Verbal Reasoning Question - can you get it right?

[john kane]

Ive been applying to some firms. So far all six have accepted my application, but 4 of them want me to do psychometric tests. Ive only done one firm's test and i failed. I think it was due to the verbal reasoning.

Here is an example of why i find them so difficult

Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.

Question: Children who have problems sleeping are more likely to suffer from abdominal pain (yes i know its not a question, more of another statement, but you know what i mean)

Please give reasons for your answer

[NT!]

Yeah if you can't see that this is insufficient info you're kinda dumb.

[man]

this statement is false because it's really obviously false.

[guids]

I voted false.

[john kane]

please explain your reasons given the text.

thanks

[guids]

abdominal pain in conjuntion with sleeping problems is related to emmotional disturbance, so with out the emmotional problems part thrown in there, the second statement would be false. But, also, inssuficient info could be right, because I dont knwo the stats on how abdominal pain (ie normal tummy aches) relate directly to sleeping patterns.

[Ra_]

Using only the information provided there is not enough information. Only by making assumptions can you come up with another answer.

[dknightx]

its insufficient because the statement only speaks of abdominal pain as a symptom of emotional disturbance. That is to say, if a child has abdominal pain, AND has a phobia and sleeping disorder, we are looking at a more likely case of emotional disturbance ... it says NOTHING about the relationship of abdonminal pain and sleep disorders ... so not enough info. It COULD be true and it COULD be false, we dont know.

[goodsamaritan]

<font color="red"> Insufficient Info </font>

[diddle]

[ QUOTE ]


Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.


[/ QUOTE ]

Insufficient

[dknightx]

[ QUOTE ]
[ QUOTE ]


Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.


[/ QUOTE ]

Insufficient

[/ QUOTE ]

its not insufficient because of the "may" ... even if you replace the "may be" with "always is", it is still insufficient.

[fiskebent]

It's false. Nowhere do they say that children with abdominal pain have a higher likelihood of sleep disorders.

[jstnrgrs]

[ QUOTE ]
It's false. Nowhere do they say that children with abdominal pain have a higher likelihood of sleep disorders.

[/ QUOTE ]

Nowhere do they say that children with abdominal pain don't have a higher likelihood of sleep disorders. There is insuficient info.

[NT!]

[ QUOTE ]
It's false. Nowhere do they say that children with abdominal pain have a higher likelihood of sleep disorders.

[/ QUOTE ]

nowhere do they say that they don't, either.

this is really [censored] easy, even i can get it and i'm a [censored] mod for chrissakes

[offTopic]

[ QUOTE ]

Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.

Question: Children who have problems sleeping are more likely to suffer from abdominal pain (yes i know its not a question, more of another statement, but you know what i mean)

Please give reasons for your answer

[/ QUOTE ]

I chose insufficient information.

You can't logically get from:

"A" may be a sympton of "B" especially with "C".

to:

If "B", then, likely "A".

without knowing more info.

[El Diablo]

All,

I can't believe people are arguing or confused about the answer.

[Mark Collins]

Insuffizle blizzle

[onthebutton]

Insufficient.

And it's easy.

[El Diablo]

jk,

"Ive only done one firm's test and i failed.
Here is an example of why i find them so difficult"

My advice is to focus on the 2 firms that don't give tests, because if you think this q was difficult, then you're pretty much guaranteed to fail.

[Maulik]

[ QUOTE ]
All,

I can't believe people are arguing or confused about the answer.

[/ QUOTE ]

[Maulik]

jk,

Have you taken introductory statistics classes?

[man]

I changed my mind. the question is whether statement 2 is correct based on the information in 1. so you're starting with zero information and you have to build up to statement 2, based on what's provided in 1. since we don't have enough information, it's insufficient information.

if they were asking if statement 1 implies statement 2, the answer would clearly be no.

[Shajen]

[ QUOTE ]
All,

I can't believe people are arguing or confused about the answer.

[/ QUOTE ]

Sklansky, is that you?

[Miggo]

I'm sticking with true, only because of the phrase "more likely".

[john kane]

dear god im retarded at these verbal reasoning. i went for true. and yes the answer is insufficient info.

my reasoning was that emotional disturbance increases the likelyhood of abdominal pains. the likelyhood is furthur increased by having sleep problems. hence children with sleeping problems are more likely to have sleeping problems.

ive been reading over people's replies. many thanks for the replies.

ill probably be putting up quite a few more of these over the next couple of days. i plan to do the other 3 tests on thursday afternoon.

[Aces McGee]

[ QUOTE ]
Ive been applying to some firms. So far all six have accepted my application, but 4 of them want me to do psychometric tests. Ive only done one firm's test and i failed. I think it was due to the verbal reasoning.

Here is an example of why i find them so difficult

Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.

Question: Children who have problems sleeping are more likely to suffer from abdominal pain (yes i know its not a question, more of another statement, but you know what i mean)

Please give reasons for your answer

[/ QUOTE ]

This is basic LSAT logical reasoning stuff, no?

-McGee

[fslexcduck]

yeah this isn't hard

[Inthacup]

[ QUOTE ]
I'm sticking with true, only because of the phrase "more likely".

[/ QUOTE ]

More likely than who?

[john kane]

[ QUOTE ]
jk,

"Ive only done one firm's test and i failed.
Here is an example of why i find them so difficult"

My advice is to focus on the 2 firms that don't give tests, because if you think this q was difficult, then you're pretty much guaranteed to fail.

[/ QUOTE ]

given my half decent intellect i should be able to grasp these bastard verbal reasoning tests given a lot of people so far have stated that this question is easy.

i just need to train my mind to think the correct way.

the company who told me to sod off was deloittes. the 2 firms are E+Y and pwc both id be happy to work for, kpmg, bdo and grant thornton are the other 3 firms who i have to do tests for.

i know i can do this, i just need to practice practice practice and learn learn learn.

[john kane]

oops, i assumed b_tard would be censored.

[offTopic]

[ QUOTE ]
jk,

"Ive only done one firm's test and i failed.
Here is an example of why i find them so difficult"

My advice is to focus on the 2 firms that don't give tests, because if you think this q was difficult, then you're pretty much guaranteed to fail.

[/ QUOTE ]

Question Applicants who are pretty much guaranteed to fail will focus on the 2 firms that don't give tests.

[Duke]

I'll guess that (if you got it wrong) you said false. If you said true, well, I can't help you there.

It's insufficient information because they're two unrelated statements that bear no actual resemblance to each other (just because they both deal with tummy aches doesn't mean that they have information about them that is relevant to the other).

To make it a little more clear, the way this could be a "false" answer would be if their were a question on the order of: "Does A imply B?" In that case it'd be false, since it doesn't imply it.

Since they're just 2 disjoint statements in this case, you can take the 'insufficient information' answer. You can't say false, because it could very well be true. 'A' says nothing about the veracity of 'B.'

To simplify it, if you had 2 statements, and 1 was a "given", and 2 was a "possibility" :

1. Bob had potatoes for lunch.
2. Bob likes potatoes.

Does 1 imply 2? Nope.
Is 2 true? Insufficient information.
Might 2 be true? Yep.

[john kane]

[ QUOTE ]
jk,

Have you taken introductory statistics classes?

