Exam One - correlation and regression

Question 1

What is the goal of correlation?

Answer 1

evaluate the extent of a relationship between two continuous variables

Question 2

What two things does the correlation coefficient tell us?

Answer 2

the strength and direction of the relationship

Question 3

What letter is the correlation coefficient represented by?

Answer 3

lowercase r

Question 4

______ = perfect inverse relation

Answer 4

-1

Question 5

_______ = no relationship

Answer 5

zero

Question 6

________ = perfect direct relationship

Answer 6

+1

Question 7

coefficient of determination is represented by…

Answer 7

r^2

Question 8

definition of coefficient of determination (r^2)

Answer 8

percentage of the total variance in Y scores that can be explained by X scores

Question 9

______ r^2 values mean you can make more accurate estimates based on knowledge of one variable

Answer 9

higher

Question 10

P value is inversely related to…

Answer 10

sample size and r value

Question 11

magnitude of r relevance scale

Answer 11

0-0.25 = little to no relation
.25-.5 = fair relation
.5-.75 moderate/good
>.75 good/excellent

Question 12

correlations (should/should not) be interpreted as being important if p-value is <0.05

Answer 12

should not

Question 13

t or f? correlation determines causation

Answer 13

FALSE

Question 14

coefficient r measures a linear or curvilinear relationship?

Answer 14

linear

Question 15

values that are correlated are not…

Answer 15

necessarily similar

Question 16

what is the goal of regression?

Answer 16

prediction

Question 17

definition of regression

Answer 17

measurement variables used to develop an equation to predict one based on the other

Question 18

in linear regression equation X is the ….

Answer 18

independent or predictor variable

Question 19

in linear regression equation Y is the…

Answer 19

dependent variable or criterion variable

Question 20

distance between the regression line and data point are called…

Answer 20

residuals

Question 21

residuals are the…

Answer 21

extent of error for each observation relative to the repression line (prediction)

Question 22

in equation Y = 64.72 + 1.39 (x) what is the dependent variable?

Answer 22

y

Question 23

in equation Y = 64.72 + 1.39 (x) what is the independent variable?

Answer 23

x

Question 24

in equation Y = 64.72 + 1.39 (x) what is the y-intercept?

Answer 24

64.72

Question 25

in equation Y = 64.72 + 1.39 (x) what is the slope?

Answer 25

1.39

Question 26

What units does r^2 use?

Answer 26

no units, just % of variability

Question 27

_____ r^2 implies a more accurate prediction equation

Answer 27

higher

Question 28

SEE

Answer 28

standard error of the estimate - magnitude of the expected error in Y based on the predictors

Question 29

what units does SEE use?

Answer 29

the units of Y

Question 30

_______ SEE implies a more accurate prediction equation

Answer 30

lower

Question 31

95% CI of the prediction =

Answer 31

predicted value +/- (1.96*SEE)

Question 32

multiple linear regression

Answer 32

method to predict the criterion variable based on more than one predictor variable

Question 33

What two methods are possible for multiple linear regression?

Answer 33

- "enter" method - forward, backward, or stepwise methods

Question 34

enter method of multiple linear regression

Answer 34

this will use all variables in the prediction equation whether they make a meaningful contribution or not

Question 35

forward, backward, or stepwise methods of linear regression

Answer 35

uses statistical criteria to max prediction accuracy using the fewest possible variables

Question 36

_______- people per predictor variable are necessary when using multiple linear regression

Answer 36

7-15 (liberal)

Question 37

The accuracy of multiple linear regression is evaluated by...

Answer 37

the coefficient of determination r^2 and the SEE

Question 38

r^2

Answer 38

proportion of the variability in Y accounted for by all the predictors

Question 39

SEE

Answer 39

expected error in our prediction of Y, expressed in units of Y

Question 40

logistic regression

Answer 40

similar to linear regression, predicted variables are evaluated for the extent they accurately predict a CATEGORICAL outcomee

Question 41

1 predictor: ____ logistic regresion

Answer 41

simple

Question 42

>1 predictor: _______ logistic regression

Answer 42

multiple

Question 43

use of logistic regresion?

Answer 43

predicting categorical clinical outcomes