### assessment research

Research carried out to determine students' levels of skill or other capabilities or characteristics.

### background factors

Variables that may cause other variables but are not caused by them and that are typically unchangeable attributes of individuals.

### categorial or discrete variable

A variable (such as gender, ethnicity, or treatment) that can take on a limited number of values.

### causal-comparative designs

Correlational designs that use categorical variables as independent variables.

### causal model

A theoretical ordering of variables in terms of their effects on other variables. Correlations between several pairs of variables may be used to evaluate a particular causal model, as in the technique of path analysis.

### continuous variable

A variable (such as age, test score, or height) that can take on a wide or infinite number of values.

### control variable

A variable used to remove the effect of some factor on the relationship between two or more other variables; also called a covariate.

### correlation

The degree to which two variables tend to vary in the same direction or in opposite directions.

### correlation coefficient

A statistic indicating the degree to which two variables are correlated. It may take on values from -1.0 (perfect negative correlation) to +1.0 (perfect positive correlation). A correlation coefficient of 0 indicates that the variables are unrelated.

### correlation matrix

A table of correlation coefficients showing all possible correlation pairs between a set of variables.

### correlational study

A nonexperimental research design in which the researcher collects data on two or more variables to determine if they are related (that is, if they consistently vary in the same or opposite directions).

### criterion-referenced test

A test designed to indicate how an individual performs in comparison to a preestablished criterion.

### dichotomous variable

A categorical variable (such as gender, on-off task, experimental-control) that can take on only two values.

### direction of causality

A determination of which variable causes the other in correlational research.

### multiple regression

A statistical method that evaluates the effects of one or more independent variables on a dependent (outcome) variable, controlling for one or more covariates or control variables.

### nonexperimental quantitative design

A research design (such as a correlational or descriptive design) in which the researcher measures or observes subjects without attempting to introduce a treatment.

### norm-referenced test

A test designed to indicate how an individual performs in comparison to others (such as others of the same grade level or age).

### partial correlation

A correlation in which the relationship between two variables is calculated with the effect of a third variable removed.

### self-selection bias

Bias introduced in a study by the fact that the subjects chose to participate or not participate in a given program.

### spurious correlation

An apparent correlation between two variables that is actually caused by other variables.