Help me with my Maths homework

[Deleted]

Hi there,

Hopefully this won't be a case of me simply over-thinking what i need to do again.

I recently got some data back from a survey I did using the Likert system. I now need to attempt to apply some statistical analysis to the data i've received.

That is where I'm coming unstuck, big time.

So some scenarios that I would like to attempt:

If you look at the image questions 1-4 are about different sources of stem cells, how would I go about determining if there is a significant difference between the 4 questions?

What test would I need (and how) to do on comparing if there is correlation between the answers for question 7 and the answers for question 1,2,3 and 4 (also I don't know if it's possible to determine if there is a correlation between question 7 and all the responses for questions 1-4)?

And any other tests that you could think of for analysing the data.




The questions asked were:

1. Stem cell research using cells taken from adults does not raise ethical concerns for you?
2. Extracting stem cells from embryos with no harm to the embryo is acceptable?
3. Stem cells taken from aborted embryos are acceptable?
4. Using excess embryos from IVF treatments for stem cell research is acceptable?
5. You are worried about the future impact of stem cell research?
6. You feel that you have a good knowledge on the issue of stem cell research?
7. You feel that stem cell research is morally correct?

[Deleted]

I don't know what tests you'd have to do, or whether you need to plot a graph or something ridiculous (been out of school too long, QQ), but;

If Q1-4 were vastly different, peoples' answers would be. If their answers (ie the numbers) are fairly uniform, then they hold the same opinion for each one and as such they aren't. Q7 should also be fairly similar to their answers for 1-4, so perhaps plot a graph with median of 1-4 against Q7 answers along with line of best fit or something, but without context of what you've been studying I've no idea whether that's right or not.

[Deleted]

I don't know what tests you'd have to do, or whether you need to plot a graph or something ridiculous (been out of school too long, QQ), but;

If Q1-4 were vastly different, peoples' answers would be. If their answers (ie the numbers) are fairly uniform, then they hold the same opinion for each one and as such they aren't. Q7 should also be fairly similar to their answers for 1-4, so perhaps plot a graph with median of 1-4 against Q7 answers along with line of best fit or something, but without context of what you've been studying I've no idea whether that's right or not.

Well I've got a comparison between the people who felt x about question 7 and also felt x with all sources of stem cells and there is clearly correlation between the two.

37 people disagreed with question 7 and all the sources of stem cells
19 people neutral with question 7 but only 7 of those also were neutral with all the sources
44 people agreed with question 7 and 40 of them also agreed with all the sources

Which demonstrates that of the people surveyed the source of the stem cells did not alter their opinion on whether it was morally correct. But i'm not sure how i could test to a scientific degree of confidence.

The frequencies for the responses are fairly equal with agree and disagree aswell, with no real clear winner when i converted the data into simplified categories. With there being a very similar split between general disagreement and agreement with few neithers.

45 41 41 41
11 19 7 7
44 40 52 52

Those are the frequencies for questions 1-4 when simplified to a 3 point scale of agree, disagree and neutral. AS you can see the agree and disagree for each is very similar.

---------- Post added 2012-01-13 at 06:17 PM ----------

I've attempted to apply a chi squared test to question 1 and question 7. However when i get to dividing by the expected value i have a problem because the expected value for some of the cells is 0. Does anyone know how i could correct that?