Experiment VS Observational study
All self tracking if done by many people would be a longitudinal study.[1]
Experimenting is to formally state spell out and control for as many variables as you can. Science. Because user makes sure to change nothing else during the experiment many issues with analyzing time based relations do no happen. Consequently linear regression, ttest, causal analysis and breakpoint detection work. Maybe some complications in establishing baseline. See T-test. If lots of people did this at the same time you would have a crossover study[2].
Second type is observational study. Just monitoring all the variables without a specific structure or interventions. This is used for sending data to research databases, establishing baseline for experiments, catching problems as they start, monitoring chronic illnesses and calibrating every day things like amounts of coffee. Also may remove placebo and stress or nervousness effect from a more formal experiment. Unfortunately this is where all those issues I mentioned in Finding relations between variables in time series and many more in the links apply.
When you are done, do not forget to visualize. Visualizing your Data
Linked content on this wiki
(The content in the table below is automatically created. See Template:Topic Queries for details. If newly linked pages do not appear here, click on "More" and "Refresh".)
| Tools related to this topic |
|---|
| Projects related to this topic |
|---|
| Self researchers related to this topic |
|---|
| We talked about this topic in the following meetings |
|---|