1 Introduction and General Aims

This book has its origin in a week-long intensive course on methods of twin data analysis taught between 1987 and 1997 at the Katholieke Universiteit of Leuven in Belgium, the University of Helsinki, Finland, and the Institute for Behavioral Genetics, Boulder, in Colorado. Our principal aim here is to help those interested in the genetic analysis of individual differences to realize that there are more challenging questions than simply ``Is trait X genetic?" or ``What is the heritability of X?" and that there are more flexible and informative methods than those that have been popular for more than half a century. We shall achieve this goal primarily by considering those analyses of data on twins that can be conducted with the Mx program. There are two main reasons for this restriction: 1) the basic structure and logic of the twin design is simple and yet can illustrate many of the conceptual and practical issues that need to be addressed in any genetic study of individual differences; 2) the Mx program is well-documented, freely available for personal computers and Unix workstations, and can be used to apply all of the basic ideas we shall discuss. We believe that the material to be presented will open many new horizons to investigators in a wide range of disciplines and provide them with the tools to begin to explore their own data more fruitfully. The four main aims of this introductory chapter are:

1. to identify some of the scientific questions which have aroused the curiosity of investigators and led them to develop the approaches we describe
2. to trace part of the intellectual tradition that led us to the approach we are to present in this text
3. to outline the overall logical structure of the approach
4. to accomplish all of these with the minimum of statistics and mathematics.

Before starting on what we are going to do, however, it is important to point out what we are not going to cover. There will be almost nothing in this book about detecting the contribution of individual loci of large effect against the background of other genetic and environmental effects (``segregation analysis"). In contrast to the first edition, there will be a chapter on linkage analysis concerning the location on the genome of individual genes of major effect, if they exist. These issues have been treated extensively elsewhere (see e.g., Ott, 1985, Sham, 1998, Lange, 1997, Lynch & Walsh, 1998) -- often to the exclusion of issues that may still turn out to be equally important, such as those outlined in this chapter. When the history of genetic epidemiology is written, we believe that the approaches described here will be credited with revealing the naivete of many of the simple assumptions about the action of genes and environment that are usually made in the search for single loci of large effect. Our work may thus be seen in the context of exploring those parameters of the coaction of genes and environment which are frequently not considered in conventional segregation and linkage analysis.