Question

How do you know what type of problem to use hypothesis testing on?

Answer 1

@jim_thompson5910

Answer 2

hey do u have any really good man questions for me

Answer 3

until jim gets here

Answer 4

lol, just this one. I don't understand how you know to do hypothesis testing when it's not that chapter of the book, ya know?

Answer 5

`How do you know what type of problem to use hypothesis testing on?`

The question seems a bit vague. Is there more to it?

If not, then the types of problems you use hypothesis testing on are the ones where you can use some known distribution (eg: the standard normal distribution). If the data doesn't fit neatly with a well known distribution, then maybe you can use the CLT (central limit theorem) to look at the sampling distribution. In most cases, the sampling distribution will be approximately normal for large values of n (usually n > 30)

Answer 6

lol, just this one. I don't understand how you know to do hypothesis testing when it's not that chapter of the book, ya know?

Answer 7

Sorry, OS went down for me. This isn't a question they have, it's my own. I just don't understand how you know to do hypothesis testing

Answer 8

like when you are not given anything other than your sample, so you check random, success/failure, and 10% condition, does that automatically mean hypothesis testing?

Answer 9

it sounds like you're describing a proportion test, so you'd use a Z distribution to conduct hypothesis tests like

"is the population proportion p equal to 50%?"

"do more than 46% of the people vote?"

things like that

Your null is ALWAYS going to have an equal sign in it somehow. This is so you can set up a fixed distribution around the center

so the null is going to be set up like p = 0.5 or p >= 46. The alternative hypothesis is based on the claim. So if the claim is "more than" then you have a right tailed test. If the claim is "less than" you have a left tailed test. If someone made the claim that "50% of the people vote" then you have a two tailed test

Answer 10

Just what one? You mention a problem and seem to be wondering whether you can or cannot apply hypothesis testing, but have not yet shared the problem with the rest of us.

Hypothesis testing ends with a decision to either 1) accept or 2) reject a "null hypothesis."

Often a hypothesis testing problem will begin with the words "At the 5% level of significance, test the null hypothesis (followed by a description of what is being tested)." Given the typical hypo. testing circumstances, you'll be given certain facts that become the bases for writing a null hypothesis and an alternative hypothesis.

For example: "Let the null hypothesis be that the population mean is 4.5 and the alternative hypothesis be that the population mean is greater than 4.5."

It's important that you read up on stating hypotheses and alternative hypotheses, as well s on the actual hypo. testing. Have you a textbook? online reference?

Answer 11

When you ask a question about some population parameter (eg: population proportion), you would ideally like to take a census. That's too expensive usually. It is not feasible to sample everyone in the population. It would take too much time, energy, money, etc

So instead you take a random sample that represents the population and you do a hypothesis test to help answer the question. The conclusion of the test isn't 100% true because there's always room for error

Answer 12

So, as long as it's set up so you can say p= something, you can use hypothesis testing?

Answer 13

Unfortunately in probability and statistics, p is used a lot.

the population proportion p is going to be different from the p-value (even though the two are connected in some way)

so be sure to keep that in mind and also let your students know as well

Answer 14

I'm just trying to understand when this test is used. This book is very calculation based without explaining when the tests are used and why

Answer 15

ok, I'll tell them about the p part.

Answer 16

Yes if someone made the claim "50% of the people vote" then you can set up

Null:
Ho: p = 0.5

Alternate:
Ha: p =/= 0.5

then do a hypothesis test to check the claim

Answer 17

Alright, thanks jim.. And thanks mathmale

Answer 18

Chick: Afraid this situation is much more complex than that.

As Jim points out, you must set up a null hypothesis AND an alternative hypothesis,
choose a level of significance, calculate probabilities, and so on.

If you're actually going to try teaching hypothesis testing, I hope you'll spend a couple of hours with a good introductory statistics textbook first. Neither Jim nor I could teach you hypothesis testing in the usual 5- or 10-minute interaction you get here on OpenStudy.

Answer 19

Were you to do that, then you could ask clarifying questions here on OpenStudy.

Answer 20

These are clarifying questions. I don't have a choice on teaching it. I teach it tomorrow. It's not that I don't get the null, alternative, and how to calculate it. I get the math part. I'm trying to get the why and figure out usage. The tests themselves are easy to do. Even writing the hypotheses are easy, but in order to explain it to my liking, I need to know the whys and whens and all my questions that I get on the application. This is why I've asked this.

Answer 21

It's not that bad really. It's something taught in many science courses

step 1) make a claim
step 2) set up hypotheses
step 3) set up rules on how to reject the hypothesis
step 4) gather data
step 5) compute the data to form a test statistic
step 6) you can use the traditional/classical approach OR the p-value
step 7) whichever method you use, they both will point to either you rejecting Ho or failing to reject
step 8) it's usually a good idea to translate the results into common english. This is handy for instance, if you're giving a presentation to your boss who isn't versed in the language of statistics

Answer 22

The hypothesis and the tests used are all based on the distribution the data is based on. Which is why it's always a good idea to identify the distribution and test method (is it a T test? a Z test? or maybe you're using the F distribution? etc)

Answer 23

I'm requiring a conclusion statement in english for it. I have a great activity set up for it(mock trial), but I just wasn't finding the when to use info. This is the only test they know

Answer 24

It's the very beginning of the section. They just learned confidence intervals

Answer 25

But anyways, thanks again @jim_thompson5910 . I think I got it enough now. I might do a practice problem or two, you think you'd be able to check it??

Answer 26

If you're looking for conditions needed to do a Hypothesis Test for a Proportion, then this page may help

http://stattrek.com/hypothesis-test/proportion.aspx?Tutorial=AP



` I might do a practice problem or two, you think you'd be able to check it??`

sure I can take a look at it