1

I am writing a thesis and have been advised to use 95% CI to determine if i have significance between my data. I have been told to use a scatter plot and place CI as error bars. I am unsure how to describe my statistical significance in my results section below my graph. Is using CI enough to validate if there is significant difference between my samples? Do i not need to use further statistical analysis (my sample size is very small and my data is not normally distributed after transformation - log, arcsine, boxcox)

claire
  • 23
  • 4
  • 1
    The short answer is no, but this is a rich and complicated subject. Consult our threads about [comparing CIs](https://stats.stackexchange.com/questions/18215) and [what CIs mean for replicability](https://stats.stackexchange.com/questions/218602), for instance. I'm guessing that you might actually need some kind of Analysis of Variance (ANOVA) or regression procedure, but it's hard to determine that just from knowing that you have data! – whuber Apr 11 '19 at 13:26

0 Answers0