It depends on how many data-points will be counted.

If it's a single pre-post  test, you could use Repeated Measures t-test (if the data are parametric) or Wilcoxon test (if the data is not parametric). But if you are going to include more data points, than One-way Repeated Measures ANOVA (parametric) or Friedman's test (non-parametric) should be used.

Hi Stergios,

So you would essentially use the same tests as a normal repeated measures experiment with many people?

Just to make sure my question was clear, I'm talking about a study with a single subject (N=1). With, let's say 30 data-points in one condition and 30 in another condition.

Hi Joshua, yes, since it's n-of-1 and not independent, I would use the repeated-measures test (actually that's what I'm doing also in my work right now).

If you want to use all the data points, you are in the second option. If you are going to use the means of each period, in the first.

There have been suggestions to use time-series approaches in the literature, but the works of Huitema I think gives the correct view. The simpler the better! Check the books:

http://link.springer.com/book/10.1007%2F978-1-4684-8786-2

especially this chapter for a shortcut:

http://link.springer.com/chapter/10.1007/978-1-4684-8786-2_10

and

http://onlinelibrary.wiley.com/book/10.1002/9781118067475;jsessionid=335DD82ED372BD6C683186061032F9CF.f03t02

Also, a very simple introductory is that one:

https://uk.sagepub.com/en-gb/eur/how-to-design-and-report-experiments/book219351

Don't think of it as useless, the PART TWO: ANALYZING AND INTERPRETING DATA is one of the simplest step-by-step text I have seen on deciding what test to choose based on your data and design.

Thanks a lot Stergios. Time to hit the books.