Is ANOVA the way to go?

#1
SOCIAL MEDIA (SM) STUDY - My study is to find out if thought polarisation in adults is affected by type and amount of SM use and age and sex

Determining the AGE and SEX of participants
Determining the AMOUNT of SM use by questionnaire (low, medium or high)
Determining a POLARISATION score by a questionnaire (1-10)
Determining the TYPE of SM use by a tendency to more echo chamber type use (polarised) or more random content (balanced)

Then analysing to see if less SM use relates to less tendency towards polarisation and if this is affected by age and sex and or the type of social media use
For example if people have high social media use but a low polarisation score is that because they frequent more balanced groups?
OR if they have low social media use but a high polarisation score is that because they tend to frequent polarised groups

I need to know if ANOVA is the suitable method of analysis
and the Dependant and Independent variables?
 

Karabiner

TS Contributor
#2
You could use multiple linear regression with the polarisation score as the dependent variable.

Sex and type of SM use will have to be recoded, e.g. dummy (0/1) coding, also amount will have
to be recoded, maybe into 2 dummy-variables (e.g. one for medium, one for high amount,
while low amout will receive no dummy and will serve as reference group).

You described age and sex as possible moderators, so you'll need to include the interactions
(age*type, age*medium use, age*high use, sex*type, sex*medium use, sex*high use).

HTH

Karabiner