ANOVA help - comparing growth rates between groups with 3 replicates in each group

#1
Hi

I'm really pretty lost, statistics have never been my strong point. I need to test whether my growth rate differences between trees are significant or not. I tried the ANOVA: 2 factor with replication on excel, but keep getting an error so I'm not sure if my data is suitable. SO first off the test; I had 6 plants of 10 different citrus cultivars, for each cultivar, 3 plants were inoculated with a virus and 3 were used as the healthy controls. I measured the shoot length every month over 1 year and worked out the growth rate (cm/day) for each growth period (month). Some values are zero when the plant did not grow for that period. I use growth rate rather than shoot length because the shoot length is variable according to the plants. So I need to establish if the growth rate of, for example Mexican Lime (a citrus cultivar), differs significantly among the three replicates that were inoculated with the virus, and if the growth rate between the inoculated plants and inoculated plants differ significantly or not. Also, some plants died during the trail so some groups would have 3 inoculated trees and 2 controls.

What is a suitable statistical test for this? Is it available on excel? I did download XLSTAT. If not is it available on other free software? Or can I do this manually?

I will really appreciate any form of help or suggestions, THANKS!
 
#2
Re: ANOVA help - comparing growth rates between groups with 3 replicates in each grou

So, I'm having a little trouble telling what the structure of your data is. Correct me if I'm wrong.

10 sets of 6 plants
60 plants total
30 inoculated, 30 control

If we are limiting ourselves to the Mexican Lime, e.g., there are 6 plants w/ 3 inoculated and 3 control. For each plant you have a set of ~12 growth measurements. So, for the Mexican Lime, e.g., we have 6*12=72 data points. We want to compare the 3*12=36 inoculated growth rates to the 36 healthy growth rates.

To compare the difference of mean growth rate for each group of Mexican Lime, I suggest Welch's t test.
http://en.wikipedia.org/wiki/Welch's_t_test