Hello,

So I have asked this before, and it seems like the more I dig into this, the more suggestions I receive for different tests!

My research question is looking at different rooms in a building with a green roof, trying to find if there is a significant difference in temperature between rooms under the green roof, and not under the green roof. My data consists of date on the x-axis, and time on the y-axis. What would be the best way to test for significance between the temperatures of the rooms over the testing period? I tried using one-way ANOVA, and then Tukey's HSD test, and got some reasonable results, but I am not sure if I chose the right test.

I am truly confused about how to test my question. I was told that ARIMA would be appropriate since I have autocorrelated time series data, but this would produce a single result, basically just saying whether or not the green roof had a significant impact on temperature. However, what I really like about the Tukey's HSD test after ANOVA is that I can compare individual rooms. I believe that when I give the results to the building manager of the building I studied, it would be more interesting and useful for them to be able to compare each room to each room, so they can look into how cool individual rooms are based on the green roof. It was easy and straightforward to carry out the ANOVA with Tukey's HSD, but I just do not know what to do about the autocorrelation. I know that it seems odd for me to treat each day over the 3 week testing period as a category, but I am not sure what else to do.

Please understand that I am a novice at statistical analysis and would appreciate easy to follow steps! I am working in R, because it is the only statistical program I know.

I attached an image of my data below for reference.

Thank you!