The question stem says, which of the following, if true, most seriously weakens the argument, so this is a weakened question.

Let's identify assumptions that this argument makes, and then pick an answer that will undermine one of those assumptions.

So out of people who went to a psychologist to help with a personal problem, the magazine asked some of those people questions.

Of the people who responded who had received treatment for 6 months or less, 20% claimed that treatment made things a lot better.

Of those responding who had received longer treatment, so more than 6 months, 36% claimed that treatment made things a lot better.

And so here are the two different groups. 6 months or less, 20% say they got better.

The longer treatment group, the people who got longer treatment, 36% said they got better.

Based on that fact, the author says, therefore, so here's the conclusion, therefore, psychological treatment lasting more than 6 months, so longer, is more effective than shorter treatment.

Now because this is a weakened question, we know that there is something flawed about this argument.

There's some reason why these statistics here do not actually prove that longer treatment is more effective.

So let's think about why it doesn't actually prove that longer treatment is more effective.

Now one framework to view this problem that might help you see what's wrong with it is the phenomenon hypothesis framework.

That is a very common framework that is being used in weakened questions and strengthened questions.

The author is looking at particular statistics from a study. This is the phenomenon, what happened during the study.

And then the author has a certain explanation or hypothesis for what explains these statistics.

The author's looking at these differing percentages and thinking, huh, I wonder why the group that was treated longer seemed to have a higher proportion say they got better.

And in the author's mind, it's because, oh, well, something about the longer treatment is actually more effective.

It's better to be treated for longer. This is the author's theory.

That's why 36% say they got better here, but only 20% say they got better here.

But of course, that's not the only explanation for what happened.

Can you think about other reasons that might explain why the group that got longer treatment had a higher proportion say they felt better?

Now there can be all kinds of explanations, so I don't think you need to come up with any specific explanation,

but let me give you an example of some of the kinds of things that might come into play.

What if the kind of person who chooses to go for a longer treatment as opposed to a shorter treatment is wealthier

and therefore has more resources to help themselves or can do what the therapist says more seriously?

They can implement the advice more effectively.

Or what if the kind of person who chooses to go for a longer treatment has problems that are easier to solve?

Now there might be all sorts of other explanations that I'm not even considering, but you don't have to think about every specific explanation possible.

You just need to generally recognize that to weaken this argument, one of the main ways to do that is to propose a different explanation for the results.

Here, that explanation I think is probably going to relate to the fact that we were not told that the kind of person who gets six months or less treatment

is similar to the kind of person who gets longer treatment.

Nothing about the way that this survey was done suggests that they controlled for relevant differences between the groups.

And if they didn't control for that, that could potentially be an explanation for differing results.

Now there are other problems in the argument too.

I mean one baseline assumption the author is making is that reports of how a therapist made things a lot better are indicative of actual improvement in treatment.

Now I don't think that's an unreasonable assumption, but technically it's still an assumption.

All we got was how these survey respondents said they felt.

The author must assume that that indicates how they actually feel.

And that how they actually feel relates in some way to actual effectiveness of treatment.

So if there were answers that suggested what they said does not actually indicate effectiveness of treatment, that could be a weakener too.

So let's take a look at these answers now. We're looking for what weakens the argument.

10% of them said that treatment made things worse.

So A is telling us that in this group, 10% said they got worse.

Now if you're tempted by this answer, you're probably thinking, oh well if 10% said they got worse, isn't that like something bad about longer treatment?

So maybe longer treatment actually isn't better.

But the problem with that line of reasoning is that you have no idea what percentage of the shorter group said things got worse.

For all we know, maybe 50% of this group said things got worse.

And so 10% would actually be better than 50% when it comes to who got worse.

So because you don't know how this 10% worse rate compares to the rate who said they got worse in the shorter group, A does not weaken this argument.

So what B is saying is, this group was more likely to actually answer the questions than this group.

And just to make sure you really understand what B is saying, let's say that this magazine sent out questionnaires to 100 people who had six months or less treatment.

And they also did the same with the longer group. They asked 100 of those people to respond to the questionnaire.

What B is saying is that this person was more likely to respond than this person.

So for example, maybe for the six months or less group, only 50% of people responded.

But for the longer group, maybe 99 people responded.

So almost everybody responded. This is the kind of thing that B is saying.

But this doesn't weaken the argument because the actual number of people who responded is completely irrelevant.

It's irrelevant because we're talking about the percentage response that said they got better.

20% said they got better versus 36% said they got better.

So we're still dealing with the idea that out of this longer group, a higher proportion of who responded said they got better.

For the shorter treatment group, a lower proportion of the people who responded said they got better.

What explains that? What explains the differing proportions of people who are saying they got better?

The overall number of responses does not offer an explanation for that.

Now if you are still not convinced that B is not a weakener here, let me give you an example of some statistics that actually would make B relevant.

Imagine that this stimulus here, instead of being about the percentage who said they got better,

imagine that what was said was that we asked 100 people who got the shorter treatment and then an equal number of people who got the longer treatment.

And then imagine that the results were not about the percentage, but just about the number of people.

Let's say, for example, 10 people said they got better in the shorter treatment group,

and then 20 people said they got better in the longer treatment group.

If this was a statistic and then the author said,

Look, more people said they got better in the longer group than in the shorter group, so I think longer is more effective.

If that's how the author made the argument, then B would be relevant because pointing out that this group was more likely to respond

would act as another explanation for why you have a higher number of people who said they got better.

Maybe it has nothing to do with longer treatment being more effective,

it's just that more people responded so you would naturally expect more people to respond that they got better than for this group.

This is the situation where B would be relevant.

But that's not what the argument said.

It wasn't about the number of people who said they got better,

it was about the proportion out of those who responded who said they got better.

