This is the meaning of factor analysis that you delete those correlation coefficients below 0.5. Then you build your different factors and name them e.g. autonomy, intrinsic reward. You can do a Cronbach's alpha of each factor and delete those items that make your factor less than alpha 0.7 which is an acceptable internal reliability.

Then you use the factors in regression analysis. Exempel:

Furthermore, read the qualitative section of the attached paper and you can see what people really think and you can also use this idea: