1

I am running an experiment in which I have three groups, and three treatments.

The treatments are three different temperatures, and I am testing how they impact DNA degradation in water.

I.e. Group 1: 4 degrees; group 2: 16 degrees; group 3: 27 degrees.

Each run lasts for 5 days, and I sample the water once each day, at the same time.

I end up with 5 data points/time series, and a trend of overall change.

My question is: what model could I use to compare these three groups over time? I can't figure out how to fit the 5 time-series in, with any model. Would it be ANOVA-RM? Or could I even use a linear regression?

Thanks so much!

Bio1211
  • 11
  • 2
  • Perhaps the simplest model is a two-way ANOVA with 3 levels of Temperature, and 4 times. See [this Q&A and links](https://stats.stackexchange.com/questions/463172/statistical-tests-when-sample-size-is-1/463198#463198) for a $2\times 3$ two-factor ANOVA. If Time effect is significant you could use linear (and perhaps quadratic) contrasts to assess the nature of trend over time. If you have only one measurement for each time/temp cell of your data table, then there is no interaction term.// Also you might investigate links in given in the right margin of this page. – BruceET Apr 28 '20 at 09:02

0 Answers0