gavinkil

Hi I am trying to analyse some stats without much knowledge within this area.

I was wondering if anyone that did have knowledge could explain the Chi-square test to me

Diagram provided below:

Crosstab
Count
If I had more money I would always buy products which support good causesTotal
Strongly DisagreeDisagreeAgreeStrongly AgreeDefinitely
GenderMale0974525
Female32271125
Total3119111650



Chi-Square Tests
ValuedfAsymp. Sig. (2-sided)
Pearson Chi-Square13.301(a)4 .010
Likelihood Ratio15.0544 .005
Linear-by-Linear Ass2.9701 .085
N of Valid Cases50
a4 cells (40.0%) have expected count less than 5. The minimum expected count is 1.50.

gavinkil

Looked a lot better in a table

Andy Jordan

do you know what degrees of freedom are?

rohan

So what's the hypothesis, and what is the value of the test statistic?

Andy Jordan

looks like it's going for statistical independence for each condition in a contingency table, if I had to guess.

find the expected table, sum (observed - expected)^2 / expected, compare to critical value.

given that there are 2 rows and 5 columns, I believe there are 4 degrees of freedom; thus, the critical chi is 9.487729037.

your test value is more extreme than test chi, so you reject null and conclude that one's gender does indeed influence their decisions.

Andy Jordan

yup, that was it.

I have an Excel spreadsheet I just cooked up if you want to see how I did it?

gavinkil

2 secs just re-loading back up SPS

The Hypothesis is that one's gender is influenced by the given question.

To be quite honest I have no idea what its trying to tell me from my figures.

Andy Jordan

at an alpha of 5%, ofc.

gavinkil

Well what ever you think is easiest I am using SPSS to analyse the data as I stated just there.

Basically just want to know what it is trying to tell me such as:

What are the key figures I should be looking at and what proves or disproves the hypothesis

Andy Jordan

I sent you the spreadsheet to the e-mail listed in your profile.