0
$\begingroup$

I'm studying structure-function relationships in neurons and mainly focusing on correlations between dendritic length and excitability. My question is:

When searching for correlations, which multiple comparison correction should I use to classify which of the correlations I find are significant?

I have measured dendrite length (apical and basal) which provided me with 3 values: apical dendrite length, basal dendrite length and total dendrite length. I've also measured 5 different electrophysiological paramaters within the same cells. My research question is: Are the electrophysiological parameters related (correlated) with the dendritic length? This yields 3×5 comparisons (3 length parameters with 5 electrophysiological). I have found some significant correlations, as expected. Now I want to find out whether the significance of these correlations must be corrected, and if so, which method I must use. Bonferoni correction seems a little too conservative, because correlations between structure and function of a neuron seems logical, and it has already been established that they exist.

$\endgroup$
  • $\begingroup$ You'll need to provide more detail. There are lots of possible ways to correct for multiple comparisons, each w/ different pros & cons. $\endgroup$ – gung - Reinstate Monica Jun 10 '14 at 0:03
  • $\begingroup$ Alright, I have measured dendrite length (apical and basal) which provided me with 3 values, apical dendrite length, basal dendrite length and total dendrite length. I've also measured 5 different electrophysiological paramaters within the same cells. My RQ is: Are the electrophysiological parameters related (correlated) with the dendritic length. This yields 3x5 comparisons (3 length parameters with 5 electrophysiological). I have found some significant correlations, as expected. Now i want to find out whether the significance of these correlations must be corrected. $\endgroup$ – user48054 Jun 10 '14 at 0:19
  • $\begingroup$ And if so, which method i must use. Bonferoni correction seems a little too conservative because correlations between structure and function of a neuron seems logical and it has already been established that they exist. $\endgroup$ – user48054 Jun 10 '14 at 0:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.