Hello everyone,
I am currently analyzing bacterial growth in cold water dispensers as part of my MSc thesis. During the experiment, I measured the bacterial counts in the effluent water of 3 different dispensers. From the raw data, it can be seen that the bacterial count in the effluent increases exponentially over time.
I analyzed the effluent water of 3 configurations, all in duplicate (so 6 different dispensers), for 8 times (on day 0, 1, 2, 7, 14, 28, 42, 58).
Right now, I would like to compare the bacterial counts of the configurations, to see whether one configurations has lower bacterial counts than the other configuration. So 1 vs 2, 2 vs 3 and 1 vs 3.
Furthermore, I would like to "validate" the results by comparing the dispensers with the duplicate dispensers. So 1a vs 1b, 2a vs 2b, 3a vs 3b.
I found a few possibilities to do this (like ANOVA and two sample t-test), but I would say these methods would not apply since the data is not normally distributed (and relatively small distribution size (n=8)) and I possibibly need a non-parametric method.
Could you maybe advise my in selecting the proper statiscal analysis?
Thanks in advance,
Best regards,
Frank