[/ QUOTE ]

not at gcse (16 year old level) but i got top grade for maths, and laid the smackdown at A Level (18 years old), and did a maths stats course at university (22 years old) and did very well in that.

but i much prefer numbers, not this written crap.

[fnord_too]

[ QUOTE ]


Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.

Question: Children who have problems sleeping are more likely to suffer from abdominal pain (yes i know its not a question, more of another statement, but you know what i mean)



[/ QUOTE ]

Because a bunch of people are saying false, which is the most wrong answer, lets analyse this a bit

Emotionally Disturbed = ED
Abdominal Pain = AP
Sleep Disorder = SD

I am going to assume that the OP is from memory and that were tighter in their language on the actual question (for instance, by problems sleeping they mean the specific disorders that are linked to being emotionally disturbed)

We are told that P(AP|ED) &gt; P(AP), otherwise it would not be a symptom (like breathing is not a symptom of having cancer, even though all cancer patients breath).

We are told that P((AP&amp;SD)|ED) &gt; P(AP&amp;SD) by more than P(AP|ED) &gt; P(AP)

That is, AP &amp; SD are not as tightly correlated when ED is not involved.

What we are never told anything about is P(ED|SD).

It could easily be that SD does increase the liklihood of ED, but that is an assumption (I do not think you can get there from bayesean probability given what we have been told without knowing P(SD)). In fact, I am pretty sure if I tried I could come up with probability distributions that would make the statement both true and false, but that a random selection of probability distributions that met the criteria would make it true (the last is just my gut feeling).

The reason "false" is more incorrect that "true" is that there is absolutely nothing to even lead you to the conclusion that this is false. That is, one could say, something like "well, we know children with ED are more likely to have SD (pretty sure this is incorrect, that is that there is an assumption there), so someone with SD is more likely to have ED (again, incorrect) so there fore they are more likely to have AP". You can in no way go from the statements to arrive at "Therefore they are equal or less likely to suffer from AP".

Not being able to prove something is true from the information given does not make it false.

[NLSoldier]

[ QUOTE ]
All,

I can't believe people are arguing or confused about the answer.

[/ QUOTE ]

Saying this w/o giving your answer is incredibly weak.

[NT!]

[ QUOTE ]
ill probably be putting up quite a few more of these over the next couple of days. i plan to do the other 3 tests on thursday afternoon.

[/ QUOTE ]

take it to SMP or POG or something then, this is not OOT material and we'll just make fun of you

[fnord_too]

[ QUOTE ]
[ QUOTE ]
jk,

Have you taken introductory statistics classes?

[/ QUOTE ]

not at gcse (16 year old level) but i got top grade for maths, and laid the smackdown at A Level (18 years old), and did a maths stats course at university (22 years old) and did very well in that.

but i much prefer numbers, not this written crap.

[/ QUOTE ]

Then translate the written crap into numbers. The ability to process words into numbers is a major stratifier in academics and life. You are not going to get many problems, especially in real life, that are set up for you.

[john kane]

duke; thanks a lot for that response, i need to keep reading responses like that to get myself sorted.

[fnord_too]

[ QUOTE ]
[ QUOTE ]
ill probably be putting up quite a few more of these over the next couple of days. i plan to do the other 3 tests on thursday afternoon.

[/ QUOTE ]

take it to SMP or POG or something then, this is not OOT material and we'll just make fun of you

[/ QUOTE ]

SMP is where it belongs. Or Probability.

[runner4life7]

i just lost some hope in humanity after seeing people get this wrong. This is a 2p2 forum, not an oprah's book club. You would think that people would be smarter than this, but then you would of course be wrong.

[ahnuld]

becuase they use the term conjunction, sleeping problems alone have no relation to A. I have a midterm in sentential derivations tomorrow, so I hope I got this right.

[man]

[ QUOTE ]
becuase they use the term conjunction, sleeping problems alone have no relation to A. I have a midterm in sentential derivations tomorrow, so I hope I got this right.

[/ QUOTE ]
make sure you answer the questions tomorrow!

[El Diablo]

NL,

"Saying this w/o giving your answer is incredibly weak."

Wow, I'm shocked that you don't think the answer to this is about as trivially obvious as possible. This is like an easy question on the PSAT. Wow, honestly, I'm really shocked that you thought there was anything at all to discuss here.

[ahnuld]

the statement we have to work with is if ( A and B) then P.

Then question asks if B and P are true. This is false since B and A need to occur for P to occur. If you guys who say insufficent info mean that we need to know the thurth value of B then I agree.

[Maulik]

jk,

An elementary logic book will help. So will more reading.

[El Diablo]

jk,

I don't think there's any problem with you posting more here. Just do them in this thread if you do.

[bwana devil]

duke gave great reasoning.

in addition you could think that perhaps 5% of the nonemotional disturbance population has abdominal pain. also 5% of the emotional disturbance population has abdominal pain.

it just so happens that in the emotional disturbance kids, this is a symptom of abdominal pain.

in this example kids w/ emotional disturbance (sleep disorders) are no more likely than nonemo kids to have abdominal pain. because the numbers are made up (and unlikely the same) we need more info.

[ZeeJustin]

I do really poorly on these questions for different reasons. They are often worded in a way, that you HAVE to make assumptions, yet the whole point of many of the questions is to say, "you can't assume that".

First of all, I hate the fact that no real instructions are given, although this may just be the OP's fault. My first assumption will be that the directions are, "Based on the Statement alone and no other outside knowledge, what conclusion follows?"

Take the word, "May" for example. Is there ANY difference between saying that "Papercuts may be a symptom of emotional disturbance."

The word may expresses possibility. If we take the word literally, there is insufficient info because statement #1 is meaningless in the sense that it draws no actual conclusion. It only states a possibility.

Maybe this is why I do poorly on these tests. I will assume the statement is unintentionally poorly worded. I will change it to say, "A non-zero percent of the time, emotional disturbance will cause Abdominal pain, especially if there is also trouble sleeping."

Also, are we to assume everything else is equal? If so, we can make some logical conclusions.

1) Children w/ emotional disturbance are more likely to have Abdominal Pain.

2) According to the statement, the word especially directly implies that children w/ ED (emotional disturbance) + SD (Sleep disorder) are more likely to have AP than children with ED but not SD. Obviously someone is more likely to have ED + SD if we already know they have SD (even if we are completely informationless about whether or not they have ED).

So now let's look at the question. It asks, if SD, is AP more likely?

If SD, then SD +ED is more likely.

If SD+ED, then AP is more likely. Therefore, the answer is true.

If I were actually taking the test, I would go through this reasoning, and still be very unsure. Which of these assumptions am I supposed to make?

Technically, it is possible that if someone has SD, they are for some reason less likely to ever develop ED. I ignored this with my assumption that, "everything else is equal".

So on one hand, that implies that we don't have enough information.

On the other hand, by using the phrase, "more likely" the question is worded in a way that we don't even need all the information.

Look at it this way: A gambler would certainly place a bet given the above information, because all he cares about is the likelihood. At the same time, it is possible that because of information we don't have, that it is not true at all.

Despite all this reasoning, I am pretty confident the answer "should" be true." Given the nature of the test, and that it is considered a verbal reasoning test, I would answer "Insufficient Information" anyway. I would do this because of the statistical nature of the gambling logic I used, and I can not see that kind of logic being encouraged in a test like this.



