Hello,

My name is Chloe. I am currently conducting a university research study considering if various golf courses can compete with nature reserves as a location for conservation and sustainability. I'm looking at 3 courses and 3 reserves in my area.

I know this is the under university section but I'm really terrible with maths!

One branch of primary data I'm collecting is the levels of animal species richness and abundance.

I have the numbers from my observations but I'm struggling to provide statistical analysis. I want to test for statistical significance between different species observed at both (types) of site. I understand the concept and what it means but all the instructions I've read ask me to choose what I'd accept as a level of error or am I doing a t test or a chi squared test yada yada

Would someone be willing to help me with the maths of it? And suggest the type of data testing method that would be suitable for my data set? It'd be much appreciated!

Thankyou

x Chloe x

2. So I hear that you want to test if the observed numbers (or ratios) on these golf courses are similar to what is expected in a nature reserve?

I could help, how much data do you have?

3. Originally Posted by Bikkembergs
Hello,

My name is Chloe. I am currently conducting a university research study considering if various golf courses can compete with nature reserves as a location for conservation and sustainability. I'm looking at 3 courses and 3 reserves in my area.

I know this is the under university section but I'm really terrible with maths!

One branch of primary data I'm collecting is the levels of animal species richness and abundance.

I have the numbers from my observations but I'm struggling to provide statistical analysis. I want to test for statistical significance between different species observed at both (types) of site. I understand the concept and what it means but all the instructions I've read ask me to choose what I'd accept as a level of error or am I doing a t test or a chi squared test yada yada

Would someone be willing to help me with the maths of it? And suggest the type of data testing method that would be suitable for my data set? It'd be much appreciated!

Thankyou

x Chloe x
Do you have a supervisor for your research project?

4. Originally Posted by Bikkembergs
Hello,

My name is Chloe. I am currently conducting a university research study considering if various golf courses can compete with nature reserves as a location for conservation and sustainability. I'm looking at 3 courses and 3 reserves in my area.

I know this is the under university section but I'm really terrible with maths!

One branch of primary data I'm collecting is the levels of animal species richness and abundance.

I have the numbers from my observations but I'm struggling to provide statistical analysis. I want to test for statistical significance between different species observed at both (types) of site. I understand the concept and what it means but all the instructions I've read ask me to choose what I'd accept as a level of error or am I doing a t test or a chi squared test yada yada

Would someone be willing to help me with the maths of it? And suggest the type of data testing method that would be suitable for my data set? It'd be much appreciated!

Thankyou

x Chloe x
When testing for significant difference it is usual to use 95% as the confidence level (that is if there is no real difference you would identify a difference erroneously 5% of the time).

CB

5. Hi,

Thanks for the replies

Mr Fantastic - I do, but the statistics documents he gave me I just cant get my head around Plus he's difficult to talk to even when having a normal conversation.

Pickslides, yes thats generally the idea. Im looking at various other similar studies. One is comparing the animals on golf courses, forests and residential areas. I've extracted the key results info for you to read rather than giving you the full article (My results will be of similar numbers when completed and also observing richness and abundance)

Total birds observed = 16,585
Total bird species observed = 138

Species richness:
Golf courses X=53.2
Residential areas X=33.4
Forests X=48.0

Stat significance for species richness :
Between GC and RE = P<0.0001

Relative abundance of birds:
Golf courses X=452
Residential areas X=361
Forests X=337

Apparently he calculated the species richness estimates using the Bootstrap method. He calculated the relative site abundances using relative index.

Once he had these estimates he then compared the species collectively between land types. The mammal estimates using T tests and ANOVAs with Tukey HSD testing for birds.

He finishes by comparing individual species between land types. Comparing birds using Kruskal Wallis ANOVA for birds and Mann-Whitney U tests for mammals.

So much information, is this overly complicated? I'm not trying to demonstrate I can interpret statistics via 5 or 6 fancy ways - Once I'd demonstrated the understanding and justification of a statistical test then 1 or 2 would surely suffice?

Lets go back to his stats. Where has he got these figures from? I cant see the sums he has done to acheive the numbers from 16585 birds over 138 species...? He must have left out some raw data too because he's mentioning richness and abundance figures for mammals too, and he hasnt mentioned the raw figures he observed for mammals.

Captain Black, where does p<0.0001 fit into 95-5% as a confidence level?

I thought significance was about reliability? And a difference of >0.05 was significant?

Thanks for helping, xChloex

6. I am hearing there has been lots of different tests for lots of different data sets.

This is the case given the distribution of each set of observations.

When p<0.0001 then there is a signifcant difference between the two data sets.

7. Originally Posted by pickslides
When p<0.0001 then there is a signifcant difference between the two data sets.
So what does that mean/signify? And what would it mean if it was no significant difference?

8. Originally Posted by Bikkembergs
So what does that mean/signify? And what would it mean if it was no significant difference?
It is clear that you have very little understanding of statistics and are currently out of your depth with what this research task requires. You need to discuss this with a live tutor, ideally your supervisor. It is his/her job to help you understand what you need to do and why you need to do it.