« Back to student resources
Glossary
A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z
C
correlation coefficient - Measures the linear relationship between two interval-ratio variables; tells us the degree to which the values of one variable change as the values of the other variable change.
curvilinear relationship - A relationship between two variables follows a curved line rather than a straight line.
D
dummy variable - A variable with two categories, coded as 0 and 1. The category coded as 0 is called the “reference category.”
F
F-statistic - In regression, used to test the statistical significance of a regression model overall by assessing whether the regression model does a better job of generating predicted values of the dependent variable than an “intercept-only model.” In analysis of variance, the ratio of between groups to within groups variation; used to test whether there are statistically significant differences among groups.
I
intercept - In regression, the expected value of the dependent variable when the independent variable(s) are equal to zero. The value of the dependent variable where the regression line crosses the y axis. Represented by “a” in the equation for the regression line.
L
linear regression - A linear model that measures the effect of an independent variable, x, on a dependent variable, y.
linear relationship - A relationship between two variables that can be represented by a straight line. In a linear relationship, every unit change in the independent variable is associated with a constant amount of change in the dependent variable.
logistic regression - A regression model that can be used with binary (dummy) dependent variables. Logistic regression predicts the probability that a dummy dependent variable (coded as 0 or 1) will be equal to 1 for given values of the independent variables.
R
r-squared (r2) - Goodness of fit measure for regression that indicates the proportion of variation in the dependent variable that is accounted for by the independence variable(s).
regression coefficient - The expected change in the dependent variable for a one-unit increase in the independent variable; represented by “b” in the regression equation. A positive coefficient predicts an increase in the dependent variable for a one-unit increase in the independent variable, while a negative coefficient predicts a decrease in the dependent variable for a one-unit increase in the independent variable. Also called slope.
regression coefficient - The expected change in the dependent variable for a one-unit increase in the independent variable; represented by “b” in the regression equation. A positive coefficient predicts an increase in the dependent variable for a one-unit increase in the independent variable, while a negative coefficient predicts a decrease in the dependent variable for a one-unit increase in the independent variable. Also called slope.
regression line - The line that best fits a scatterplot; represented by the equation yˆ = a + bx, where a is the intercept, b is the slope (coefficient), and yˆ is the predicted value of the dependent variable (y) for a given value of the independent variable (x).
S
scatterplot - A graph showing the paired values of two interval-ratio variables for each case in a data set.
slope - The expected change in the dependent variable for a one-unit increase in the independent variable; represented by “b” in the regression equation. A positive slope predicts an increase in the dependent variable for a one-unit increase in the independent variable, while a negative slope predicts a decrease in the dependent variable for a one-unit increase in the independent variable. Also called regression coefficient.
Standard error of the estimate - Goodness-of-fit measure for regression that indicates how far, on average, the actual values of the dependent variable fall from the values predicted by the regression line.
standard error of the slope - Used to test the statistical significance of a single slope in regression; indicates how much, on average, a sample slope varies from the population slope due to sampling error.
T
t-value - Refers to values of the t-distribution; the distance between any value of a sample statistic that follows a t-distribution and the population parameter, expressed in standard errors.
