Paired or Unpaired T-test?


New Member
If I have traps set to capture a certain animal at night vs. during the day, would I use a paired or unpaired T-test? I'd be comparing the mean number of species 1 collected during the day with the mean number of species 1 collected at night (using the same trap). Each individual animal is either collected during the day or at night (not both), but it's the same trap used each time and the same species being measured. Does this make sense?
I hope I understand this correctly.

A paired t-test is the most appropriate to use since you are using the same trap at night as in the day. However, if you do not expect that there is a difference in effectiveness between the traps, a two sample ttest should be okay.


TS Contributor
the number of animals captured will be influenced by the location of the trap and the time of the day. If you use a paired t-test then you eliminate the time of the day influence. If you use a simple t test then the variation from location to location will reduce the power of your test, i.e. you will need a greater difference between day and night to see it as statistically significant.