I know this is long winded, but I'd appreciate it if someone would tell me which of these assumptions I'm supposed to make, which I'm not, and why. Yes, I realize I'm probably over thinking this, but I must be sick in the head because I can't NOT overthink it.

[mason55]

[ QUOTE ]
jk,

I don't think there's any problem with you posting more here. Just do them in this thread if you do.

[/ QUOTE ]

Agree; maybe we'll end up with another thread like when jason_t smashed that vase or when that plane tried to take off from the moving runway.

[Maulik]

mason,

Do you have links? That [censored] sounds hilarious.

[Duke]

[ QUOTE ]
[ QUOTE ]
jk,

I don't think there's any problem with you posting more here. Just do them in this thread if you do.

[/ QUOTE ]

Agree; maybe we'll end up with another thread like when jason_t smashed that vase or when that plane tried to take off from the moving runway.

[/ QUOTE ]

The plane one was fun.

Let's re-do that one but make the plane a car with wings/glider thing.

[ZeeJustin]

Instructions: Based on the Statement alone, and no other information, what conclusion do you draw? (answer this as you would if you saw it on an LSAT test)

Statement: The best basketball player of all time currently plays for the Lakers.

Conclusion: The Lakers are more likely to win more than 50% of their games this season than they are to lose more than 50% of their games this season.

[fslexcduck]

i don't understand how anyone could ever even think of positing that there is enough information here.

do you not see how you jumped like 8 logical steps here? important information being what the makeup of the rest of the team is, how team makeup influences chance of winning, etc...

[dknightx]

zj, the reason its not enough info is that we can't assume that there is a relationship between "having the best player" and "winning games".

[ZeeJustin]

Reply: Question #2, this time it's a real life question.

There is a sport you have never heard about called baseketball. Your friend likes it, explains the concept to you, and the only thing he tells you is that the best baseketball player in the world plays for the Gyros.

Your crazy gambly foreign friend from another country is looking for action. You are 100% sure he has never heard of baseketball until you told him about it 5 minutes ago. Since you told him about it, he assumes you are knowledgeable about it, and understands (yet accepts) that you will have an edge in any bet you place.

He tells you he likes Oranges, and will bet on them to win any game this week based purely on his love of the fruit. The Oranges play the Gyros in 2 minutes.

[fslexcduck]

try this one:

Instructions: Based on the Statement alone, and no other information, what conclusion do you draw? (answer this as you would if you saw it on an LSAT test)

Statement: The best basketball player of all time currently plays for the Lakers. Every team that has had the best basketball player on it in the past 50 years has won more games than it has lost.

Conclusion: The Lakers are more likely to win more than 50% of their games this season than they are to lose more than 50% of their games this season.

[ZeeJustin]

Duck and DK,

I realized most people will answer the 3rd option to that question (as did I). But to my next question, I think the 1st option is the best, despite the questions essentially being the exact same.

The only difference is one is in test form, and the other is a real life scenario. Why should the answers be different?

[fslexcduck]

do you not see how likelihood is different than certainty ?

of course you would bet on the team, because you would assume all other things to be equal since you have no other information stating otherwise. so likelihood is they are better.

but you can't say with certainty for the same reason, you have no information stating that it is one way or the other.

in your original question, you asked to give an affirmative statement that something is true. in the second, you asked "would we bet on it?" those are two very, very different questions.

[Razor]

[ QUOTE ]
I realized most people will answer the 3rd option to that question (as did I). But to my next question, I think the 1st option is the best, despite the questions essentially being the exact same.


[/ QUOTE ]

In your second question I voted both Yes and No.

[CrazyEyez]

ZJ #2: I bet on the Gyros. They have the best player ever =&gt; increased chance they have overall better than average players. Better than average players = increased chance you have a better than average team. Oranges are a random/average team.

I don't know if this is analagous to the OP and therefore increases the likelihood that I'm an idiot, but I did manage a 1430 on my SATs.

[ahnuld]

Duck, not sure how you wanted to word your example, but this is how I read it.

A Statement
If A then P Statement

Therefore A Conclusion.

Ya, seems true to me. A = Team with best basketball player
P = Likely to finish with a winning record

[dknightx]

fslexcduck, you question is STILL not enough info (right?). The reason is, you can not assume that having the best player is a DIRECT cause for having won more games than lost ... it could be a completely unrelated factor.

[fslexcduck]

ahnuld,

i made that one especially tricky. the point is, it is still not enough information.

just because history says that it is more likely than not that the best basketball player will lead to a winning record, we don't know for sure.

imagine the following scenario: every year, the team has been 4 average players + all star. this year the team w/ best player is 4 10 yr old midget children + all star. would we still say they have a &gt;50% chance of winning record?

[fslexcduck]

[ QUOTE ]
fslexcduck, you question is STILL not enough info (right?). The reason is, you can not assume that having the best player is a DIRECT cause for having won more games than lost ... it could be a completely unrelated factor.

[/ QUOTE ]

causation v correlation, bingo

[dknightx]

[ QUOTE ]
Duck and DK,

I realized most people will answer the 3rd option to that question (as did I). But to my next question, I think the 1st option is the best, despite the questions essentially being the exact same.

The only difference is one is in test form, and the other is a real life scenario. Why should the answers be different?

[/ QUOTE ]

ZJ:

The thing is, I in the "real life" scenario, I am making an assumption ... that assumption is having the "best" player means a higher chance of winning ... but we have no reason to know that 100%. It may make sense to us logically, but what if baseketball has this rule which says "if you have the best player in the world, you automatically lose by forfiet". That could be the case! But yeah, i'd bet on the gyros cause i love gyros and i love to gamble.

[CrazyEyez]

[ QUOTE ]
do you not see how likelihood is different than certainty ?

of course you would bet on the team, because you would assume all other things to be equal since you have no other information stating otherwise. so likelihood is they are better.

but you can't say with certainty for the same reason, you have no information stating that it is one way or the other.

in your original question, you asked to give an affirmative statement that something is true. in the second, you asked "would we bet on it?" those are two very, very different questions.

[/ QUOTE ]
Hasn't every question in this thread been about likelihood?

[ahnuld]

yeah the way my post interpreted was "every team with the best player WILL win" which is very different in logic then has won.

[man]

[ QUOTE ]
[ QUOTE ]
fslexcduck, you question is STILL not enough info (right?). The reason is, you can not assume that having the best player is a DIRECT cause for having won more games than lost ... it could be a completely unrelated factor.

[/ QUOTE ]

causation v correlation, bingo

[/ QUOTE ]
so you're saying that you could assemble every possible statistic that's been an indication of a win, and that still wouldn't be enough right?

[fslexcduck]

man,

right.

[snagglepuss]

zeejustin,

how have you scored on the sat/act/lsat/mcat or whatever, if you have taken any of them?

john kane,

post more questions.

[dknightx]

statistics never give us 100% accurate causation or correlation ... they just give us levels of certainty.

[CrazyEyez]

[ QUOTE ]
man,

right.

[/ QUOTE ]
theory: Having the best player ever makes it more likely you will win.

scenario A) Every team in history that had the best player won more than half it's games.
scenario b) 60% of teams in history that had the best player won more than half their games.

I'm supposed to believe that the theory is equally likely to be true whether scenario A or B takes place?

[SNOWBALL]

I love this thread. I used to think I was just much smarter than the general public, but only somewhat above average for 2p2. Now I know that I'm much smarter than most 2p2ers also. /images/graemlins/smile.gif

[fslexcduck]

crazy,

you're misinterpreting the question ZJ posited.

your phrase: Having the best player ever makes it more likely you will win

his phrase: Having the best player ever makes it more likely that you will win than that you will lose

yours: yes
his: we don't know enough

yours: qualitative
his: quantitative

i'm probably being really unclear... does everyone else see what i'm saying or am i just insane?

[danzasmack]

I took one of these tests and they are not just about knowledge, decision making, but it depends on the test. Seeing as how you are going to be taking a bunch I figured I'd fill you in.

I read some of the replies (to the original question) before reading the question and doubted myself, because I felt like I would be dumb for not knowing the answer. Don't think this isn't the factor in a test. You have to be confident when going in. I guarantee you they know how much time you spent on each question. Plus, reading the original question with a clear head, it is super easy. But if you are nervous/head is clouded you want to put true because you are not sure of yourself.

For example, when I took the test they gave you pattern recognitions. I had 3 minutes to do something like 30. I was flying through them, no WAY could I have answered them faster without knowing the answers and I still didn't finish. The test was designed that way to fluster me. They do things like ask you the same question 20 times. The job it was for was on a trade desk at a hedge fund and my mornings would basically be the same thing every day so the test was designed to see if I would snap and hate it.

Basically you have to go into these things with confidence and be yourself. Nobody sees if you pass or fail and most of the time it benefits you because the results are apparently scarily good at determining whether a person fits the traits you want or not.

[ZeeJustin]

[ QUOTE ]
do you not see how likelihood is different than certainty ?


[/ QUOTE ]

The question where you answered "not enough information" was about likelihood, was it not?

The phrase "more likely" seems pretty unambiguous to me.


Here's the conclusion that I'm drawing:

In a Verbal Reasoning test, we are supposed to assume (yet never told to assume) that the phrase "more likely" has a different meaning than it would in any other scenario.

[ZeeJustin]

[ QUOTE ]
zeejustin,

how have you scored on the sat/act/lsat/mcat or whatever, if you have taken any of them?

john kane,

post more questions.

[/ QUOTE ]

Only taken the SAT, and I scored very well (I have a feeling if I give an actual number it will turn this thread into a flame war).

[fslexcduck]

zeej,

no... you said "x means they are more likely to win than to lose" which is different than saying "x means they are more likely to win"

let's say x is something that increases probability of something happening from 10% to 30%. you could say "x means it is more likely to happen" but you could not say "x means it is more likely to happen than not to happen."

[dknightx]

theres a difference between:

Every team in the past that has had the best player was more likely to win than lose.

and

If you have the best player YOU ARE more likely to win than lose.

[CrazyEyez]

[ QUOTE ]
[ QUOTE ]
do you not see how likelihood is different than certainty ?


[/ QUOTE ]

The question where you answered "not enough information" was about likelihood, was it not?

The phrase "more likely" seems pretty unambiguous to me.


Here's the conclusion that I'm drawing:

In a Verbal Reasoning test, we are supposed to assume (yet never told to assume) that the phrase "more likely" has a different meaning than it would in any other scenario.

[/ QUOTE ]
This is precisely what I gather from this thread.
/images/graemlins/confused.gif

[chesspain]

[ QUOTE ]
...that plane tried to take off from the moving runway.

[/ QUOTE ]

Link me, please.

[ZeeJustin]

[ QUOTE ]
let's say x is something that increases probability of something happening from 10% to 30%. you could say "x means it is more likely to happen" but you could not say "x means it is more likely to happen than not to happen."


[/ QUOTE ]

That's only true because your line is drawn at 30%. Mine was at 50%.

[danzasmack]

Can someone post some of these questions (like the bball one) where the answer is True or False? I think that would make this much clearer.

[fslexcduck]

[ QUOTE ]
[ QUOTE ]
let's say x is something that increases probability of something happening from 10% to 30%. you could say "x means it is more likely to happen" but you could not say "x means it is more likely to happen than not to happen."


[/ QUOTE ]

That's only true because your line is drawn at 30%. Mine was at 50%.

[/ QUOTE ]

dude, wtf. my point is only that those two statements are very different statements, and i was using an example to illustrate why. you have been conflating those statements. you likely did it bc you drew your line at 50% thus they were easy to conflate, w/e.

[TiK]

[ QUOTE ]
[ QUOTE ]
...that plane tried to take off from the moving runway.

[/ QUOTE ]

Link me, please.

[/ QUOTE ]

linked (http://forumserver.twoplustwo.com/showflat.php?Cat=0&amp;Number=4011389&amp;page=0&amp;fpart=1&amp;v c=1)

[dknightx]

basically the point of this thread is that there is nothing in the future that you can be 100% certain about!

[fslexcduck]

Statements: Team A has the best basketball player in the nation. It is impossible for the team with the best basketball player to have a losing record.

Question: Is team A more likely to win more than 50% of its games than lose more than 50% of its games?

Answer: Yes.

[man]

[ QUOTE ]
basically the point of this thread is that there is nothing in the future that you can be 100% certain about!

[/ QUOTE ]
nah the point is that past events give no information about the probability of future events and my head's probably gonna explode

[john kane]

thank you to all those who replied, especially the longer ones which i found of particular interest.

me and zeejustin have the same analysis i believe that:

[ QUOTE ]

) Children w/ emotional disturbance are more likely to have Abdominal Pain.

2) According to the statement, the word especially directly implies that children w/ ED (emotional disturbance) + SD (Sleep disorder) are more likely to have AP than children with ED but not SD. Obviously someone is more likely to have ED + SD if we already know they have SD (even if we are completely informationless about whether or not they have ED).

So now let's look at the question. It asks, if SD, is AP more likely?

If SD, then SD +ED is more likely.

If SD+ED, then AP is more likely. Therefore, the answer is true.



[/ QUOTE ]

btw the wording i gave in the statement and question are identical to the actual question (copied from a test website).

can someone please explain where this trail of thought is incorrect or what we are assuming which we shouldnt.

ill post another question soon once ive definitely got the above problem solved.

many thanks

[CrazyEyez]

[ QUOTE ]
zeej,

no... you said "x means they are more likely to win than to lose" which is different than saying "x means they are more likely to win"

let's say x is something that increases probability of something happening from 10% to 30%. you could say "x means it is more likely to happen" but you could not say "x means it is more likely to happen than not to happen."

[/ QUOTE ]
light bulb

I get that, but that's not how I've been interpreting the questions. Is this right -

If you have symptom x, you're more likely to have symptom y than not have symptom y. Not enough info.

If you have symptom x, you're more likely to have symptom y than someone who doesn't have symptom x is likely to have symptom y. True.

??

[dknightx]

[ QUOTE ]
Can someone post some of these questions (like the bball one) where the answer is True or False? I think that would make this much clearer.

[/ QUOTE ]

When you are talking about using past information to predict future outcomes, the answer is ALWAYS "not enough information" ... however, in these test questions, there are times where the test will allow you to assume certain things to be true.

