May I perform Pearson/Spearman correlation between the ∆ of some interval parameter (e.g. the dynamics of BP) and the ∆ of another interval parameter (e.g. HR) if the ∆ between 2 time-points is not statistically significant?
Hello Mikhail,
Yes, you may correlate differences in scores across time for each of two interval-strength variables.
The fact that the mean difference for a given variable isn't different from zero doesn't pre-determine anything about the potential relationship between difference values across two variables.
One concern which may be of interest is, if either of the scores is less than perfectly reliable (e.g., has some measurement error) then score differences will be less reliable than either of the individual values from which the differences were computed.
Good luck with your work.
Thanks a lot. I was wondering why there is a sense in correlating these variables. If I have no diffrences during an experiment over time, how I interpret the results of (lets say) positive and significant correlation coefficient?
Hello again Mikhail,
In such a circumstance, you would conclude that mean scores don't change, and that persons' scores are (at least somewhat) stable over that time frame. That is, people who had high initial scores tended to have high ending scores, whereas people who had low initial scores tended to have low ending scores.
Good luck with your work.
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology