I am trying to find the relationship among 3 variables: values orientation, awareness of environmental threat (as IV) and ecological behavior (pro-environmental behavior) as DV. Any ideas on the best statistical tool to use? Values has 3 domains treated individually, by the way.
I also wish to look for interactions among all variables.
Thanks for the help!
The first step is to determine what kind of variable your DV is. If you are treating it as ordinal, then ordinal regression is the best appropriate method. It is fairly easy to perform at least in some software packages, including R.
If you are able to treat your scale DV as a continuous variable*, then you can use common multiple regression/ANOVA.
If you think your DV is count or proportion data, then you would use something different.
In any case be sure you understand the assumptions of each approach, and what kind of data is appropriate.
* For example, if your scale is made up of several Likert items, and the results have many values, then you might be able to treat it as continuous.
If your scale has traditionally been treated as interval within your field, than simple regression would be appropriate. In particular, regression can handle four separate IVs with more clarity than ANOVA.
The answer depends on your variables level of measurement and if you are interested to find Relationship vs. differences.
So, if your DV is continious (interval or ratio) and you are interested to study relationship iwth more than one IV, then you need to use multiple linear regression.
if you have only one variable as IV with 3 grous or more, and the DV is continious, in this case you need to use ANOVA.
- Badges
- Science method