Initial approach to parallel flaw/reason

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Initial approach to parallel flaw/reason

Postby izzy895 » Mon Oct 14, 2013 9:27 pm

When this question comes up do you typically diagram automatically, or do many of you attempt the reasoning in your head and match from there. I know there are some very difficult questions where it makes sense to diagram but I'm just wondering what the first approach is. Sometimes I find myself trying it in my head and then being forced to diagram which wastes a lot of time. I do have trouble extracting the formal logic sometimes (sufficient assumption is my other weak point) which is why I find myself trying to do without the diagram.

Normally I can get the diagram correctly but I am a bit. So, for the more difficult questions usually after question 14 do some of you waste no time and automatically diagram?

User avatar
Jeffort
Posts: 1896
Joined: Wed Jun 18, 2008 4:43 pm

Re: Initial approach to parallel flaw/reason

Postby Jeffort » Tue Oct 15, 2013 12:43 am

Since not every parallel reasoning question is based on conditional reasoning, you shouldn't try to force diagramming on every one. If the reasoning is conditional and you don't easily see the pattern of reasoning (such as A-->B-->C therefore A-->C) just from reading the argument and thinking it through in your head, diagramming can be helpful to better see the particular pattern of reasoning you need to parallel.

Part of the process of getting good at this type is recognizing when diagramming would be useful vs. when it may do more harm than good. Trying to force diagramming onto an argument that is not conditional typically hurts instead of helping since it distracts you from the real pattern of reasoning going on that is not actually a conditional relationship and can cause you to totally misunderstand the true reasoning pattern you need to parallel.

When there is conditional reasoning in the argument, immediately diagramming it out to see the reasoning pattern can be helpful for solving the question accurately and more efficiently as long as you are able to diagram properly without taking forever. Don't worry about being slow at diagramming in the early stages of prep, with practice you should get better and faster at doing it when needed.

A big part of prep with diagramming is gaining experience to better be able to quickly determine when it is going to help vs when it isn't right after reading the argument the first time. It takes a good amount of trial and error practice to get a better feel for when diagramming will be helpful for you vs. when it won't since it is dependent on the context, ie what type of argument you are dealing with, how quickly you recognize conditional logic when it's being used, and how good your diagramming skills actually are.

User avatar
Howl
Posts: 293
Joined: Fri Jan 20, 2012 1:11 am

Re: Initial approach to parallel flaw/reason

Postby Howl » Tue Oct 15, 2013 3:20 am

This is the quickest, most methodical way I've found.

I first ignore the premises and "match" the conclusions before anything else. In this part and other parts, the modifiers and verbs in the sentences are SUPER important. For example, if the conclusion of the stim says "therefore, x should do y," you can look for any conclusions that do not match the same stress and automatically rule them out (ex. factual statements or deviations like "therefore, x is all of y" or therefore, "x most likely will do y" will be eliminated right away, while "x ought to do y" will remain). Similarly, if the conclusion of the stim contains the modifiers "most," "some," "all," or "none," the conclusion of the correct answer choice will ALWAYS have the same modifier or a synonym of it (i.e. "not all" instead of "some," "majority" instead of "most" can be kept).

Doing that really quickly with the conclusion of the stim should eliminate at least one or two ACs right off the bat. If you're lucky, it will sometimes eliminate 3 ACs, leaving only 2 to work with.

Then I repeat the same process of matching modifiers/verbs with premises. Let's say the logic chain in the in the stim goes "All a are b. most b are c..." Then the premises in the ACs need to have the same modifiers in their logic chain. It can't go "Most a are b. All b are c..." or God forbid, "All a are b. Some b are c..." Eliminate those. Doing this will quickly eliminate another 1-2 ACs.

Sometimes, you're really unlucky and most of the ACs will exactly mirror the stim in their modifiers/verbs. THEN I diagram, so I can get a really clear sense of the logic going on in the stim. This is pretty rare though, and if I can't quickly eliminate at least 2 ACs quickly using the method above, I'll skip the question and come back to it at the end of a section.

tl;dr - match match match. Correct ACs will never contain "some" if the stim only mentions "all/most," never contain "should/ought" when the stim is a fact set, etc.

User avatar
Howl
Posts: 293
Joined: Fri Jan 20, 2012 1:11 am

Re: Initial approach to parallel flaw/reason

Postby Howl » Tue Oct 15, 2013 10:10 am

^If this is unclear, PM me and I'd be happy to explain in a bit more detail with an example.

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Tue Oct 15, 2013 10:42 am

Jeffort wrote:Since not every parallel reasoning question is based on conditional reasoning, you shouldn't try to force diagramming on every one. If the reasoning is conditional and you don't easily see the pattern of reasoning (such as A-->B-->C therefore A-->C) just from reading the argument and thinking it through in your head, diagramming can be helpful to better see the particular pattern of reasoning you need to parallel.

