PSY 370 Lecture Notes - Lecture 22: Factor Analysis, Psy
Document Summary
***a correlation between two items implies that the items have something in common. We call that "something" a common factor that is driving responses to all of our items. But we need to wait for the validation stage to be able to test that. If our items don t fit the model we predicted, we haven"t measured our construct in the way we intended to. When crafting a seating arrangement for a fancy party, you will be more successful if you group people together based on what they have in common. With cfa, we are making predictions about which items should go together. And then we test those predictions with the way that the data actually falls out. *sometimes the things you think your items have in common are not what they actually have in common. F is weighted by the factor loading. How closely related it is to the common factor.