There is a difference between:

In the past, the team with awesome power X, won all of their games.

vs.

If a team has awesome power X, they automatically win.

In the first case, the only "true/false" statements you can make can only pertain to the past. Anything about the future is "not enough info" because we CAN NOT make a DIRECT correlation between X and winning just because past statistics have said so.

In the second case, we CAN make "true/false" statements pertaining to the future, because the test allows us to "assume" a certain statement to be true in all cases (which is X -&gt; win). Therefore a statement like "Team A has X, they will beat Team B who does not have X" will be true.

[skiier04]

The only conclusion that I can come to is that these questions suck.

[fslexcduck]

crazy,

if there's a correlation between having symptom x and having symptom y, then yeah, exactly.

[john kane]

basically that if i have a sleeping disorder, i am more likely to have emotional distress than if i did not have a sleeping disorder (edited to add: and with emotional distress i am more likely to have abdominal pains). i am therefore more likely to have abdominal pains with a sleeping disorder than without a sleeping disorder.

where am i going wrong?

[El Diablo]

ZJ,

I think you got hustled by your two friends.

[CrazyEyez]

[ QUOTE ]
crazy,

if there's a correlation between having symptom x and having symptom y, then yeah, exactly.

[/ QUOTE ]
Well in the OP, there was a correlation between x and a, and y and a, therefore x and y.

[dknightx]

jk:

the original statement is that SP in CONJUNCTION WITH SD means that the SP is more likely to be a symptom of MD. in this case, there is NO relation between SP and SD.

[danzasmack]

Duck,

did you take the LSATs?

[CrazyEyez]

[ QUOTE ]
basically that if i have a sleeping disorder, i am more likely to have emotional distress than if i did not have a sleeping disorder. i am therefore more likely to have abdominal pains with a sleeping disorder than without a sleeping disorder.

where am i going wrong?

[/ QUOTE ]
This is exactly where I'm at.

[ZeeJustin]

I think I can simplify this.

Almost all scientiests agree, a coin will land on heads 50% of the time. A few scientists say that due to the uneven weight distribution and aerodynamics, heads is ever so slightly more likely than tails. No scientists say that tails is more likely, because that would be absurd. Assume there is absolutely no chance that tails is more likely than heads.

Conclusion: Heads is more likely than tails.

On an LSAT test, we are supposed to say "Insufficient Information" because we don't know which scientists are right.

In a gambling setting, we understand that there is a non-zero chance that heads is more likely than tails. As a gambler, we would therefore always bet on heads, and never tails.

As far as I'm concerned, there is sufficient information.
As far as the LSAT test is concerned, there is not succifienct information.

The difference stems from the LSATs unorthodox definition of the phrase "more likely".

[danzasmack]

ZJ,

But this coin thing you just mentioned is different from betting on the bball team, correct?

[dknightx]

ZJ: in the original question, you say "Assume ...", because of that, we are allowed to assume that whatever follows the "assume" is always 100% true.

because of that, your conclusion would be "true" on the LSAT, and "not enough information" in real life.

[dknightx]

sorry, its actually "not enough info" on the LSAT too, because based on the "Assume ..." P(heads) = P(tails) is a possibility.

[ZeeJustin]

[ QUOTE ]
ZJ,

But this coin thing you just mentioned is different from betting on the bball team, correct?

[/ QUOTE ]

They are really similar. There's an obvious correlation between having the best player, and winning more than 50% of the time.

[IronUnkind]

I've constructed a mathematical version of this question, the terms of which are somewhat arbitrary but consistent with the question's verbal details.

1. There are 100 children in the world.
2. 49 children have abdominal pain.
3. 20 children with abdominal pain are emotional disturbed.
4. 15 children with ab pain are disturbed and have sleeping problems.
5. x children in the world have sleeping problems.

If x = 100, then 49 percent of them have ab pain and 51 percent of them do not, and the answer is false.

If x = 50, then 98 percent of them have ab pain and 2 percent do not, and the answer is true.

Since the statement's truth or falsehood depends on x, which we don't know, then it is clear that we have insufficient information to solve the question. (Note: emotional disturbance is irrelevant to the question).

[fslexcduck]

[ QUOTE ]
Duck,

did you take the LSATs?

[/ QUOTE ]

ya, 3 yrs ago. did pretty well

[NT!]

wait, how is there still serious discussion going on in this thread? i'm moving this pile of [censored], and don't ever post something like this again in OOT

[danzasmack]

[ QUOTE ]
[ QUOTE ]
ZJ,

But this coin thing you just mentioned is different from betting on the bball team, correct?

[/ QUOTE ]

They are really similar. There's an obvious correlation between having the best player, and winning more than 50% of the time.

[/ QUOTE ]

I guess the way your phrased that question though I would have to say don't bet it/not enough info to bet it.

[ZeeJustin]

[ QUOTE ]
"not enough information" in real life.

[/ QUOTE ]

Scientists discover a way to test this theory using a machine that will flip the coin a quadrillion times.

After 100,000 flips (obviously, a very small sample), the machine determines there is a 1 out of 10,000 chance that heads is .0000001% more likely. After a quadrillion flips, we will know for sure, but that will take time.

Now that we have numbers, the question hasn't changed. However, we can still do math, and heads will be &gt; 50%.

Edit: Accidentally wrote tails where I meant to write heads.

[fslexcduck]

[ QUOTE ]
wait, how is there still serious discussion going on in this thread? i'm moving this pile of [censored], and don't ever post something like this again in OOT

[/ QUOTE ]

um, are you serious? dictator much?

[drexah]

I'm 19, and possibly a downie (according to NT) and i knew it definitely wasn't true. initially thought it was not enough info and ended up picking it. although i could make an argument for false as well.

[ZeeJustin]

[ QUOTE ]
[ QUOTE ]
wait, how is there still serious discussion going on in this thread? i'm moving this pile of [censored], and don't ever post something like this again in OOT

[/ QUOTE ]

um, are you serious? dictator much?

[/ QUOTE ]

There is not enough information to determine whether or not he is serious.

[IronUnkind]

[ QUOTE ]
Almost all scientiests agree, a coin will land on heads 50% of the time. A few scientists say that due to the uneven weight distribution and aerodynamics, heads is ever so slightly more likely than tails. No scientists say that tails is more likely, because that would be absurd. Assume there is absolutely no chance that tails is more likely than heads.



Conclusion: Heads is more likely than tails.


[/ QUOTE ]

This conclusion is false. The only reason we would bet on heads is because even if the crackpot scientists are wrong (which is probable), then we are still on a freeroll.

[bwana devil]

[ QUOTE ]
The difference stems from the LSATs unorthodox definition of the phrase "more likely".

[/ QUOTE ]

ZJ, when logic problems dont work out how you expected (from well respected tests), dont disregard definitions to make wrong answers fit your own logic.

youre trying to fit your scenarios into gambling situations where youre willing to take a risk that you may be right 51% of the time to win money. these logic questions are asking you to definitively state that one event occurs 51% more of the time than the other. dont you see the difference?

[dknightx]

[ QUOTE ]
[ QUOTE ]
"not enough information" in real life.

[/ QUOTE ]

Scientists discover a way to test this theory using a machine that will flip the coin a quadrillion times.

