Stats Analysis

[amazingtoxic]

I am starting to analyse the results for my dissertation and am a bit stuck on how to analyse them, mainly thanks to my disstertation supervisor, who is really not interested in helping me. I have worked out that I am doing the paired samples t-test but what I am not sure of is how to entre the data. I have two sets of twenty-five results for each person (30 subjects) and I want to anaylse the difference bwtween each set. Do I entre all 25 pieces of data for each person or do I take the mean of each twenty five?

[Bev2007]

Hi toxic

I am assuming the pairing will be between corresponding values in the 2 sets for each subject. For example, Subject 1's first value in Set 1 (lets call it x1) is paired with Subject 1's first value in Set 2 (x2). Because you are pairing you need to calculate the difference between the pairs (x1 - x2)and will then work out the mean of your 25 calculated values.

If you had been doing a straight t-test you would calculate the two means first, but not when you are pairing.

Good luck

[Bev2007]

perhaps this will help - has a worked example....

http://helios.bto.ed.ac.uk/bto/statistics/tres s5.html

[StudentGal]

it depends what sort of data you have, qualitative or quantitative. I am also doign my final year project and have found the book 'Essential mathematics and statistics for science' by Currell and Dowman fantastic. Get a copy of it and read it. Should explain what you need to do.

i would suggest using software to do your stats though like minitab for quantitative or SPSS for qualitative.

Good Luck :D