# Linear models and multiple comparisons on time-series data

I have conducted a study on subjects corresponding to two groups G1 and G2. Each subject in the group was tested under two conditions C1 and C2. The testing under each condition was done across a period of time of several hours, taking one record every ten minutes. Our hypothesis is that there is an interaction between Group and Condition, the latter affecting the dynamic of the time course differentially between the groups. I have constructed a linear mixed model on R in this way:

model = lme(var ~ time + condition*group, random = ~1|subject, data = data1)


I analyzed the results of this model, which revealed a significant interaction between Group and Condition. (I also compared the previous model with a similar one in which time was included as another interacting term, but the former came up as the best fitting model).

So after that, I thought that it would be elegant to set some post hoc multiple comparisons to determine which time-points show differences according to the conditions, within each group. Now, it doesn't seem practical to me to compare each of the ten-minute bins throughout the testing period, because they are numerous and not necessarily relevant in their own; I figured I could compare hourly intervals.

So here is my question. I can think of two ways of doing that: 1) Assign each time-point datum to an hourly interval and build a model and the multiple comparisons from it; or 2) Average or sum all time-points data from an hour and build the model and comparisons from the new dataset.

I figure option 2 is the most straightforward but it would significantly simplify my dataset and reduce the power of the analysis, or at least that's what I feel. My problem with option 1 is that I don't know whether each 10-minute datum can be considered as a replicate datapoint in an hour interval.

Can anyone help me with this matter? Also, please let me know if my reasoning is wrong from the beginning. I've just started learning the ropes of this kind of analyses and my base understanding could be already wrong. Also excuse my grammar as only speak English as a second language. Thanks to all.