And by the way, that phrase there is important.

The proportion out of those who responded.

Because I think if you were reading these percentages as if they were about the percentages out of all the people who we asked,

regardless of whether they responded, if that's how you were interpreting this, then I could see why you might like B.

But that's not what it means.

It's about percentage out of those who responded.

And B is not relevant to that comparison.

Now, by the way, I think sometimes people pick B because they think it weakens by showing that one group was like a lot larger than the other group.

And so they think, oh, well, if the number of responses in the longer group was bigger than the number of responses in the shorter group,

doesn't that make the study less reliable?

And the answer is no.

If you had, let's say, like a million responses in the longer group,

and then you had a thousand responses in the shorter group,

and let me make clear that we're talking about responses, not the number of people who were asked.

If you had this disparity in number of responses, that's completely irrelevant,

because you can still say, okay, 36% of that number versus 20% of that number.

What explains the difference?

And no, you have no basis to say that, like, one of these is too small of a sample size to reach a conclusion.

That's not an appropriate line of reasoning either,

because you have no reason to think that any of these numbers are too small to reach a conclusion.

So ultimately, although answer choice B is very tempting,

there is nothing about it that's actually relevant to the reasoning of the argument.

It does not provide another explanation for why a higher proportion of the longer group said they got better than for the shorter group.

But take a look at answer choice C.

while those who are doing poorly tend to quit earlier.

If this is true, this does provide a potential explanation for why the longer group,

out of those who responded, had a higher proportion say they got better.

Put yourself in the shoes of a patient who is seeking help.

What C is saying is, if you don't think psychotherapy is working for you,

then you will tend to quit earlier.

So for example, let's say this is a person who is getting treatment,

and then this is month one through six,

and then this is everything after month six.

What C is indicating is that for the average person,

if therapy is not working for them, they are more likely to drop out.

They'll continue their therapy over here, and then they're more likely to drop out earlier.

Let's say they put a stop right there.

So on average, this kind of person is not really going to make it to the six-month group,

or rather to the after-six-month group.

And let me just put an unhappy face to say it's not working for them.

But if you are the kind of person for whom therapy is working, so you're very happy,

then on average, you're going to be more likely to make it into the over-six-month category.

Now of course, this doesn't happen for every single person who's unhappy,

or every single person who is happy.

But what C is saying is that on average,

the unhappy people for whom treatment is not working

will be less likely to make it into the over-six-month category.

On average, the people for whom treatment is already working,

they're going to be more likely to make it to the over-six-month category.

So then if you send out a survey to the people who got less than six months

or just six months of treatment, and ask them how therapy went for them,

your sample is going to include a higher percentage of people for whom therapy is not working

than if you asked the over-six-month group.

If you ask the over-six-month group, then a higher proportion of the people you ask

are going to be those for whom therapy is working.

So in other words, what C is suggesting is,

it might not be that the length of the therapy is actually helping people get better,

that it's more effective.

It could be that someone for whom therapy is already working

is more likely to be in the longer group.

That could be the other explanation for what's happening.

Because the people for whom therapy is not working are tending to quit

before they even make it to the over-six-month group,

you don't even get those people as part of your sample,

or at least you don't get as many of those people,

and to be precise, you don't get as much of a proportion of those people in your sample

if you just ask the over-six-month crowd.

And in case it helps, here's an analogy to show you how C is working.

Let's say you look at everybody who is a professional athlete.

And then you say, hmm, you know what?

Almost every single professional athlete is like super into their sport.

They're just super into athletics.

They love working out.

They love playing sports.

And then you look at the average person and you say, you know what?

Only a small percentage of people actually love to work out and play sports.

Imagine someone looks at that and says, oh, you know what?

I bet becoming a professional athlete causes one to start loving sports.

It increases the love you have of athletics and working out.

Answer choice C is like someone pointing out,

well, no, maybe someone becomes a professional athlete

because that kind of person, on average,

is already more athletic and into sports and working out.

That's what helped them become a professional athlete in the first place.

It's not that becoming a professional athlete

causes someone to then like to work out and to like sports.

It's the other way around.

That's the same point that answer choice C is making.

It's not that the longer treatment is causing people to get better.

It's that, on average, the kind of person

who is already getting better from treatment

will be the kind of person who is more likely

to want to get over six months of treatment.

And that's why answer choice C weakens the argument.

D is presenting another possible explanation

for the results of this study

besides the explanation that the author offers.

Let's understand why D and E are wrong now.

were more likely to feel a need to express their feelings about it

and thus to return a survey.

But the problem with D is that I don't have any reason

to think that it creates a difference between these two groups.

What D is suggesting is that within both groups,

people who were dissatisfied were more likely to respond to the survey,

but I already know that 20% say they got better here

versus 36% in the longer group.

What explains this difference?

Maybe there's like a certain percentage of people

who said that they were dissatisfied

and that they really hated what they got,

but maybe that's like 10% of this group

and then 10% of that group.

I don't have any reason to think there's a difference here.

So that's why answer choice D has no effect on this argument.

encourage their patients to receive treatment

for longer than six months.

Okay, but that's not offering any alternate interpretation

of these statistics here.

And moreover, the conclusion is trying to say

that more than six months is more effective.

So an answer that says,

oh, this is what a lot of psychologists recommend,

doesn't even go in the direction of something

that would go against the author's position.

So E, if true, does not weaken this argument.

Now I do want to note that

even if E had said the opposite and said

many psychologists encourage their patients

to get shorter than six months treatment,

that still would not weaken the argument

because we don't actually care what psychologists recommend.

What we care about is the results of this study

and whether the author's interpretation of these results

is a good one or a bad one.

Outside recommendations from other people

doesn't affect the logic of the argument

that the author gave here.

So that's a separate reason to give it a B.