Question: I'm not sure whether I have to use parametric or non-parametric tests.
0
gravatar for rahulgenexpress
3.3 years ago by
rahulgenexpress10 wrote:

I have obtained NGS data from stool samples of disease subjects and healthy controls, performed QIIME analysis. To compare gut microbiota composition and to know significant difference, which statistical tests I should perform?

next-gen • 1.3k views
ADD COMMENTlink modified 3.3 years ago • written 3.3 years ago by rahulgenexpress10

You should ask your question to http://stats.stackexchange.com/ as it's not really a bioinfo question

ADD REPLYlink written 3.3 years ago by Nicolas Rosewick8.3k
1
gravatar for EagleEye
3.3 years ago by
EagleEye6.5k
Sweden
EagleEye6.5k wrote:

If your data follows normal distribution, use Parametric otherwise non-paramertic.

More information,

http://blog.minitab.com/blog/adventures-in-statistics/choosing-between-a-nonparametric-test-and-a-parametric-test

Finding the distribution,

http://www.originlab.com/doc/Origin-Help/Normality-Test

Perform normality test using R,

http://m.dummies.com/how-to/content/how-to-test-data-normality-in-a-formal-way-in-r.html

ADD COMMENTlink modified 3.3 years ago • written 3.3 years ago by EagleEye6.5k
1
gravatar for agata88
3.3 years ago by
agata88790
Poland
agata88790 wrote:

It depends how many samples you have for analysis. Sometimes if the number of samples is small the distribution of this group is not normal. I had the same problem with my analysis. One group was normally distributed and second was not. I wanted to find statistical difference between groups. In that case I used two test: student test and Mann-Whitney test. Student test for normally distributed data, and Mann-Whitney test for not normally distributed data. I am not 100% sure that is correct solution but as far as I know it is not necessary to do a normalization of not normally distributed group, since there is small amount of samples (the biggest amount of data the more probably that group have normal distribution). But in case you have a large number of samples and still your data is NOT normally distributed I would go straight for non-parametric test.

What is your opinion about that?

I am also looking forward for some tips.

Best,

Agata

ADD COMMENTlink written 3.3 years ago by agata88790
0
gravatar for rahulgenexpress
3.3 years ago by
rahulgenexpress10 wrote:

Thank you very much Agata. As per your suggestion I checked my data, it has large number of samples and not normally distributed; I am going for non-parametric test.

ADD COMMENTlink written 3.3 years ago by rahulgenexpress10
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1653 users visited in the last hour