Multiple Regression

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/17

encourage image

There's no tags or description

Looks like no tags are added yet.

18 Terms

1
New cards

Multiple Regression

A statistical technique used to explain the variance in a continuous outcome measure using a linear equation with more than one predictor variable.

2
New cards

R2R^2 (R-Squared)

A statistic indicating the proportion of variance in the outcome variable explained by the regression model for the sample, which tends to be an overestimate of the population value.

3
New cards

Adjusted R2R^2

A downward-adjusted estimate of R2R^2 that accounts for sample size to provide a more accurate reflection of the true variance explained in the population.

4
New cards

Regression Degrees of Freedom (dfregdf_{reg})

The degrees of freedom associated with the regression model in the ANOVA table, equal to the number of predictor variables (kk).

5
New cards

Residual Degrees of Freedom (dfresdf_{res})

The error degrees of freedom in multiple regression, calculated using the formula N−k−1N - k - 1, where NN is the total sample size and kk is the number of predictors.

6
New cards

Unstandardized Regression Coefficient (bb)

The estimated contribution of a predictor variable to the outcome while holding constant or controlling for all other predictors in the linear model.

7
New cards

Standardized Coefficient (β\beta / Beta)

A regression coefficient expressed in standard deviation units that enables direct comparison of the relative importance and effect sizes among different predictors.

8
New cards

Dummy Variables

Categorical variables converted into numeric indicators coded with 00 and 11 to allow their inclusion in a linear regression model.

9
New cards

Multicollinearity, new assumption but others remain the same

An assumption violation that occurs when predictor variables are moderately to strongly correlated with each other, leading to unstable bb estimates and artificially deflated R2R^2.

10
New cards

Variance Inflation Factor (VIF)

A collinearity diagnostic indicating the degree to which a predictor is correlated with other predictors, where values approaching 1010 indicate serious concern, used to check for multicollinearity

11
New cards

Tolerance

A collinearity statistic calculated as 1/VIF1 / \text{VIF}, which should remain above 0.20.2 to confirm the absence of severe multicollinearity.

12
New cards

Cook's Distance

An influence statistic assessing the impact of individual cases on the overall regression model, where values should generally remain below 11.

13
New cards
<p>Bootstrapping</p>

Bootstrapping

A robust non-parametric resampling technique that derives parameter estimates and confidence intervals by repeatedly drawing samples from the observed data without assuming normally distributed residuals.

14
New cards

Forced Entry Multiple Regression

A regression approach where all chosen predictor variables are entered into the linear model simultaneously at once.

15
New cards

Hierarchical Multiple Regression

A multiple regression method where predictor variables are entered in planned, sequential blocks or steps to determine whether adding new predictors increases explained variance after controlling for previous variables.

16
New cards

FF-Change (ΔF\Delta F)

A statistical test evaluating whether the additional variance explained (R2 ChangeR^2\text{ Change}) by entering a subsequent block of predictors in hierarchical regression is significantly different from zero.

17
New cards

Deep Learning Approach

A study approach focused on seeking meaning and conceptual understanding when studying academic material.

18
New cards

Surface Learning Approach

A study approach focused on rote memorization and factual recall rather than understanding.