Question

All the following would be evidence that you do not have a normal distribution, except:

the presence of multiple modes.

the fact that 75% of the observations have a value below the mean.

a normal quantile plot suggests an obviously curved line.

the likelihood for all values of x is the same.

None of the above

I believe the correct answer is A, since a normal distribution does not have more than one mode according ot its definition.

Answer 1

@perl

Answer 2

having multiple modes is a clear evidence that the distribution is not normal right ?

so that cannot be your answer, strike it off

Answer 3

oh i see i misread the question

Answer 4

yes its a tricky one, go thru remaining options and see if u can eliminate a few more

Answer 5

Well since the normal curve has a theoretical area of 1 all the time, then if it were to come down to chance in any part of the area under the curve, the answer most suitable to not indicate it is not a normal distribution would have to be D

Answer 6

haha lol, hbu @ganeshie8

Answer 7

let me give u a quick example : suppose u have a normal distribution of ages of "all the living people "

are the numbers "20 years" and "200 years" equally likely ?

Answer 8

yup

Answer 9

Actually, that's not right, it's just worded weirdly so lets cross that out as well.

Answer 10

how many living people you know are 200 years old ?

Answer 11

All the following would be evidence that you do *not* have a normal distribution,

Answer 12

Perl gave away the answer xD

Answer 13

no i didn't

Answer 14

i just emphasized a part of the question

Answer 15

that was the one i chose dude lol

Answer 16

technically the probably of any x value is zero, because it is a continuous distribution

Answer 17

byt ok ty guys for helping me learn how to read a question lol

Answer 18

There we go perl

Answer 19

P(X = x) = 0 , for any x

Answer 20

what do we have ztables for

Answer 21

for intervals

Answer 22

P ( a < X < b )

Answer 23

P ( X > a) , P ( X < b ) , etc

Answer 24

you would be right if it was a discrete distribution, P( x = 20) is different than P( x = 200 )

Answer 25

So the answer is D?

Answer 26

can't give out answer

Answer 27

so if the SAT scores are normaly distributed, the probability of getting a score of exactly 350 is 0 ?

Answer 28

correct

Answer 29

if you are using a continuous scale

Answer 30

There's a probability that x is equal to a value of 0

Answer 31

a better example is weight or temperature , continuous variables

Answer 32

yeah you're right, probability of getting a score of 350 or higeher is P(X > 350)

Answer 33

this has to do with the way we define probability, as area under the curve

Answer 34

Right.

Answer 35

P( X = a ) = integral { a, a} f(x) , but the area is zero

Answer 36

P ( a < X < b ) = integral { a,b} f(x) , which is not zero if a =/= b

Answer 37

It's just the way it's defined a better way would be to say specific value is 0

Answer 38

this is for f(x) 'density' curve

Answer 39

iam, you mean a better way to say it? any specific value is zero

Answer 40

That's right

Answer 41

I think its none of the above

Answer 42

here is a classic example.

suppose you have to pick a decimal randomly between [0,1]

Answer 43

that gives uniform distribution right

Answer 44

right

Answer 45

the probability of picking .348 , for example, is zero

Answer 46

How do I give a thumbs up emoticon on here?

Answer 47

the probability of picking a number between 0 and .5 is going to be 1/2