Part of the process of getting good at this type is recognizing when diagramming would be useful vs. when it may do more harm than good. Trying to force diagramming onto an argument that is not conditional typically hurts instead of helping since it distracts you from the real pattern of reasoning going on that is not actually a conditional relationship and can cause you to totally misunderstand the true reasoning pattern you need to parallel.

When there is conditional reasoning in the argument, immediately diagramming it out to see the reasoning pattern can be helpful for solving the question accurately and more efficiently as long as you are able to diagram properly without taking forever. Don't worry about being slow at diagramming in the early stages of prep, with practice you should get better and faster at doing it when needed.

A big part of prep with diagramming is gaining experience to better be able to quickly determine when it is going to help vs when it isn't right after reading the argument the first time. It takes a good amount of trial and error practice to get a better feel for when diagramming will be helpful for you vs. when it won't since it is dependent on the context, ie what type of argument you are dealing with, how quickly you recognize conditional logic when it's being used, and how good your diagramming skills actually are.


I see what you are saying, I normally do not try to force diagrams but for some of the more difficult stims I will run out of time and guess. When I come back to checking my answers I find that if i force a set up I can find the answer even if it isn't an exact conditional set up. For example, PT 40 S4 Q 17 I was running low on time so I did my best to reason it in my head and picked incorrectly. When I went to review I wrote out X -> Y, Alex is Y, so X. I was able to match this up with the correct answer after I had diagrammed it out.

I think this is a question that people could handle without diagramming and if I had more time I could possibly as well but normally I see a flaw question like this, attempt it and by that time I've already wasted over a minute since I didn't diagram. Once I do diagram I see the flaw clearly or the reasoning clearly. My biggest issue is the time wasted if I try to not diagram.

I'd say diagramming and conditional logic are my weakest areas, I am in the process of drilling to refine some of these things. Would I just be better off practicing to recognize conditional vs. non-conditional stims? Or maybe work on trying to diagram first since I end up having to for most of the time.

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Tue Oct 15, 2013 10:46 am

Howl, your method is exactly what I've been attempting to do more of now. I have found it extremely helpful in certain questions since I am trying to do that process of elimination a bit more now. There was actually a difficult stimulus for a parallel question I just did where in the end you could have eliminated each of the Incorrect AC's due to conclusion mismatch that I didn't see. When you do eliminate to like two AC's do you normally diagram here?

User avatar
Howl
Posts: 293
Joined: Fri Jan 20, 2012 1:11 am

Re: Initial approach to parallel flaw/reason

Postby Howl » Tue Oct 15, 2013 11:52 am

izzy895 wrote:Howl, your method is exactly what I've been attempting to do more of now. I have found it extremely helpful in certain questions since I am trying to do that process of elimination a bit more now. There was actually a difficult stimulus for a parallel question I just did where in the end you could have eliminated each of the Incorrect AC's due to conclusion mismatch that I didn't see. When you do eliminate to like two AC's do you normally diagram here?


Sometimes I do, sometimes I don't. It depends on the stimulus - sometimes, matching up the premises' verbs/modifiers is enough to get me to eliminate down to 1 AC. This happens more with all/some/most/none kinds of stims, because the wording match in both the premise and the conclusion has to be basically exact. But if the stim has a bit of confusing conditional reasoning and there are still 2 answer choices left, I'll diagram the stim really quickly/loosely (e.g. "flaw: A -> B, C -> B, so C -> A") and compare that to the two remaining answer choices. If the wording is really tough, then I'll occasionally diagram the 2 remaining ACs as well (no shame in being thorough), but with practice, identifying the right match becomes easier so that some ACs you'll be able to eliminate without necessarily diagramming the ACs.

Also, watch out for LSAT tricks of bringing new info in - for example, if the stim's reasoning concerns itself with just car engines, the correct AC must also just stay on 1 topic and shouldn't bring in a second, new element to complete its reasoning, etc. This is the next best way to eliminate wrong ACs, I found.

I hope these tips help - I used to be SO slow at parallel reasoning/flaw, and since I've used this method it's cut down my time a lot, so I thought I'd share.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: Initial approach to parallel flaw/reason

Postby bp shinners » Tue Oct 15, 2013 12:24 pm

Another important note - Parallel Reasoning and Parallel Flaw are different question types with different requirements for correct answers.

Parallel questions need all relevant elements of the stimulus to match up in the answer - number of terms, logical force, conclusions, etc... They have a very high burden for a correct answer.

Parallel flaw questions only need to flaw to match up. While this might mean the logical force, number of terms, etc... match up, they don't have to if those elements are not relevant to the flaw.

So if I had an argument stating:
A+B -> C
C
______________
A+B

And an answer choice stating:
E -> F
F
__________
E

That would be correct for a parallel flaw question, but not a parallel question.

