parents’ reasoning is questionable.

The question asks you to address the assumption upon which the researchers’ argument depends. You should immediately notice the logical gap in it. The researchers question whether reducing class size can make students more engaged in the learning process, because such reductions have not so far caused any change in students’ average grades. This line of reasoning clearly assumes that grades and the level of student engagement are somehow related, so that students’ average grades might provide a suitable proxy for measuring their engagement. Prephrasing the missing link between the premise and the conclusion of the researchers’ argument immediately reveals answer choice (D) to be correct.

Answer Choice (A): This answer choice is irrelevant. The researchers do not need to assume that only large elementary schools are appropriate for study. It is possible that smaller schools could be appropriate as well. In fact, the argument here is about class size not school size and so the researchers do not need to make any assumptions about school size.

Answer Choice (B): The researchers do not need to assume that teachers’ attention to their students is equal. Even if teachers’ attention is not equal, reducing class size may still have no effect on the students’ level of engagement.

Answer Choice (C): This answer choice implies that reducing class size could negatively affect the level of student engagement, because it would require reducing the number of teachers. This would strengthen the researchers’ argument that reducing class size is unnecessary. However, their argument does not require a decrease in the number of teachers as a result of the reduction in class size. Therefore, this is not an assumption of the argument.

Answer Choice (D): This is the correct answer choice. The researchers must assume that there is a connection between the degree of student engagement and students’ average grades. To prove that this answer choice contains an assumption, apply the Assumption Negation Technique: what if the degree of student engagement does not correlate well with students’ average grades? The researchers’ argument would be seriously weakened, because grades would not be a suitable proxy for student engagement. Consequently, the fact that student grades were unchanged would have little or no bearing on whether or not these students had become more engaged thanks to the reduction in class size. Since negating this answer choice weakens the conclusion of the argument, it is an assumption upon which the argument depends.

Answer Choice (E): The researchers’ argument does not require any assumptions about parental support. The researchers’ argument merely rests on the assumption that there is a connection between students’ grades and their level of engagement.

[voodoochild]

Experts,
If we say that "If x, then Y" in the conclusion-- does it mean that the author is implying correlation between X and Y ? I have seen this format in many causation/correlation problems.

Any thoughts are greatly appreciated.

[Jason Crandall]

Correlation, in a technical sense, is about how observed changes in one variable are related to changes in some other variable (as in, "As X goes up/down, Y goes up/down").

"If X, then Y" simply indicates a conditional relationship between the variables, but this format can be applied far more broadly than simple correlation. It's possible for a conditional relationship to also be a correlation, if X and Y are processes or phenomena that can change over time.

Bear in mind that correlation appears far less often than conditional relationships on the LSAT. Just as "correlation does not imply causation", "conditionality does not imply correlation".

[voodoochild]

Jason,

Thank you for your response. Can you please explain the above concept with some examples so that the concept is hammered in my brain?

I have seen "implied causality" in the conditional framework. For instance, "Had I not eaten the ice-cream sold at Walmart, I wouldn't have fallen sick last night." (In this example, I believe that causality is implied. I have seen a lot of such forms on the LSAT).

Thanks in advance.

[Jason Crandall]

My phrasing above was only meant to suggest that you should not expect to automatically find causation where there is correlation or correlation where there is conditionality.

See my PM for more details here.

[moshei24]

Can you assume correlation if there's causation?

Let's say A causes B. Does that mean that A and B are automatically correlated? How does that work?

If as A goes up, B goes up, that means they're correlated, but B might not be caused by A.

And also, if A then B implies conditionality, does that also imply causation? How would that relate to correlation?

Thanks.

There's a question where these ideas are crucial, but I can't remember off the top of my head which test it's on.

It was a question with teachers and lowering class sizes. Ring a bell?

[moshei24]

The question is from June 2010 LR #1 Q: 14.

Also, correlation can be reversed. If A is correlated to B then as one goes up, so does the other. Correct?

But if A causes B, then B occurring has nothing to do with A, while if A is correlated to B, B increasing means A increases, if it's a direct correlation. Correlation has to do with time or changes in something, correct?

Causation has to do with anything, really. Just that A causes B, so if A occurs, B occurs. And conditionality goes even further, if A then B, then if A occurs, B occurs. And if B does not occur, A does not occur. Do I have the different concepts down now?

Can you clarify, please?

Like in the question I mentioned earlier, the correlation applies because it's necessary for the degree of student engagement to correlated with student grades, because that proves the researchers' point that since the grades didn't change it must mean that the degree of student engagement also didn't.

One last thing. A is correlated with B is the same thing as saying B is correlated with A, correct?

Thanks!

[Nikki Siclunov]

Correlation refers to any statistical relationship involving dependence. Assuming a direct correlation between A and B, when A increases B increases as well (and vice versa). Such a dependence is not sufficient to prove that A causes B, because there could be a third factor (a confounding variable) that causes both. Alternatively, it is possible that the relationship is reversed, i.e. that B causes A.

While we never doubt correlations, it is possible that the causal conclusions drawn from them are flawed for the above-mentioned reasons. Causation is a matter of argumentation: it is rarely a de facto premise; usually, it is something we conclude based on data, correlations, or other empirical observations. Even if we can prove that A causes B, the occurrence of B still does not prove that A was the cause. This is because there could be a myriad alternate causes that produce the same effect.

Conditional and causal reasoning are frequently conflated. In conditional reasoning, the statement A B simply suggests a dependent relationship between A and B. That does not mean that A causes B, or that B causes A. Of course, conditional reasoning does not preclude causation. For instance, the following claim has both conditional and causal aspects to it:

If it rains, it must be cloudy.

Conditionally, the relationship can be diagrammed as R C

There is an inherent causal relationship between rain and clouds, but it goes the other way around: clouds cause rain (C R). It does not mean that every time the cause occurs the effect occurs. Although "clouds" cause rain, they are not a sufficient condition for rain. It is imperative NOT to confuse conditionality with causation, as they are distinctly different ways of creating arguments and one neither presupposes nor precludes the other.

In the question you mentioned, the correlation between grades and the degree of student engagement was an assumption required by the researchers argument, which, in its simplest form, is this:

Premise: Increased time with students does NOT cause student grades to increase

Conclusion: Increased time with students do NOT cause student engagement in the learning process.

Obviously, we need to assume that grades and engagement correlate. And yes, to say that A is correlated with B is the same thing as saying B is correlated with A.

[moshei24]

That was very helpful. Thank you, Nikki.