# What test to get a P-Value?

#### crisisinmedicine

##### New Member
So I have a group of students all divided by different categories in response to a poll. There is a total of 34 students. Depending on each answer, I assigned the student a 1, 2, 3, or 4 for that category. For example, favorite candy 1 = twix 2 = snickers 3 = gum 4 = none. Now that I've divided them up into 4 categories, how can I find if there is a statistical difference between the candy they prefer? Is this also called categorical data? Thanks!

#### bruin

##### Member
I've divided them up into 4 categories...Is this also called categorical data?
Yes.

how can I find if there is a statistical difference between the candy they prefer?
If by this you mean that you are curious whether the scores are evenly distributed among your four categories (frequency = 34/4 = 8.5 in each category), use a chi-square goodness of fit test.

Is this for a class project?

#### crisisinmedicine

##### New Member
Yes.

If by this you mean that you are curious whether the scores are evenly distributed among your four categories (frequency = 34/4 = 8.5 in each category), use a chi-square goodness of fit test.

Is this for a class project?
Homework/trying to learn general statistics for just everday kind of use. I like doing surveys just to find out stuff about my student body (I plan on pursuing a PhD when I'm older in sociology so I need to learn how to do this kind of stuff).

How do I run a chi-square goodness of fit? When I input the data into my SPSS I got an error. The first question in the survey was do you like candy 1=yes 2=no, I have 32 "1s" in column 1 and 2 "2s" in column 2. What do I click/etc. to get a p value? Thanks!!!!!!!!!!!!

#### Karabiner

##### TS Contributor
Andy Fields "Discovering statistics using SPSS" is often recommended for beginners.

With kind regards

K.

#### crisisinmedicine

##### New Member
Andy Fields "Discovering statistics using SPSS" is often recommended for beginners.

With kind regards

K.
OK Thanks. If I want to run 4 variables up against another, how can I input them to get 1 p-value? I keep messing around with the rows/columns but ultimately get 3 p values

#### Karabiner

##### TS Contributor
OK Thanks. If I want to run 4 variables up against another, how can I input them to get 1 p-value? I keep messing around with the rows/columns but ultimately get 3 p values
Yes. You didn't describe your data well, seemingly you did structure them inadequately, you didn't know which tests to run and how to perform them within your software. Given that you have been asking here to solve your data analysis problems for a year or so, you should maybe give it a try to learn the basics.

E.g., it would be useful if you first learned how to describe your data and to state your research question(s) properly. You said you divided the group into 4 subgroups, but now you say each participant has 3 values, not just 1 (his grouping category). That's confusing.

With kind regards

K.

#### bruin

##### Member
Homework/trying to learn general statistics for just everday kind of use. I like doing surveys just to find out stuff about my student body (I plan on pursuing a PhD when I'm older in sociology so I need to learn how to do this kind of stuff).
OK. I just asked because, frankly, sometimes people think that sprinkling a little statistics in their research will make an uninteresting research question suddenly interesting or worthwhile. In this case, I don't think there's any reason to think candy preferences ARE evenly divided among the arbitrary categories of "twix" "snickers" "gum" and "none", so it wouldn't be particularly interesting to prove that they AREN'T in fact so divided up.

However as an example of categorical data on which to try out a chi square test your example works fine. Practicing stats if you plan on a grad degree is certainly a good idea.

Try following this for doing it on SPSS:
http://www.stat.vcu.edu/help/SPSS/SPSS.GoodnessOfFit.PC.pdf