After 100,000 flips (obviously, a very small sample), the machine determines there is a 1 out of 10,000 chance that heads is .0000001% more likely. After a quadrillion flips, we will know for sure, but that will take time.

Now that we have numbers, the question hasn't changed. However, we can still do math, and heads will be &gt; 50%.

Edit: Accidently wrote tails where I meant to write heads.

[/ QUOTE ]

ZJ:

You are a smart guy, have you taken any statistics classes before? The thing you will learn is that in statistics there is never anything that is 100% certain, nor is there every a 100% direct correlation between A and B. The only thing we can ever assume is "levels of confidence". Sure eventually these levels of confidence approach 100%, and they eventually become commonplace understanding.

As I was saying earlier, on tests like the LSAT, they will often allow you to assume things (like a coin is perfectly fair and 50/50), that we could not otherwise assume in real life.

[danzasmack]

Why is it false? Isn't it not enough info?

Edit - in reply to scientist one a few posts up

[fnord_too]

[ QUOTE ]
basically that if i have a sleeping disorder, i am more likely to have emotional distress than if i did not have a sleeping disorder. i am therefore more likely to have abdominal pains with a sleeping disorder than without a sleeping disorder.

where am i going wrong?

[/ QUOTE ]

The bolded part relies on an assumption.

You cannot conclude that if a sleeping disorder AND stomach pains increase the liklihood of having an emotional disorder than just having a sleeping disorder increases the liklihood of havina an emotional disorder.

That is P(ED|(SD&amp;AP) &gt; P(ED) and P(ED|AP) &gt; P(ED) and P(ED|SD&amp;AP) &gt; P(ED|AP) DOES NOT IMPLY P(ED|SD) &gt; P(ED)

Let me put some numbers to it.
Say P(ED) = .05
P(ED|AP) = .06
P(ED|AP&amp;SD) = .07
P(SD) = .5

P(ED|SD) can be below .05

To show this, consider 100,000 subjects
5000 have ED
10,000 have AP, of which 600 have ED
1,000 have AP&amp;SD so 70 of those have ED
If 10,000 have SD, you are only guaranteed that P(ED|SD) &gt;= .007
.007 &lt; .05, so we have a counter example to the assertion that P(ED|SD) &gt; P(ED)

These numbers all fit within the constraints.

[ZeeJustin]

[ QUOTE ]
This conclusion is false. The only reason we would bet on heads is because even if the crackpot scientists are wrong (which is probable), then we are on a complete freeroll.

[/ QUOTE ]

Do you change your view in my example with the numbers?

If no, why not? We have mathematical proof that says &gt;50%.

If yes, tell me how the question is any different.

[IronUnkind]

[ QUOTE ]
wait, how is there still serious discussion going on in this thread?

[/ QUOTE ]

Because there are still a bunch of people who don't get it.

[ZeeJustin]

[ QUOTE ]
dont you see the difference?

[/ QUOTE ]

Yes. The difference is in the definition of the phrase "more likely".

[bwana devil]

[ QUOTE ]
[ QUOTE ]
dont you see the difference?

[/ QUOTE ]

Yes. The difference is in the definition of the phrase "more likely".

[/ QUOTE ]

party on wayne.

[SNOWBALL]

[ QUOTE ]
wait, how is there still serious discussion going on in this thread? i'm moving this pile of [censored], and don't ever post something like this again in OOT


[/ QUOTE ]

NT!,

This thread is good and it belongs in either OOT or SMP, but since it started here, there's no good reason to move it. It already has an OOT flava, so I say leave it alone. Also, I can't believe that you are singling this thread out when there is so much total garbage in OOT right now.

[IronUnkind]

[ QUOTE ]
Do you change your view in my example with the numbers?

[/ QUOTE ]

Didn't see it, but obviously if you prove that heads is more likely (even slightly), then your conclusion is fine.

[ZeeJustin]

[ QUOTE ]
[ QUOTE ]
Do you change your view in my example with the numbers?

[/ QUOTE ]

Didn't see it, but obviously if you prove that heads is more likely (even slightly), then your conclusion is fine.

[/ QUOTE ]

Scientists discover a way to test this theory using a machine that will flip the coin a quadrillion times.

After 100,000 flips (obviously, a very small sample), the machine determines there is a 1 out of 10,000 chance that heads is .0000001% more likely. After a quadrillion flips, we will know for sure, but that will take time.

Now that we have numbers, the question hasn't changed. However, we can still do math, and heads will be &gt; 50%.

[john kane]

fnord, youve made two great posts. edit: i think ive got it now

so, (btw i understand all your annotations, just replying in writing though);

as an example. disease A gives symptoms B and C.

if there is a massive rise in symptom C, that could be becuase of say a disease D, and so a rise in symptom C would therefore not mean a rise in sympton B so long as disease D didnt have symptom B

therefore, in such a case, a rise in symptom C would not lead to a rise in symptom B. but it could if symptom C had been triggered by an outburst in disease A.

this is identical to an above poster about the heads and tails with some scientist thinking heads is a fraction more likely.

the problem is in my brain i think there are 2 scenarios one where it is caused by disease A and the other by disease D. i need to view this as two seperate cases (in which case you 'cannot say/insufficient info) rather than taking an average of the two (and therefore it leans to an increase in symptom C leads to a rise in symptom B).

[IronUnkind]

The proper conclusion is not: 1. Heads is more likely than tails.

In fact, even though this too would be technically false, a better conclusion than the first would be: 2. Heads is NOT more likely than tails.

Of course, the correct conclusion is (to be precise): 3. There is some chance that heads is more likely than tails.


Since statement 2 directly contradicts statement 1, and since it is mathematically more likely to be correct, then statement 1 must be false, given the information we have.

However, we would still bet on heads because you stipulate that there is a 0 percent chance that tails is more likely, so at worst, we are even money, and we might even have a slight edge.

But what if we were to bet on the truth of Statement 1 versus the truth of Statement 2, with the winner to be determined after one quadrillion flips? Would you still take statement 1? Even if I laid 1000 to 1?

[FortunaMaximus]

So how do you determine the probability valuation for the answer then?

[madnak]

Ahh! A big thread just fell out of the sky!

[FortunaMaximus]

Huh? There was a thread? /images/graemlins/grin.gif

[Jcrew]

[ QUOTE ]
[ QUOTE ]

fslexcduck, you question is STILL not enough info (right?). The reason is, you can not assume that having the best player is a DIRECT cause for having won more games than lost ... it could be a completely unrelated factor.

causation v correlation, bingo

[/ QUOTE ]

[/ QUOTE ]

I don't think that is right. The insufficient data is the prior distributions of wins for teams. If you are given the win distribution for a random team, and the win distributions for the teams with the best player each year, that would be sufficient data to claim whether the team with the best star would make a team have likely more wins than not.

I think the claim you need the actual causal mechanisms to make plausible inferences is incorrect. Bayes outlook.

[iversonian]

[ QUOTE ]
NL,

"Saying this w/o giving your answer is incredibly weak."

Wow, I'm shocked that you don't think the answer to this is about as trivially obvious as possible. This is like an easy question on the PSAT. Wow, honestly, I'm really shocked that you thought there was anything at all to discuss here.

