I have a dataset that tracks barnacle size (shell length - dependant variable) growing on 3 independant buoys (that are identical). The measurements are taken every 1 week interval (independant variable). I want to find out if the growth of my barnacle is mainly influenced by time and that my 3 replicates (buoys) are not an influencing factor.
I was about to run an ANOVA but then realised that an assumption of the ANOVA is normally distributed dependant variables (in this case shell size). However shell size increases with time and this does not conform to normality and breaks the assumption of the ANOVA.
My question - is it appropriate to use an ANOVA to test for variation between my three media replicates (Buoys). Or should I look to some other statistical test. If the later any direction would be much appreciated.