Bloody statistics again...

nearly_there

Ok its that time of year again where Im beginning to write up my thesis. Statistics, my old enemy is back to haunt me


Im just looking for a bit of advice re this. Im using SPSS which is very user friendly but just to make sure Id like a pointer please.

I have 14 time points, at each of these I have a measurment taken from 6 different treatments. This is an ANOVA right?

and i want to compare each at the end so a tukey test? assuming data is normal/normalised.

but its not a two-way parametric ANOVA. does a 3/4/5 etc way anova exist?

2Scoops

nearly_there wrote: »

I have 14 time points, at each of these I have a measurment taken from 6 different treatments. This is an ANOVA right?

Yes, assuming scale variables.

nearly_there wrote: »

and i want to compare each at the end so a tukey test? assuming data is normal/normalised.

Choice of test depends on many things; you might want to pick the one best suited to the mean comparison in question, repeated measure, treatment effect, variance etc.

nearly_there wrote: »

but its not a two-way parametric ANOVA. does a 3/4/5 etc way anova exist?

You have a 2-way with repeated measures on time - a mixed model of sorts. There are such things as 3-, 4- and 5- way ANOVAs but they are so difficult to interpret that you don't see them much past 3-way. Make sense of a 5-way interaction anyone?? :0

nearly_there

Thanks for your reply!


I ended up doing many 2-way ANOVAs. Time consuming but I really knew what I was doing by the end of it!

Am getting to grips with SPSS now, but no doubt will be back for more help in time !

tayto potatoe

Hey,

I need help on an SPSS assignment. Am looking for an hour or so tutorial, will pay.

Thanks

Ostrom

What exactly are you doing?