Correct test?

Correct test?

by Rachel -

Hi! 

I am working through my data and encountering a bit of a Stata learning curve meets insufficient knowledge of statistics. I am hoping to compare two continuous variables by a categorical category. Specifically, I have a difference in survey 1 scores versus a difference in survey 2 scores across three education categories (<= grade 8, grade 9-12, grade >12). This would add p-values to a clustered bar graph.

I ultimately came up with the following command, but I was hoping to run it by someone to be sure it's doing what I think it's doing.
anova educationcat diffipssfreq##diffvpssfreq

 

Thank you so much!

Rachel Edlin

Re: Correct test?

by Chloe Atreya -

I'm not sure I follow the difference in survey scores. Is each survey taken twice by students in each age group?
What are the axes and clusters on your bar graph?

 

 

Re: Correct test?

by Rachel -

Each survey (survey 1 - IPSS and survey 2 - VPSS) is taken independently and then with assistance by each participant. The difference in those scores, or "error", is the y-axis.

The difference for each instrument (IPSS and VPSS) is then clustered by education category.

Thanks for your help, hopefully that was more clear?

Re: Correct test?

by Chloe Atreya -

Okay, becoming more clear. Are you trying to determine if "error" for IPSS vs VPSS is significantly different within a given education category, or are you trying to make comparisons between educations categories? Can you state the question that you wish to answer?

Re: Correct test?

by Rachel -

Is it possible to answer both questions?

Thanks again for all of your help!

Re: Correct test?

by Chloe Atreya -

I don't want to misguide you - it will be easier to figure this question out while looking at the data.

Please bring questions about the final project to Tuesday's lab!