I'm trying to work out if LCA is the appropriate analysis for the data we have. We are trying to determine whether particular types of advertisements have better/worse effects on attitudes and behaviours. To do this we have coded each advertisement on a number of variables (i.e. message, approach, execution), and want to know whether a certain combination of these factors is associated with a better attitudinal outcomes.
The data comes from cross-sectional surveys conducted over a number of years, where one advertisement was evaluated in each survey. Is LCA the appropriate analysis for this style of data?