Q18

[hilarykustoff]

Can someone please explain this whole game in detail? I don't even know where to begin. Thanks!

[timmydoeslsat]

This is a binary grouping game with a 2 sets of information, so you can think of this as 3D binary in a way.

This game tells us up front that we are selecting 5 of 9, which means 4 will be out.



After diagramming the rules, that is the setup I had initially.

I know that we cannot have 3 A's. I know that I have 1 D. Of the 4 that are out, I knew that we had to have 2 D's out and at least 1 A out. There is only 1 more spot left in the out column. So this means that at the very least we must have 2 P's. We can have at most 1 P out.

This is a very restrictive idea. Either we have 3 P's in or 2 P's in. What a great framing opportunity with a clear split.

I know that the 3 P's being in frame will really give me limited options, I expect to have only 1 hypothetical dealing with this idea.

However, with the 2 P's being in, I know that I have a variety of options: Either UV, UW, or VW.

I am prepared to move on to the questions if I feel that this part of the frame gets too involved. I think it is a great thing to explore, however.

The entire framing below took me about 3 minutes. You can then answer the questions in literally 2 minutes.



To accomplish this framing, do the following:

Show the 3 P's frame: UVW being in.

For the next frame, draw out 3 hypothetical sketches.

Show one with UW, one with UV, one with VW.

I also feel obligated to tell you that in all instances, I would approach the typical first hypothetical question by applying the rules 1 by 1 until one choice remains. This is always the most efficient approach.

[jkang]

i'm a bit confused. couldn't you make the inference combining rules 3 and 4 ?

A > B
A > C
therefore, B > C.

F > O
F > /W

therefore, O > /W.

i understand after trying to make a working diagram using this rule, i see that it is impossible. i found myself stuck on what should have been an easy acceptability questions (18) because of this inference. is my inference incorrect or is it the application that cancels the inference?

[ohthatpatrick]

No, you cannot make that Inference from
A --> B
A --> C

Those don't chain together.

If they did, how would you arbitrarily pick that B-->C rather than the other way around?

Consider these two statements:
play in the NFL --> male
play in the NFL --> rich

Can we infer
male --> rich
or
rich --> male
??

No.

You can only create chains when the necessary (right side) idea of one statement is the sufficient (left side) idea of another.

play in the NFL --> male
male --> have Y chromosome

we can chain to say
play in the NFL --> have Y chromosome

make sense?