# Statistics help for resit

Watch
Announcements
#1
So I've got a resit coming up in about 3 weeks time for a stats module as part of my university degree (not maths), and for the assignment we are expected to use the software 'R' to perform certain statistical tests on a set of data we're given. Is there anyone who has any idea about which statistical tests (e.g. Generalised Linear Models, ANOVAs, t-tests, Chi-Square tests, etc.) are appropriate with which types of data (poisson, normal, binomial distribution etc.) and how to identify what kind a data set is? Also anyone with experience with R? Any advice/tips/pointers/hints would be greatly appreciated! Thanks!
0
3 years ago
#2
(Original post by CrossedHammers6)
So I've got a resit coming up in about 3 weeks time for a stats module as part of my university degree (not maths), and for the assignment we are expected to use the software 'R' to perform certain statistical tests on a set of data we're given. Is there anyone who has any idea about which statistical tests (e.g. Generalised Linear Models, ANOVAs, t-tests, Chi-Square tests, etc.) are appropriate with which types of data (poisson, normal, binomial distribution etc.) and how to identify what kind a data set is? Also anyone with experience with R? Any advice/tips/pointers/hints would be greatly appreciated! Thanks!
This is a question with an extremely wide scope! My first recommendation for learning "R" is Dalgaard's book "Introductory Statistics with R".

Whenever you're considering which statistical test that you want to use, the first question that you should ask yourself is "what is the scientific question?", then the second question is "what type of data do I have?". To answer this part of your question, it would be very useful if you gave some examples - along the lines of "how do I...?" then I can give you some more specific answers.
0
#3
Thanks for replying. The scientific aim is to determine how much of an influence anthropogenic activities have on a species of fish.

I am provided with a range of data (some anthropogenic some natural) for 21 coral reef sites studies, and for each I have data for the number of fish species counted; the reef size; whether fishing and water sports occur; what proportion of the year the reef is subjected to scuba diving activity; the minimum mean temperature; density of the fish species' primary food competitor; and the Simpson index of diversity.

I suppose my main "how do I" question would be how would you suggest I efficiently summarise the effect human activities (fishing, watersports, scuba diving) have on fish species when compared to fundamental (natural) variables (min. Temp., competitor density, Simpson's diversity index).
0
3 years ago
#4
(Original post by CrossedHammers6)
Thanks for replying. The scientific aim is to determine how much of an influence anthropogenic activities have on a species of fish.

I am provided with a range of data (some anthropogenic some natural) for 21 coral reef sites studies, and for each I have data for the number of fish species counted; the reef size; whether fishing and water sports occur; what proportion of the year the reef is subjected to scuba diving activity; the minimum mean temperature; density of the fish species' primary food competitor; and the Simpson index of diversity.

I suppose my main "how do I" question would be how would you suggest I efficiently summarise the effect human activities (fishing, watersports, scuba diving) have on fish species when compared to fundamental (natural) variables (min. Temp., competitor density, Simpson's diversity index).
So just one question of clarification: does this mean that you have 21 observations, reef_1, reef_2, ..., reef_21, and for each of these you have a number of fish species, a number that represents reef size, a yes/no field for fishing, a yes/no field for water sports, and so on?
0
#5
Yes that's correct...and a percentage value for scuba diving, a number for min temperature, and a number for competitor density and Simpson's diversity.
0
3 years ago
#6
(Original post by CrossedHammers6)
Yes that's correct...and a percentage value for scuba diving, a number for min temperature, and a number for competitor density and Simpson's diversity.
OK - so your outcome is the number of fish species per reef and you're interested in the effect that a number of different covariates (reef size etc.) have on that outcome variable.

If you had more data, then the obvious answer to this question would be to use Poisson multilinear regression (a form of GLM) with number of fish species as the outcome and all the other variables as covariates. (n.b. if the numbers of fish species per reef were fairly large, then ordinary multilinear regression would work too). Unfortunately, with only 21 observations, this will not work (the rule of thumb is that you need 10-20 observations per covariate in the regression equation).

So what you're going to have to do here is a simple univariate (Poisson or ordinary) regression for each covariate of interest.

Now, you haven't mentioned what sort of course this is (other than "not maths"!), so it is possible this is too sophisticated an approach - let me know what subject/level course you're on and I'll advise if I think you need to tone it down!
1
#7
Apologies for the delay!

That makes sense I think about the inappropriateness of using the poisson multi linear regression.

I mentioned that this is for a university degree course. The subject is geography, and in terms of level it is a 3rd year module so it sounds like the sophistication of your approach so far is exactly what is required!
0
X

new posts
Back
to top
Latest
My Feed

### Oops, nobody has postedin the last few hours.

Why not re-start the conversation?

see more

### See more of what you like onThe Student Room

You can personalise what you see on TSR. Tell us a little about yourself to get started.

### Poll

Join the discussion

#### Do you think receiving Teacher Assessed Grades will impact your future?

I'm worried it will negatively impact me getting into university/college (145)
42.9%
I'm worried that I’m not academically prepared for the next stage in my educational journey (39)
11.54%
I'm worried it will impact my future career (28)
8.28%
I'm worried that my grades will be seen as ‘lesser’ because I didn’t take exams (72)
21.3%
I don’t think that receiving these grades will impact my future (34)
10.06%
I think that receiving these grades will affect me in another way (let us know in the discussion!) (20)
5.92%