This is a comprehensive introduction to configural frequency analysis, its rationale, and various modifications. When more individuals than expected display a pattern of characteristics, these individuals belong to a type. When fewer individuals than expected display a pattern they belong to an antitype. This book deals with methods that focus on groups of subjects in contingency tables. A many-faceted statistical method, CFA is useful whenever categorical variables are analyzed. The original contribution of this text is the linking of CFA to log-linear modeling and the general linear model, enabling the reader to relate CFA to a well-known statistical background.
Okay, it's a statistics book so I wouldn't normally give that 5 stars, but 4 would be undermining its rationale and usefulness. This method is something that every psychology student and researcher must understand if they really want to do research on which they could make more precise conclusions. Over are the analyses where all you could say is "X is on average more likely to do Y than Z is". With this you can say "The co-occurrence of A, B and C under X is way above expected and thus if A, B and C occur together it predicts X..." and "None of those who were X and Y could do A, but more than predicted of those who were X and Z could do A and hence there's K happening." and something like that. Read it. Use it.