I can’t speak to your first question, but for your second question, the first common factor identified will explain the most variance in your measured variables, with additional common factors explaining additional variance, so you’ll typically need to determine how many common factors (and not which measured variables) to retain. The factor loadings are simply representing the strength of a given factor on a measured variable, with the rotation producing a simplified structure that aids interpretation of the data. So as you might expect, the underlying factor affecting the battery of one set of measured variables won’t have much influence on variables outside of that battery (and ideally it will have none – one of the assumptions of factor analysis is that latent factors are orthogonal). This means you should see strong loadings on one variable from one component but ideally corresponding with weak loadings from another component. Just make sure you understand the use case for factor analysis; you’re trying to find the latent factors influencing the variables you can measure, with the hope that common factors are underlying sets of those measured variables (and not the other way around – that would require a different approach).