|
Title: How to calculate standard error for % change and/or percent difference? Post by: sc_val on Jun 19, 2015 Hi everyone,
my research question is whether genetic manipulation has a higher effect in male mice than in female mice (or in another experiment, whether treatment has a higher effect in males than in females). We measures pain scores in males and females in three genotypes each (so 6 groups) and I would like to compare for each sex gentype1 vs genotype2, and genotype1 vs genotype3, and then finally compare these differences between sexes (i.e. is the difference in genotype1vs2 greater in males than in females?). I thought it'd be best to report the differences in pain scores as a percent change, but am unable to calculate the corresponding standard error. I found a pdf online (attached) The STEs I get are incredibly high with this formula and probably incorrect. Does anyone know what the right way of calculating the standard error for percent change would be? Some more info on my data: I calculated the % change from means, i.e. ((mean female_genotype1 minus mean female_genotype2)/mean female_genotype2)*100), because not each genotype1 mouse has a corresponding genotype2 mouse. I tried implementing this in SPSS unsuccessfully, so calculated these % changes in excel. Thanks for any help and sorry if I posted this in the wrong thread (I did check if there was a similar question in here already :-)). Title: Re: How to calculate standard error for % change and/or percent difference? Post by: bio_man on Jun 20, 2015 Have you tried ANOVA?
|