chapter 9

Question 1

What can researchers do to get closer to a causal claim when they originally only have a bivariate correlation?

Answer 1

Use multivariate techniques.

Question 2

what causal criteria are longitudinal designs good at showing

Answer 2

temporal precedence

Question 3

what type of correlations can be used from longitudinal designs that help us determine temporal precedence? how?

Answer 3

compare the relative strength of 2 cross lag correlations

Question 4

longitudinal designs make what 3 types of correlations

Answer 4

cross sectional
autocorrelations
cross lag correlation

Question 5

cross sectional correlation

Answer 5

corr between the 2 key variables at any one time period

Question 6

autocorrelations

Answer 6

corr between one variable and itself over time

Question 7

regression design

Answer 7

starts with a bivariate correlation , then measures other possible third variables

Question 8

why do we use multi regression analysis?

Answer 8

we can see if the basic relationship is still present even when controlling for one or more third variables

if beta is still significant for key variables after controlling, relationship isnt due to third variables

Question 9

if beta approaches 0 when controlling for 3rd variable, what do we assume about the relationship

Answer 9

its because of the third variable

Question 10

what can regression analysis NOT do?

Answer 10

establish causation

Question 11

mediation hypothesis

Answer 11

specify a variable that comes between the 2 variables of interest as a potential reason for why these variables are associated

Question 12

To increase causal certainty we can …

Answer 12

use pattern and parsimony to specify a mechanism for causal relationship and combine results from variety of research questions

Question 13

Multivariate design

Answer 13

test an association w more than 2 measured variables

Question 14

Cross lag correlation

Answer 14

corr between 2 variables that are measured at same time

Question 15

Multiple regression or multivariate regression

Answer 15

statistical technique to compute relationship between predictor and criterion variable while controlling for other predictor variables

Question 16

Criterion variable

Answer 16

variable that researchers are most interested in understanding (also called dependent variable).

Question 17

Predictor variable

Answer 17

variable used to explain variance in criterion variable (also called independent variable).

Question 18

Parsimony

Answer 18

the simplest way to explain a pattern of data, requiring the fewest exceptions or qualifications

Question 19

Mediator

Answer 19

helps explain relation between 2 other variables