What test to use for longitudinal data and to test interactions among factors?

#1
My study involves five plant varieties subjected to three cultivation methods in two different sites. Every week for six weeks, we record the specific daily growth rate of all the plants. I’d like to analyze differences in growth rates in terms of variety, cultivation method, sites, and time (six measuring times). I’d also like to look into interactions among these four factors. I initially thought of doing factorial ANOVA until I came across repeated measures ANOVA. I’m now confused about what statistical method to use since my growth rate data are longitudinal data or repeated measures data given that the same plants are measured every week. furthermore, after testing assumptions for anova, i found that my data violates normality and homoscedasticity. What is the appropriate analysis for this?
 
Last edited:

Karabiner

TS Contributor
#2
I’m now confused about what statistical method to use since my growth rate data are longitudinal data or repeated measures data given that the same plants are measured every week.
A "mixed" ANOVA with a repeated-measures factor and several
between-groups factors would be woth thinking of. How many
plants are there in each cell? And how did you record "growth
rate", is this a percentage?

With kind regards

K.
 
#3
Growth rates are in percent per day. I have unequal cell sizes because some of the plants died during cultivation. If I consider measuring times as repeated measures (falling into one cell), n ranges from 27 to 60 plants per cell. Sorry I'm not really familiar with statistics.