Is it appropriate to compare normalized data between treatment groups if each is not different compared to it's respective contr

1 post / 0 new
Susan Murphy
Susan Murphy's picture
Is it appropriate to compare normalized data between treatment groups if each is not different compared to it's respective contr

I am not sure if our results are meaningful. We are comparing the molecular effects of blast injury to the brain. We have 2 different blast/control groups:<!--break-->The animals receive a single blast, with their control group in which the animals receive the same treatment without the blast (e.g., anesthesia, etc.).<!--break-->The animals receive 3 blasts, 1 hour apart, with their control group.<!--break-->N=10 in both blast and control groups per treatment.<!--break-->When we compare the data (ANOVA with either Tukey's or Bonferoni), p>0.05 for blast vs. control for both groups and when comparing the 1x vs 3x controls. But when we normalize the data to its respective control and then compare the 1x vs 3x blast p<0.05.<!--break-->My question is if neither treatment is different compared to its control does it make sense to compare between the blast groups?<!--break-->Also is the result meaningful for the 1x blast to be significantly different to the 3x blast if neither is different compared to its control?