Measurement Test 2 Flashcards
(36 cards)
squared correlation (r squared)
percent of variance explained
Distance between X and the line is the . . .
error
error =
observed - predicted
b0
The expected value of y when x = 0
b1
the expected value of y given a 1 unit increase in x
In a regression analysis, we aim to minimize . . .
the sum of square errors
Error variance = 0 when
- There was no relationship 2. There was a perfect relationship
If x is centered, b0 will equal . . .
the mean of y
When there is no relationship between X and Y
- The slope is flat 2. The intercept = the mean of Y 3. Sum of Squares error = sum of squares 4. Error variance = total variance
Tolerance
Used to assess the extent of collinearity
Beta is calculated using the . . .
matrix algebra formula
matrix of betas =
(matrix of co variances between x and y) / (variance - covariance matrix for all x variables)
What are the three questions of multiple regression?
- Effect of each x ignoring the others 2. Effect of each x controlling for the others 3. Total effect of all xs taken together
What descriptive statistic is used when measuring the effect of each x while ignoring the others?
r (correlation)
What inferential statistic is used when measuring the effect of each x while ignoring the others?
t (t-test)
What descriptive statistic is used when measuring the effect of each x while controlling for the others?
standardized beta
What inferential statistic is used when measuring the effect of each x while controlling for the others?
t (t-test)
What descriptive statistic is used when measuring the total effect of all xs taken together?
R squared
What inferential statistic is used when measuring the total effect of all xs taken together?
F
What descriptive statistics are biased?
R and R squared
When measuring the effect of each x controlling for the others, beta =
the predicted increase in Y for a one unit increase in X when all other variables are held constant
When calculating partial correlation, if beta is negative, then . . .
make the correlation negative
Why is delta R squared helpful?
It can be used with more than 2 predictors
Suppressor effect
The partial correlation is larger than the correlation