Statistical testing

[amazingtoxic]

I have to analyse the knee angles of a subject when performing a vertical jump.

I have 6 different sets of data I need to analyse and compare. What statistical tests do I use? I was thinking the ANOVA test but would i need to use the post-hoc comparison?

Thanks H

[Gnu]

What is the hypothesis that you are testing, and what data have you collected? I think that both of these are relevant in order to decide on the most appropriate statistical test.

[amazingtoxic]

The use of an arm swing does affect knee angle and velocity parameters during a vertical jump performance.

We recorded six jumps, 3 with arms, three without and have to compare the knee angles and volocity of the jumps with arms compared to the one without.

So I have six pieces of data for the arms swing and six pieces of data for velocity.

Thanks

[Gnu]

I think that you would carry out comparisons for the knee angle and velocity paramenter separately.

The data has a reasonably normal distribution, I'd have thought, with two conditions (arms used, or not). The hypothesis would be that the means would be different.

If the jumps are in matched pairs (e.g. one jumper with arms, and then again without), the t-test for correlated samples could be used, otherwise the t-test for independent samples.

It seems to me that 6 items of data would not give enough information to reach statistical conclusions, if I've understood your situation correctly.

A useful book on statistical design and methods is 'Experiment, Design and Statistics in Psychology' by Colin Robson, published by Penguin. It is easy to use and relevant to a wide range of social sciences.