Hi!
I have a question about the best way to analyze the data from duplicates in a sampling insects by way of traps. The reasons for use duplicates in the same point are:
1 To avoid the complete lost of information in this point due to the lost of single samples by incontrolled events.
2 To decrease the number of false negatives due to lost samples or wrong emplacement of one of them due to unrecognizable adverse factors (i.e. a plant not known as repellent, etc.)
Usually, the data for every point is calculate as the mean of both counts. But, if the counts are very different (even no captures in one of them) the use of the mean really could be falling the real data if the lower value is artifact.
Any recommendation o reference about?
Thanks
The extra traps represent subsamples but have equal value to duplicate traps. Just use the means and don't worry about the outcome. Otherwise your personal concerns will become type of bias. At least the power of your experiments will increase, even though there is 'pseudo-replication'
You must clearly state what you did and not claim replication. I suggest you seek expert statistical advice
One way to minimize the variance between 2 traps is to have more traps in the same area, say 4 tarps. If this is feasible for you, that would increase the reliability of your data.
I think treating these 'pseudoreplicates' (I don't think they are really pseudoreplicates) as samples at different time points and analyzing them as separate samples (e.g. repeated measures) would be one of the accurate options to deal with that. Especially if the intent was to cover all possible captures and not to replicate the first sample.
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Invertebrate Zoology
- Arthropods
- Insects