Lecture 11: Non-binary linear regression

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

1/30

encourage image

There's no tags or description

Looks like no tags are added yet.

31 Terms

1
New cards

What is the purpose of linear regression for prediction?

To model the linear relationship between predictors and an outcome so we can predict outcomes for new observations.

2
New cards

What is a predictor variable?

A variable used to predict an outcome, denoted X.

3
New cards

What is an outcome variable in regression?

The variable we want to predict, denoted Y.

4
New cards

What does Y-hat represent?

The predicted value of the outcome variable.

5
New cards

What is simple linear regression?

A linear regression model with one predictor variable.

6
New cards

What does the intercept represent?

The predicted value of Y when X equals zero.

7
New cards

What does the slope represent?

The predicted change in Y associated with a one-unit increase in X, on average.

8
New cards

What does a positive slope indicate?

A predicted increase in Y as X increases.

9
New cards

What does a negative slope indicate?

A predicted decrease in Y as X increases.

10
New cards

What does a slope of zero indicate?

No predicted change in Y as X changes.

11
New cards

What is a residual?

The difference between an observed outcome and its predicted outcome.

12
New cards

What is the residual formula?

Observed Y minus predicted Y.

13
New cards

What does a positive residual indicate?

The model underpredicted the observed outcome.

14
New cards

What does a negative residual indicate?

The model overpredicted the observed outcome.

15
New cards

What does a residual of zero indicate?

The predicted outcome equals the observed outcome.

16
New cards

What does a residual represent visually?

The vertical distance between an observed data point and the fitted regression line, with its sign.

17
New cards

What is the least squares method?

A method that selects the line minimizing the sum of squared residuals across all observations.

18
New cards

Why are residuals squared in least squares?

To prevent positive and negative errors from cancelling and to penalize larger errors.

19
New cards

What are the two steps involved in prediction?

Fit a model using observed X and Y, then predict Y using new observed values of X.

20
New cards

How do we predict an outcome using a fitted regression line?

Multiply the predictor value by the slope and add the intercept.

21
New cards

How do we predict the change in Y associated with a change in X?

Multiply the slope by the change in X.

22
New cards

Do we include the intercept when predicting a change in Y?

No, the intercept cancels when comparing predictions.

23
New cards

What does lm() do?

Fits a linear regression model.

24
New cards

In lm(Y ~ X, data = dataset), which variable is the outcome?

Y, the variable before the tilde.

25
New cards

In lm(Y ~ X, data = dataset), which variable is the predictor?

X, the variable after the tilde.

26
New cards

Does a regression slope automatically represent a causal effect?

No, it generally represents an association unless causal assumptions are satisfied.

27
New cards

What language should be used when interpreting a noncausal slope?

A one-unit increase in X is associated with a predicted change in Y, on average.

28
New cards

What determines the unit used to interpret a non-binary outcome intercept?

The original unit of measurement of Y.

29
New cards

What determines the units used when interpreting a non-binary outcome slope?

The change in Y's units associated with a one-unit increase in X.

30
New cards

What is extrapolation?

Making predictions for predictor values outside the observed range of data.

31
New cards

Why might an intercept have limited substantive meaning?

X = 0 may be unrealistic or outside the observed range.