Hello,
I've run a Proc Mixed function on a set of data in SAS.
The data was a result of an experiment that measured % viability over time and I
wanted to compare a Large sample lets say 50L to a small sample say 5L. And
compare the % viability between the 2 sizes to see if I get the same answer
there are 8 samples of each size and the same 8 samples are tested every day
over 10 days (day 0 - day 10).
Now, Given that the same bottles are repeatedly tested each day – then doing an
ANOVA without taking account of the repeated measures nature of your experiment
could lead to erroneous conclusions. The problem is that when I test a bottle 11
times over so many days – each recorded value from that bottle cannot be
considered statistically independent from the others. E.g. bottle A, day 0 – the
percentage recorded is ‘high’ – then it is more likely to be quite high on day 2
than to be quite low – hence the two measures are not statistically
independent.There are ANOVA models that can take account of repeated measures –
I have used the SAS procedure Proc Mixed for this. And I have produced a
significant P-Value for size and not for time so I've shown there is an
effect of time but not for size. I have also produced confidence intervals for
each day over time how can I output the data in SAS to be read into R then plot
a graph of the confidence intervals??Also is there a way of doing this whole
analysis in R?Kind Regards,Al
[[alternative HTML version deleted]]