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Tue Oct 15, 2013 1:04 pm

Howl wrote:
izzy895 wrote:Howl, your method is exactly what I've been attempting to do more of now. I have found it extremely helpful in certain questions since I am trying to do that process of elimination a bit more now. There was actually a difficult stimulus for a parallel question I just did where in the end you could have eliminated each of the Incorrect AC's due to conclusion mismatch that I didn't see. When you do eliminate to like two AC's do you normally diagram here?


Sometimes I do, sometimes I don't. It depends on the stimulus - sometimes, matching up the premises' verbs/modifiers is enough to get me to eliminate down to 1 AC. This happens more with all/some/most/none kinds of stims, because the wording match in both the premise and the conclusion has to be basically exact. But if the stim has a bit of confusing conditional reasoning and there are still 2 answer choices left, I'll diagram the stim really quickly/loosely (e.g. "flaw: A -> B, C -> B, so C -> A") and compare that to the two remaining answer choices. If the wording is really tough, then I'll occasionally diagram the 2 remaining ACs as well (no shame in being thorough), but with practice, identifying the right match becomes easier so that some ACs you'll be able to eliminate without necessarily diagramming the ACs.

Also, watch out for LSAT tricks of bringing new info in - for example, if the stim's reasoning concerns itself with just car engines, the correct AC must also just stay on 1 topic and shouldn't bring in a second, new element to complete its reasoning, etc. This is the next best way to eliminate wrong ACs, I found.

I hope these tips help - I used to be SO slow at parallel reasoning/flaw, and since I've used this method it's cut down my time a lot, so I thought I'd share.


These tips will most definitely help I actually just started to implement this strategy. I just did a pt where there was a flaw and a reasoning towards the end where I was only able to eliminate 1 ac on only 1 question. I got them both right without diagramming actually but I just didn't feel too confident. When you can't eliminate do you try to do the reasoning in your head?

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Tue Oct 15, 2013 1:09 pm

bp shinners wrote:Another important note - Parallel Reasoning and Parallel Flaw are different question types with different requirements for correct answers.

Parallel questions need all relevant elements of the stimulus to match up in the answer - number of terms, logical force, conclusions, etc... They have a very high burden for a correct answer.

Parallel flaw questions only need to flaw to match up. While this might mean the logical force, number of terms, etc... match up, they don't have to if those elements are not relevant to the flaw.

So if I had an argument stating:
A+B -> C
C
______________
A+B

And an answer choice stating:
E -> F
F
__________
E

That would be correct for a parallel flaw question, but not a parallel question.


That is a good point, thank you. How about for a reasoning question if you have :

X -> Y
X
__________
Y

With an answer

Y->X
Y
_________
X

Would that be an acceptable answer? I may not be explaining that correctly I recently saw a question similar to that same logical structure but different placement of variables.

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: Initial approach to parallel flaw/reason

Postby bp shinners » Wed Oct 16, 2013 4:44 pm

izzy895 wrote:That is a good point, thank you. How about for a reasoning question if you have :

X -> Y
X
__________
Y

With an answer

Y->X
Y
_________
X

Would that be an acceptable answer? I may not be explaining that correctly I recently saw a question similar to that same logical structure but different placement of variables.


Since the AC will not use the same terms, you could just replace X and Y in your second example and it would be the same as the first.

To answer the implicit question, I believe, no - the order doesn't have to be the same; just the logic.

jd2121
Posts: 28
Joined: Fri Sep 13, 2013 2:24 am

Re: Initial approach to parallel flaw/reason

Postby jd2121 » Thu Oct 17, 2013 12:16 am

Heres what I do:

1) read through the argument twice, consecutively.
2) go through each answer choice one time, instinctively eliminating those which dont "feel" right; this usually narrows it down to two choices.
3) i then go through and compare the argument structures of the two attractive choices to that of the prompt.
4) choose the one that is parallel

this may seem time consuming, but it honestly doesn't take longer than a minute. the idea of diagramming this stuff seems grossly excessive given the simplicity of the arguments. dont underestimate your ability to mentally retain a basic syllogism (which is essentially what you'll run into a majority of the time).

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: Initial approach to parallel flaw/reason

Postby bp shinners » Thu Oct 17, 2013 12:15 pm

jd2121 wrote:Heres what I do:

2) go through each answer choice one time, instinctively eliminating those which dont "feel" right; this usually narrows it down to two choices.

the idea of diagramming this stuff seems grossly excessive given the simplicity of the arguments. dont underestimate your ability to mentally retain a basic syllogism (which is essentially what you'll run into a majority of the time).


I think you might be underestimating your ability relative to others. I've worked with plenty of students, and eliminating things that don't "feel" right usually results in either everything or nothing being eliminated. And many, many students can't keep track of this stuff in their heads throughout a stimulus and 5 ACs.

