Biology, images, analysis, design...
Use/Abuse Principles How To Related
"It has long been an axiom of mine that the little things are infinitely the most important" (Sherlock Holmes)

Search this site



Just a note

Irrespective of how close to normal the transformed data are notice that, by pooling the data in this way, we are implicitly assuming all these results represent random samples of the same population. For these data, as with many other time series data, this assumption is likely to be unreasonable for two reasons.
  1. The biological population is likely to have changed over time - so we cannot possibly be estimating a single 'population' mean (or geometric mean) catch.
  2. Consecutive sample are likely to be correlated - so cannot possibly be random.
Although in this instance we are dealing with data from a single trap, even if we consider catches of several traps on the same day, pooling their results in this way is only reasonable to the extent we consider them to be taking random samples of the same population. If the traps are widely separated, or are in very different vegetation types, we cannot pretend flies from the same population are equally likely to be caught in each trap.

In either case we are not dealing with random samples of a single population, and treating our results as if they were is simply misleading.