Midterm 1 - Correlation & Regression (Ch. 3)

Question 1

Correlations serve to determine the extent to which ______

Answer 1

changes in scores for 2 variables are related

Question 2

Pearson’s correlation coefficient provides a quantitative description of the ___ and ___ of a straight-line relationship between two variables

Answer 2

direction and strength

Question 3

Pearson’s product-moment correlation coefficient describes relationships between scores from ____

Answer 3

interval and ratio scales

Question 4

what is the range for correlation coefficients?

Answer 4

-1 to 1

Question 5

Pearson correlation (r) can be calculated by dividing _________ by _______

Answer 5

degree to which X and Y vary together (covariability)
degree to which X and Y vary separately

Question 6

(T/F) Pearson correlations can sometimes be used to describe non-linear relationships

Answer 6

FALSE

Question 7

Pearson correlations are (slightly/strongly) affected by outliers

Answer 7

strongly!

Question 8

A regression line is a straight line that describes how _____ changes as ____ changes. Essentially, it provides an average statement about the change in __ that is ass w change in __

Answer 8

a dependent (outcome) variable changes as an independent (predictor) variable changes

change in Y ass w change in X (often used to predict value of y for a given value of x)

Question 9

What is the criterion to satisfy when placing a regression line?

Answer 9

sum of all vertical distances from each point to line should be as small as possible

Question 10

What is the least-squares regression line?

Answer 10

line of Y on X that makes the sum of the squares of the vertical distances of points from line as small as possible

Question 11

in a regression line, vertical distance to the line is calculated by subtracting ___ from ___

Answer 11

subtract predicted y(w hat; on line) from observed y (point)

Question 12

In a regression line, Vertical distance to the line is also called the ____ or ____

Answer 12

error of prediction or residual

Question 13

What is the regression equation?

Answer 13

ŷ = bx + a

a = intercept (y when x is 0)
b = slope (amount y changes when x increases by 1)

Question 14

What does the regression line look like when the correlation is 0?

Answer 14

horizontal to x axis

Question 15

to determine how useful/accurate a regression line is for prediction, we compute the ___ which measures the ______ in one variable that can be determined from the relationship with the other variable

Answer 15

r(squared) - aka coefficient of determination

measures the proportion of variability….. (basically shared variability)

Question 16

The prediction of ŷ based on x is stronger when the correlation is (strong/weak)

Answer 16

strong!