If you can do it, then this approach would be a great way. I just don't think most people prepping for the LSAT are going to find it effective.

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Thu Oct 17, 2013 1:19 pm

The 'feels right' approach I've used with success but I don't like to get answers correct by instinct because I worry I won't be consistent. I actually finished the parallel flaw packet from Cambridge and now I can really see the difference between some regular parallel reasoning questions. Like you said above bp, about the not as high burden of matching in ac's.

While we are on topic of diagramming I have a question for some high scorers like yourself. I recently took a preptest that had two difficult sufficient questions that I got right without diagramming. I like to double check reasonings by going over to manhattan lsat and mostly all and the advice was to solve it via diagramming. Would I be smart to make myself diagram? I wasn't 100% confident with my answer but when I saw the explanation through conditional logic it becomes very clear. I am weak with this but a lot of the times I can get them right without doing so. Do you like to use conditional logic more often than not on more difficult suff. Assumptions?

bp shinners
Posts: 3091
Joined: Wed Mar 16, 2011 7:05 pm

Re: Initial approach to parallel flaw/reason

Postby bp shinners » Fri Oct 18, 2013 12:59 pm

izzy895 wrote:While we are on topic of diagramming I have a question for some high scorers like yourself. I recently took a preptest that had two difficult sufficient questions that I got right without diagramming. I like to double check reasonings by going over to manhattan lsat and mostly all and the advice was to solve it via diagramming. Would I be smart to make myself diagram? I wasn't 100% confident with my answer but when I saw the explanation through conditional logic it becomes very clear. I am weak with this but a lot of the times I can get them right without doing so. Do you like to use conditional logic more often than not on more difficult suff. Assumptions?


I break sufficient assumption questions down into 2 types:
1) There's a new term in the conclusion;
2) There's not a new term in the conclusion.

For (1), I almost never diagram or, really, think much about it. When that happens, the new term absolutely has to show up in some form in the conclusion, and in the same condition (suff/nec) as it does in the conclusion. This trivializes these sufficient assumption questions, and it applies to a large majority of them. Seriously, it almost feels like cheating. This works for non-diagrammable ones as well.

For (2), even if I'm pretty sure of an answer, I diagram it out. There's going to be a gap between two premises, and it's easy for you to get the suff/nec conditions turned around in your head. You'll actually probably get through it faster, too, since you won't have to think. It never feels that way, because you feel like writing is eating up time. But thinking also eats up time, though it doesn't feel like it because your brain processes time differently when it's thinking than when it's doing something rote.

izzy895
Posts: 45
Joined: Tue Sep 17, 2013 4:59 pm

Re: Initial approach to parallel flaw/reason

Postby izzy895 » Fri Oct 18, 2013 3:32 pm

bp shinners wrote:
izzy895 wrote:While we are on topic of diagramming I have a question for some high scorers like yourself. I recently took a preptest that had two difficult sufficient questions that I got right without diagramming. I like to double check reasonings by going over to manhattan lsat and mostly all and the advice was to solve it via diagramming. Would I be smart to make myself diagram? I wasn't 100% confident with my answer but when I saw the explanation through conditional logic it becomes very clear. I am weak with this but a lot of the times I can get them right without doing so. Do you like to use conditional logic more often than not on more difficult suff. Assumptions?


I break sufficient assumption questions down into 2 types:
1) There's a new term in the conclusion;
2) There's not a new term in the conclusion.

For (1), I almost never diagram or, really, think much about it. When that happens, the new term absolutely has to show up in some form in the conclusion, and in the same condition (suff/nec) as it does in the conclusion. This trivializes these sufficient assumption questions, and it applies to a large majority of them. Seriously, it almost feels like cheating. This works for non-diagrammable ones as well.

For (2), even if I'm pretty sure of an answer, I diagram it out. There's going to be a gap between two premises, and it's easy for you to get the suff/nec conditions turned around in your head. You'll actually probably get through it faster, too, since you won't have to think. It never feels that way, because you feel like writing is eating up time. But thinking also eats up time, though it doesn't feel like it because your brain processes time differently when it's thinking than when it's doing something rote.


That's very helpful I'm going to try to look at them this way. I think a lot of times I tend to go off what feels right in the answer since I'm not confident enough in my diagramming. My evaluation process definitely shortens. I'm making my way over to drilling these so I will work on what you suggested. Much appreciated.

User avatar
sashafierce
Posts: 454
Joined: Sun Aug 04, 2013 11:44 am

Re: Initial approach to parallel flaw/reason

Postby sashafierce » Sun Nov 24, 2013 12:31 pm

The information in this thread is very helpful. Thanks guys :D




Return to “LSAT Prep and Discussion Forum”

Who is online

Users browsing this forum: Google [Bot], TakeItToTrial and 9 guests