Multiple Correspondence Analysis

[Daniela Rozzi]

Hello everyone,

I hope someone could help me.
I have a dataset with 154 observations referred to individuals who have participated in a project and for these subjects I have:
- the reason they left the project (1=objectives achieved; 2=other reasons)
- sex
- citizenship
- title of study
- employment
- age

My goal is to see if there are some characteristics which are most related to people who left the project because achieved objectives.
In order to show relations among the different modalities of these variables I was thinking to use the Multiple Correspondence Analysis.

mca reason sex citizenship age title of study employment
mcaplot, overlay origin


Do you think it make sense to use MCA? Do you suggest something else?

Another question, do you think that I should put some of these variables as supplementary?

Thanks in advance,
Daniela

[Carlo Lazzaro]

Daniela:
provided that I'm totally unfamiliar with -mca-, I would consider -logit- (coding 1=objectives achieved; 0=other reasons):

logit reason i.sex i.citizenship c.age##c.age i.education i.employment

[Daniela Rozzi]

Thank you for suggestion!
I tried but I'm not sure how to interpret these results (I'm not really familiar with that).



In addition I was thinking to use the MCA because I would like to show categories relations in a plot.

[Carlo Lazzaro]

Daniela:
please note that screenshots are deprecated on this forum for wise reasons well exaplined in the FAQ.
in your future posts, please use CODE delimiters to share what you typed and what Stata gave you back (as per FAQn again). Thanks.
That said, your output tabe tells you that an increase of 1 year of age increase the log odds of leaving the project upon its accomplishment by 0.9112 (other things being equal).
Testing if adding a squared term for -age- (età) makes sense, too.
Eventually, with 123 observations it is hard that you can get memorable results from your dataset.
As an aside, I cannot be helpful about -mca-.

[Daniela Rozzi]

Thank you for your explanation.

I'm sorry to have used screenshot...I will use CODE delimiters in my future posts.

Anyway, if someone could help me with MCA I will be grateful!

Daniela