Just as it is said that the index of a book is the last chance for the author to shape how the book is read, a glossary can convey the sensibility of a book. Below is the glossary for Taylor, Peter J. (2014) Nature-Nurture? No: Moving the Sciences of Variation and Heredity Beyond the Gaps. *Items in italics* are described elsewhere in the glossary.

**Additive model**: a) Equation that connects the values of the trait for an individual to the summation of several *contributions* (sense b); b) A parallel equation that adds up *variances* related to these contributions.

**ANOVA or Analysis of Variance**: The *variance* of a trait is divided into parts or *components* corresponding to each term (*contribution* [sense b]) in an *additive model*. For example, for observations of a set of *varieties* raised in a set of *locations*, one part would be the variance of the *variety means*. The other parts would be the location means, the variety-location-interaction means (after subtracting the first two means), and noise, residual or “error” contributions (see equation 1 in Item D).

**Behavioral genetics:** *Quantitative genetics* applied to behavioral traits, most commonly human behavioral traits.

**Cause**: A difference that makes a difference. The former difference refers to either a) an intentional modification (e.g., adding fertilizer to an agricultural plot); or b) a distinction between points of data with regard to a measured *factor* that is not, by and large, modifiable (e.g., male versus female sex).

**Component:** (as used in this book) The parts into which the *ANOVA* divides the *variance* of a trait.

**Contribution** (as used in this book): a) as colloquially understood, i.e., something given or supplied in common with other contributors; b) a term in an equation that connects the values of the trait for an individual to the summation of several such terms (see *additive model*). Contribution sense b. substitutes for the technical term *effect* because that term is causally ambiguous.

**Dizygotic (DZ) twins**: Two offspring gestated together from two separately fertilized eggs. Contrast *Monozygotic twins*.

**Effect**: See *contribution*, sense b.

**Environment**: a) Set of specific environmental *factors*; or b) Synonym for *location*, a term that can be used without reference to or knowledge of the environmental factors present.

**Epigenetics** (as the term is used in recent molecular biology): Chemicals from outside the cell can modify the activity of genes for the rest of the organism’s life and sometimes even into subsequent generations.

**Experimental data: **Data derived from explicit manipulation of measured *factors* or conditions.

**Factor** (as used in this book): Something for which its presence or absence can be observed or its level can be measured—a quality that is emphasized in some places in the text by adding the adjective “measurable.” For any given trait, the factors of interest are those associated with changes or variation in the trait’s development, but the *causal* quality of a factor is a secondary matter. Measurable genetic factors include the presence or absence of variants (alleles) at a specific place (locus) on a chromosome, repeated DNA sequences, reversed sections of chromosomes, and so on. Measurable environmental factors can range widely, for example, from fertilizer application per hectare of crop to average daily intake of calories to degree of maltreatment that a person experienced as a child.

**Genetic** (as used in this book): An adjective used in reference to *factors* that are transmitted through the germline from parents to offspring and whose presence or absence can, in principle, be observed.

**Genotype**: a) The pair of variants (alleles) for a given place or locus on two paired chromosomes; b) A synonym for *variety*, a term that can be used without reference to or knowledge of the genetic factors present.

**Genotypic values**: See *variety means*.

**Heritability**: a) The ratio of the *variance* of the *variety means* for a given trait to the overall *variance* of that trait in a population consisting of a specific set of *varieties* raised in some specific *location*(s); b) (recent definition, referred to in this book as new heritability): The fraction of *variance* in a trait associated with variation in Single-Nucleotide Polymorphisms (SNPs) as examined by Genome-Wide Association studies.

**Heritability studies**: (as used in this book) Research that uses methods of *quantitative genetics* for partitioning variation in a trait into *heritability* and other *components*, such as, the *variance* of the location means.

**Heterogeneity**: A state in which one kind of thing can be separated into a spectrum, range or mixture of many different kinds. See also *underlying heterogeneity*.

**Heterozygous**: The two variants that make up a *genotype* (sense a) are different.

**Homozygous**: The two variants that make up a *genotype* (sense a) are the same.

**Interaction**: a) A variety-by-location interaction is a *contribution* (sense b) and variety-by-location interaction *variance* is one part in an *Analysis of Variance*. When this variance is high for the trait in question, the ranking of varieties varies across locations and the best variety in one location is not the best in other locations; b) A gene-environment interaction derives from a *regression analysis* involving measured genetic *factors* and measured environmental factors in which included in the *additive model* are additional terms that are the product of a genetic factor and an environmental factor. More generally, interaction means that the quantitative relation between the trait and one of the factors varies according to the measured value of the other factor.

