chocolate11001 Posted March 10, 2013 Report Share Posted March 10, 2013 (edited) I have outliers in my data according to Moore and McCabe when i use their criteria for determining outliers Q1 - (1.5 x IQR)I was wondering if it would be okay to remove them when i do the inferential statistics bit Edited March 10, 2013 by chocolate11001 1 Reply Link to post Share on other sites More sharing options...
lizzie james Posted March 10, 2013 Report Share Posted March 10, 2013 My IA had a massive outlier and it made the inferential stats accept the null hypothesis, which was annoying.. but our teacher said not to get rid of them because it gives you more to discuss. You can comment on the faults in your method and extraneous variables, sample bias.. ect. If you analyse why your results didnt agree with all of your research it shows that you understand the limitations of experiments and a lot more easy to come up with improvements for the method 2 Reply Link to post Share on other sites More sharing options...
Ezak Posted March 20, 2013 Report Share Posted March 20, 2013 Don't remove it, discuss why it is present, if it is due to error or if some kind of information may be extracted from it. Reply Link to post Share on other sites More sharing options...
carpediem Posted March 20, 2013 Report Share Posted March 20, 2013 Just to reiterate, I say don't remove them. My outliers allowed me to evaluate the methodology more, because my experiment did in fact cause stress to the participants, (because in fact I had a withdrawn participant, which indicates that the level of stress was too much for them to handle etc, etc). And if you're low on words, the outliers will help you fill the space. Reply Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.