I attached the results some one please tell me that which value i have to see in the result to know whether it is significant or not. Is it that which i marked in red?
I guess t-test is a parametric test, and 8 subjects can hardly fulfill the requirements, why not try Mann- Whitney-U test instead?
Dear Alex Russell, as we are discussing the assumptions, I need a concise, simple enough and autoritative enough to be cited guide for assumptions of statistical tests (especially t-test, ANOVA, correlations, regressions and the non parametric counterparts), because I am having difficulty in reaching good-enough sources. Can you recommend a source like this (preferably a book or article).
By the way, this is somehow off-topic, yet since it is related to the assumptions I wanted to ask the question under the thread, and your answer may be useful for dear Mehjabeen Javed.
Dear Cinar Kaya,
You can go through the book entitled "Methods in Biostatistics" by B. K. Mahajan, a very comprehensive & yet easy to understood. It is especially designed for medicals.
Dr B. K. Sharma
Associate Professor (Biostatistics)
Sri Aurobindo Medical College & P. G. Inst, Indore, India
I'm with Alex - as long as your assumptions hold, your t-test is valid. I personally avoid non-parametrics so long as assumptions hold or a single transformation fixes my data.
Additionally, for Cinar - a source I would recommend that I used in preparing for my qualifying exam is:
Underwood, A.J. (1997). Experiments in ecology: Their logical design and interpretation using analysis of variance. Cambridge University Press.
table on bottom. Sig value on right hand side is 0.000 which indicates Alternative hypothesis supported at 5% level (below 0.05) and also 1% level (below 0.01) therefore the two variable means are statisticall different (at p
- Badges
- Science topic