**Intraclass Correlation**: Ratio of *variances* of the contributions that do not vary within the class (and are thus included in the class averages) to the overall variance. For classes of size two this quantity is mathematically equivalent to the usual linear correlation of the two values in each class or pair, when the order in each pair is arbitrary. Arbitrary ordering would apply if one wanted to know the correlation, for example, of heights in same-sex couples.

**Location**: The situation or place in which a *variety* is raised, such as a family (for humans) or a specific experimental research station (for agricultural varieties)

**Measurable factor**: See *factor*.

**Monozygotic (MZ) twins**: Two offspring gestated together after a single fertilized egg splits and forms two embryos.

**Observational data: **Data derived on individuals that can be subdivided into relevant categories (e.g., raised in low socioeconomic status) but have not been assigned randomly to be subject to specific conditions (e.g., 50 kg/ha of nitrogen fertilizer) (see *experimental data*).

**Path Analysis**: Data analysis technique that defines a network of interrelated variables, which may be measurable *factors* or *contributions* estimated in an *Analysis of Variance*, and estimates the relative contribution of each variable to the variation in a focal variable after allowing for the intervening variables. The estimates of relative contributions are called path coefficients. Their reliability depends on the assumptions built into the networks, such as, in *heritability studies*, the similarity of relatives of different degrees and the inclusion in the *additive model* (or exclusion) of coefficients for variety-location *interaction*.

**Quantitative genetics**: A field in which variation in traits of humans, other animals, or plants is analyzed in ways that take account of the genealogical relatedness of the individuals whose traits are observed. (See also *behavioral genetics* and *heritability studies*.)

**Regression analysis**: For an equation (often in the form of an *additive model*) that combines a set of measured *factors* (the “independent” variables), the coefficients to each variable are estimated that make the equation predict the trait (the “dependent” variable) better than other values of the coefficients. “Predict best” can be assessed by the lowest residual *variance* (the “least squares”) or other criteria.

**Replicates**: Two or more individuals raised in the same variety-location combination (for heritability studies) or with the same values of the measured factors for *regression analysis*.

**Rerun Predictability**: The closeness of a match in the following situation: For a set of variety-location combinations trait values are a result of unknown dynamics that include some unsystematic noise. The same variety-location combinations are observed again where the only change in this “rerun” is noise at the same level as the original, but uncorrelated with it. All possible pairs of values are considered in which the first value is from the original situation and the second is from the rerun, where for each pair the original and rerun *variety* is the same (alternatively, the *location* is the same). The closeness of match is assessed by the correlation of the original and rerun values, making use of an *additive model* of *contributions* and assuming that the actual noise is given by the residual or noise *contribution*.

**Structural Equation Modeling**: A generalization of *path analysis* to include unmeasured “latent” variables that add together several measured *factors* (see *additive model*) in a way that fits the actual data well, not in a predetermined equation.

**Trait: **Observed characteristic of an individual. (See Item B.1 for explanation of why the term commonly used in *heritability studies*, phenotype, is not used in this book.)

**Underlying Heterogeneity**: When similar responses of different *varieties* or individual types are observed, but it is not the case that similar conjunctions of genetic and environmental *factors* (or, in epidemiology, risk and protective factors) have been involved in producing those responses (i.e., values for the trait in question).

**Variance**: The common measure of variation. The variance for the population considers the size of the deviation of an individual’s trait from the mean value for the population, squares that, then finds the average over all the individuals. The variety variance considers the size of the deviations of the variety means from the overall mean, and so on.

**Variety**: A group of individuals whose relatedness by genealogy can be characterized, such as offspring of a given pair of parents, or a group of individuals whose mix of genetic *factors* can be replicated, as in an open pollinated plant variety or pure (genetically identical) lines. The term can be used without knowledge of the underlying genetic factors.

**Variety mean**: The mean or average of the values observed for the trait across all locations in which the *genotype* (sense b) or *variety* is raised, minus the overall mean for the population.