I am looking at age of people in 10 distinct groups. The number of persons in each age bracket is noted and there are 10 age brackets. The total number in each group is quite varied but data is normally distributed within each group.
I am looking to compare one group (in blue in the attachment) against the others to see if if the age brackets are significantly different.
I am struggling to work out how to do this and would be grateful for ideas! Thank you in advance.
Hi James. Hope you're well. Stats are definitely not my strong point so take my advice with a pinch of salt. However, it looks like you've got a number of categorical predictors (the age brackets) with quantitative outcomes. From what you've said the data within each group is normally distributed and the variance is likely to be similar. If so, you could compare the groups with ANOVA. If significant this will only tell you that one of the groups is significantly different and not which one. You could then use a post-hoc test like Tukey's to compare between the groups to see where the significant difference lies.
I think you can use an ANOVA with age groups between 35 and 54. Due to the few cuantities in the other groups. It would be useful to use non-parametric Kruskall-Wallis.
Thank you both - I will try to get something meaningful with ANOVA.
Jameel Muzaffar - nice to hear from you. Hope you are well too! Must be a Prof by now?!?
- Badges
- Science topic
- Similar topics
- Developmental Psychology
- Attachment