• Preliminary remarks
• 1. Formal description of mating systems
• 1.1 Mating references (potential mating partners)
• 1.2 Mating preferences
• 1.3 Mating success
• 1.4 Conditional random mating
• 1.5 Mating equivalence
• 1.6 Application to an experimental data set
• 2. Classification of mating systems
• 2.1 Random mating in cosexual populations
• 2.2 Random mating in dioecious populations
• 2.3 Sexual selection
• 2.4 Classification of mating systems by product structures
• 2.5 Assortative mating
• 2.6 Random preferential mating
• 3. Models of random mating
• 3.1 Cosexuality: Effects of random mating in the haplophase on the mating frequencies in the diplophase
• 3.2 Dioecy: The effect of random mating in the diplophase on the mating frequencies in the haplophase
• 3.3 Gynodioecy and androdioecy
• 3.4 Trioecy
• 4. Evolution of systems of sexuality
• 4.1 Gynodioecy
• 4.2 Trioecy
• 5. Self-mating or self-fertilization
• 5.1 Functional sex and self-mating proportions
• 5.2 Self-mating and random cross-mating
• 5.3 A model based on female self-mating proportions
• 6. Genetic aspects of self-mating
• 6.1 Dynamics of genotypic frequencies
• 6.2 Complete self-mating
• 6.3 Absence of differential proportions of self-mating
• 7. Models of assortative mating
• 7.1 Asynchrony of sexual maturity; Positive and negative assortative mating
• 7.2 Synchrony of sexual maturity; Positive assortative mating (homotypic predisposition); Negative assortative mating (heterotypic predisposition)
• 8. Selection with two alleles and two types
• 8.1 Complete dominance of an allele
• 8.2 Mating equivalence of the homozygotes
• 9. Heterophasic mating (gametophytic incompatibility); General representation
• 9.1 Differential pollen tube growth; A model
• 9.2 Evolution of differential pollen tube growth;  Negative assortative mating; Two alleles
• References
• Subject index

1. Introduction
2. General procedure for an analysis of causal mechanisms: Testing models; Critical regions of observations
3. Estimation of target characteristics and determination of confidence regions
4. Statistical aspects of system analysis - incomplete observations: Testing models without free parameter values; Testing models with free parameter values; Confidence regions and estimation; Measures of discrepancy
5. Model dependence of analysis: Design of model

The principle of hypothesis falsification forms the epistemological basis of systems analysis. It is realized by specification of a method of investigation suitable for testing a set of hypotheses on system characteristics. From this perspective, the principle has a special position in statistics, in that its observance calls for controlling the chances of erroneous falsification for each of the considered hypotheses. Falsification by improbability thus replaces strict falsification. Regions of rejection of hypotheses are then solely determined by the levels of significance set to the probability of not obtaining a result more probable than the observed under the respective hypothesis. Being a function of hypothesis and observation, this probability is termed the confidence zeta in a hypothesis supported by an observation, and it establishes the relation to the theory of exact testing. Taking advantage of the analogy to hypothesis testing, "exact" confidence regions can then be determined as the set of hypotheses that is supported by the observation with confidence zeta above a given level alpha of significance. The problem of hypothesis (including parameter) estimation is introduced as a problem of selecting hypotheses from confidence regions, and the roles of maximizing likelihood and confidence for obtaining and testing estimates is discussed. Maximization of confidence over subsets of hypotheses is also shown to yield unconditional tests of composite hypotheses. Application to the determination of "exact" confidence regions for the frequency parameter of the binomial distribution yielded results that are in many cases closely approximated by the direct method of Clopper and Pearson but also show distinct and unfamiliar features.