[/ QUOTE ]

Be that as it may, the fact is that the answer is not undisputed. In fact, that is precisely why you posted this, isn't it? Because you couldn't believe that there was argument over it?

If you state that the (disputed) problem is trivial, then it is presumptuous to assert that and leave out your answer if for only the reason that if you could have misinterpreted the question and come to a wrong conclusion that you could be later called on. Whether El Diablo is entitled to be so presumptuous, well that is another question altogether.

[john kane]

anyone know where i can practice for my remaining verbal reasoning tests? websites? books? does anyone have any university/college material they could email me or give me the link to.

many thanks for any help.

[ZeeJustin]

[ QUOTE ]
The proper conclusion is not: 1. Heads is more likely than tails.

In fact, even though this too would be technically false, a better conclusion than the first would be: 2. Heads is NOT more likely than tails.

Of course, the correct conclusion is (to be precise): 3. There is some chance that heads is more likely than tails.


Since statement 2 directly contradicts statement 1, and since it is mathematically more likely to be correct, then statement 1 must be false, given the information we have.

However, we would still bet on heads because you stipulate that there is a 0 percent chance that tails is more likely, so at worst, we are even money, and we might even have a slight edge.

But what if we were to bet on the truth of Statement 1 versus the truth of Statement 2, with the winner to be determined after one quadrillion flips? Would you still take statement 1? Even if I laid 1000 to 1?

[/ QUOTE ]

You are making this much more complicated than it needs to be.

If everything is taken literally, heads is more likely than tails. We can prove this mathematically.

You are taking everything in an abstract metaphysical sense. In a metaphysical sense, you are right. We do not know if heads is more likely than tails or not.

[IronUnkind]

[ QUOTE ]
You are taking everything in an abstract metaphysical sense. In a metaphysical sense, you are right. We do not know if heads is more likely than tails or not.

[/ QUOTE ]

There was nothing "metaphysical" about my response. A very strange accusation, given the fact that I answered your question AS LITERALLY AS POSSIBLE.

Once you run the experiment, our answer may change because we will know for certain. Until such a time, the best conclusion is the most precise one, and the second best answer is the one which represents the most likely outcome of the experiment.

[Bork]

[ QUOTE ]
[ QUOTE ]
The proper conclusion is not: 1. Heads is more likely than tails.

In fact, even though this too would be technically false, a better conclusion than the first would be: 2. Heads is NOT more likely than tails.

Of course, the correct conclusion is (to be precise): 3. There is some chance that heads is more likely than tails.


Since statement 2 directly contradicts statement 1, and since it is mathematically more likely to be correct, then statement 1 must be false, given the information we have.

However, we would still bet on heads because you stipulate that there is a 0 percent chance that tails is more likely, so at worst, we are even money, and we might even have a slight edge.

But what if we were to bet on the truth of Statement 1 versus the truth of Statement 2, with the winner to be determined after one quadrillion flips? Would you still take statement 1? Even if I laid 1000 to 1?

[/ QUOTE ]

You are making this much more complicated than it needs to be.

If everything is taken literally, heads is more likely than tails. We can prove this mathematically.

You are taking everything in an abstract metaphysical sense. In a metaphysical sense, you are right. We do not know if heads is more likely than tails or not.

[/ QUOTE ]

No we don't know heads is more likely than tails. We know that this could be true. Since we ruled out that tails is more likely than heads the only alternative possibility is that they are equally likely. The equally likely state of affairs is exactly as probable as heads being more probable than tails. We just don't have any information either way until we do further tests.

Of course if you are betting on the outcome of a flip you would bet on heads, because you are at worst 50/50. However we are not betting on a flip we are trying to draw a justified conclusion about whether the flip is 50/50 or 51/49. Could be either we don't know until we test for more info.

[soon2bepro]

voted true because, even if there's insufficient information to come to a relevant percentage of higher likeliness, there's still a higher chance of this, however small; when compared to children who don't have trouble sleeping as the other group.

[john kane]

just to say thanks to all your helpful comments. i just did the online tests for one of the other three companies and i was accepted, and maybe without all this debate i may not of done. so thanks.

i may still [censored] up the other two, but its made me feel better.

[Lestat]

Wow, how did I miss this thread?

It seems to me the answer is: Insufficient information. At least I don't have enough info to say whether the statement is true or false.

[Lestat]

The tendency I have on some tests is to over complicate things, expecting that there's a "catch" or that it's much harder than it really is.

So just use your own common sense. It really was an easy question. Trust yourself and you'll do fine.

[DcifrThs]

this is rediculous the # of peole whose responses have read that have not done this simple proof. it may be later down int he thread so sorry if i'm repeating something.

A=abdominal pain
B= emotional disturbance
C=phobias/sleep disorders, nightmares etc.

Statement: A may cause B especially where C is present

Question: C leads to A.

are you kidding me?

Barron

[DcifrThs]

[ QUOTE ]
The tendency I have on some tests is to over complicate things, expecting that there's a "catch" or that it's much harder than it really is.

So just use your own common sense. It really was an easy question. Trust yourself and you'll do fine.

[/ QUOTE ]

in ALL of these types of tests, substitution is the best thing ever. see my response below for an example.

when given brain teasers/gmat qeustions etc. using real #sthat you choose is best.

for example:

assume a population where 5% are drug users. a drug test is 95% accurate.

assume a person is picked at random from the population and tested, given that their test result is positive, what is the probability that they use drugs?

thinking about this problem without numbers is retarded.

population= 1000
drug users=50
drug test results = 95%

how many drug users would test positive? 47.5
how many not drug users would positive negative? 47.5

what is the probability a positive test result is a drug user? 50%

Barron

[Lestat]

It may seem like a frivelous endeavor for you, but I wish you'd post more in the SMP forum.

[DcifrThs]

[ QUOTE ]
It may seem like a frivelous endeavor for you, but I wish you'd post more in the SMP forum.

[/ QUOTE ]

i'll certainly spend more time here as you guys definately helped me (and OOT did as well) prep for my interview today which i feel i performed well on.

i'll get the results back in a few days/a week at most.

i do enjoy the stuff here and i'll be posting more for ya /images/graemlins/smile.gif

i also want to apologize for how i came accross in those 2 posts in this thread. it was just kinda frustrating i guess to see like 100 responses of which i read maybe 25 and nobody did a single logical A, B, C kinda thing. but in reality i know that knowledge seems easy when you have it and i'll keep that in mind in the future. so im sorry for that lapse in the thread.

Barron

[john kane]

Barron, thanks for your replies.

firstly the question is not 'does C lead to A' its whether 'A is more likely if C is present'.

now in my mind, if you have C, you are more likely to have A, as you are more likely to have B.

as you say, giving numbers.

There are 100 diseases which give symptom C at equal probability distribution. One of these diseases is Disease B, which also gives out symptom A. No other disease gives both symptoms.

So if you are showing signs of symptom C, there is a 1% chance it is due to Disease B. Therefore you will have a 1% greater chance of showing signs of symptom A. therefore people with C are more likely to have A.

I dont see where my logic is wrong. Please explain, id be very appreciative.

Well done on doing well on your interview,

john

[Alex-db]

[ QUOTE ]
this is rediculous the # of peole whose responses have read that have not done this simple proof. it may be later down int he thread so sorry if i'm repeating something.

