I am currently writing my dissertation on the potential effect of bilingualism on intelligence scores. The latter is represented by a scale on the Y axe. My independent variable is the linguistic status of the participant (Monolingual, Bilingual and Multilingual).
I tried to conduct a simple linear regression on SPSS, but since my IV consists of three separate categories, I obtained a graph with three columns instead of a cloud of points.
Is my calculation still valid? Even if the current graph impacts my R² value?
Is there another way of calculating it or distribute the data for my IV?
I would use ANOVA in this instance. See this discussion:
https://www.researchgate.net/post/Three-means-comparison-by-t-test-or-ANOVA
Yousra Dakhch I agree with Adam N. Letchford It is better to use ANOVA here. Linear regression with two variables is not a good idea because intelligence scores may depend on a set of IVs.
It also depends on your objectives that when only you want to see the mean differences of intelligence scores among three different linguistic statuses and do not have enough data to control for other IVs then you should only use ANOVA here.
Regards,
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs