Hey, statistics question for people who know about data science---I have a few thousand .csv rows. Each one has a timestamp, a speaker ID, the speaker's province, and some quantitative data. If I want to test for patterns in each of those data, over time and between provinces, I'm using a mixed-effects model, right? Will a linear ME model be enough, or (assuming there are dips and swells in the data, not just a linear progression) do I need something else?