A=abdominal pain
B= emotional disturbance
C=phobias/sleep disorders, nightmares etc.

Statement: A may cause B especially where C is present

Question: C leads to A.

are you kidding me?

Barron

[/ QUOTE ]

I am very comfortable with this type of logic, but occasionally make mistakes on this type of tests because when, in your example: "Question: C leads to A." I get a strong internal reaction knowing that claim cannot be made and I say to myself that a suggestion that that assertion is "certainly true" is certainly false.

I then put "False" even though I understand perfectly well on reflection that the answer they were looking for was "Insufficient Info" and I know I have just tripped over my own semantics (or something like that?) by thinking to myself "that's false because you can't back it up" and forgetting about the chance its true by coincidence.

So that is an error to watch out for.

You should also be very careful to check whether the specific test is looking for strict legal-style wording analsis, or a judgement on which correct answer is implied.

Some companies test the latter on the logic that business correspondence doesn't always contain perfect information and using common sense to extract sensible meaning from vague facts can be what they are looking for. This leads to the correct answers more often being true/false than insufficient info, and is really important to check for and practice beforehand.

[bkholdem]

A (abdominal pain in children) may be a result of E (emotional disturbance), especially where it appears along with P/SD (phobia's/sleep disorders).

Children who have SD have greateer &gt; A
than
Children on average

I say it is true if I am understanding the above since Children who have SD's are a subset of all children and we know that all children are equally likely or equally unlikely to have abdominal pain (given no other info).

When we add one piece of known info (SD's), we also know that there is a coorelation between children who have SD's and A. This is because of all the children who have sleep problems, some of them have SD's (sleep disorders). And children who have SD's have a higher insidence of abdominal pain than children on the whole. The connection is casual, the question does not stipulate that sleep disorders have a cause effect to abdominal pain. It stipulates only that of all children who exist, those who have sleeping problems will have a higher instance of abdominal pain.

[bkholdem]

[ QUOTE ]
[ QUOTE ]
The tendency I have on some tests is to over complicate things, expecting that there's a "catch" or that it's much harder than it really is.

So just use your own common sense. It really was an easy question. Trust yourself and you'll do fine.

[/ QUOTE ]

in ALL of these types of tests, substitution is the best thing ever. see my response below for an example.

when given brain teasers/gmat qeustions etc. using real #sthat you choose is best.

for example:

assume a population where 5% are drug users. a drug test is 95% accurate.

assume a person is picked at random from the population and tested, given that their test result is positive, what is the probability that they use drugs?

thinking about this problem without numbers is retarded.

population= 1000
drug users=50
drug test results = 95%

how many drug users would test positive? 47.5
how many not drug users would positive negative? 47.5

what is the probability a positive test result is a drug user? 50%

Barron

[/ QUOTE ]

Are you crazy? The test is 95% accruate. The probability they use drugs is much higher than 50%. You are not factoring in all the other 900 people who could have been given the test and 95% of those would be negative.

[madnak]

Yeah, in a finite population when quality A correlates positively with quality C, quality C will also correlate positively with quality A. If As are more likely than not-As to have C, then Cs are more likely than not-Cs to have A.

But that's a huge technicality, and is much too sophisticated and tricky to apply here. If the question were that tricky, it should also be phrased more carefully.

[DcifrThs]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
The tendency I have on some tests is to over complicate things, expecting that there's a "catch" or that it's much harder than it really is.

So just use your own common sense. It really was an easy question. Trust yourself and you'll do fine.

[/ QUOTE ]

in ALL of these types of tests, substitution is the best thing ever. see my response below for an example.

when given brain teasers/gmat qeustions etc. using real #sthat you choose is best.

for example:

assume a population where 5% are drug users. a drug test is 95% accurate.

assume a person is picked at random from the population and tested, given that their test result is positive, what is the probability that they use drugs?

thinking about this problem without numbers is retarded.

population= 1000
drug users=50
drug test results = 95%

how many drug users would test positive? 47.5
how many not drug users would positive negative? 47.5

what is the probability a positive test result is a drug user? 50%

Barron

[/ QUOTE ]

Are you crazy? The test is 95% accruate. The probability they use drugs is much higher than 50%. You are not factoring in all the other 900 people who could have been given the test and 95% of those would be negative.

[/ QUOTE ]

what would you like to wager?

OOPS EDIT:before you state a $figure between 0 and your net worth, note that i meant to state in teh original post

"how many NON-drug users will test POSITIVE? 47.5"
"how many drug users will test POSITIVE? 47.5"

Barron

[DcifrThs]

[ QUOTE ]
No other disease gives both symptoms.

[/ QUOTE ]

i think this is where you get off track.

let me reread the question and reply in more detail.

Barron

[DcifrThs]

let me try this again:

Statement: A may be a result of B. This is more likely the case when C or D are present as well.

Question: Children who have D are more likely to suffer A.

I still see this as a very clear NSI.

B is more likely w/ D present given A is present.

we now are told we dont know if A is present. we are told D is present and asked to assess a relative probability of a patient having B.

there is not enough info.

Barron

[soon2bepro]

But the original example makes you assume that C is also a symptom of B, which explains where the extra probability that A is a symptom of B when C is present comes from. Since B sometimes causes both A and C, children with C are more likely to have B, and therefore are more likely to have A aswell.

[Hiltonian]

[ QUOTE ]
But the original example makes you assume that C is also a symptom of B, which explains where the extra probability that A is a symptom of B when C is present comes from. Since B sometimes causes both A and C, children with C are more likely to have B, and therefore are more likely to have A aswell.

[/ QUOTE ]
This is not true.

[Mickey Brausch]

[ QUOTE ]


Statement: Abdominal pain in children may be a symptom of emotional disturbance, especially where it appears in conjunction with phobias or sleep disorders such as nightmares or sleep-walking.

Question: Children who have problems sleeping are more likely to suffer from abdominal pain.
TRUE?
FALSE?
INSUFFIENT INFORMATION?

[/ QUOTE ]

Since the statement of facts is not to be disputed, I need to get away from the specifics of the statement. So I replace the facts with symbols (which also helps clarify the statement) and concentrate on the logic.

A = Children have abdominal pain
B = Children have emotional disturbance
C = Children have phobias or sleep disorders, eg sleep walking.

OK, now the Statement says

<u>When A then possibly B, especially when also C</u>

In other words, this is something like A &amp; C --&gt; B

And the Question is

<u>When C, then A is more likely?</u>

Irrespective of what A, B or C are actually about, the above construction is not logical, as you can see : If we get C on top of A, then this makes B more likely -- but C on its own does not necessarily make A more likely.

Using a more mundane situation probably makes things clearer.

A = I'm having diarroea
B = I have not eaten properly
C = I'm feeling stomach pains

When A then possibly B, especially when also C.
But C does not necessarily make A more likely.

Mickey Brausch

[yukoncpa]

[ QUOTE ]
Using a more mundane situation probably makes things clearer.

A = I'm having diarroea
B = I have not eaten properly
C = I'm feeling stomach pains

When A then possibly B, especially when also C.
But C does not necessarily make A more likely.

Mickey Brausch


[/ QUOTE ]
If I'm having stomach pains vs. Not having stomach pains. Isn't there a greater probability that I'll be having diarroea with the former condition?