Scribd
Upload
9 views

PSYC344 - Ch. 9 Multivariate Correlational Research Part 1

Uploaded by

emilynunezz0703
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
9 views

PSYC344 - Ch. 9 Multivariate Correlational Research Part 1

Uploaded by

emilynunezz0703
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 38

Multivariate Correlational Research Dr.

Liz Shin
(Part 1)
Learning Objectives

1. Explain how longitudinal designs are conducted.


2. Identify three types of correlations in a longitudinal
correlational design.
3. Explain how multiple-regression designs are conducted.
4. Identify and interpret data from a multiple-regression table and
explain what each coefficient means.
Reviewing the Three Causal Criteria: Part 1

• Multivariate designs (such as longitudinal and multiple


regression designs) involve more than two measured variables.
Reviewing the Three Causal Criteria: Part 2

• Three criteria for establishing causation


• Covariance
• Temporal precedence
• Internal validity
Interpreting Results from Longitudinal Designs

• Longitudinal Designs

Time 1 Time 2 Time 3

Variable 1: Sense of Belonging Variable 1: Sense of Belonging Variable 1: Sense of Belonging


Variable 2: Self-Efficacy Variable 2: Self-Efficacy Variable 2: Self-Efficacy
Cross-Sectional Correlations
Cross-Sectional Correlations
Cross-Sectional Correlations
C E
E DEN
P REC
O RAL
E MP
SH T
B L I
ESTA
NOT
CA N
Autocorrelations
Autocorrelations
C E
E DEN
P REC
O RAL
EMP
SH T
B L I
ESTA
NOT
CA N
Cross-Lag Correlations
Cross-Lag Correlations
Cross-Lag Correlations

NC E
ED E
R EC
A L P
M POR
G TE
SHI N
B L I
ESTA
ER TO
LO S
C
Cross-Lag Correlations

What if this was significant??


Longitudinal Studies and the Three Criteria for
Causation
• Longitudinal designs can provide SOME evidence for causation by
fulfilling three criteria
• Covariance
• Temporal precedence
• Internal validity
Why Not Just Do an Experiment?
• In many cases participants cannot be randomly assigned to a
variable.
Ruling Out Third Variables with
Multiple-Regression Analyses
• Multiple regression: a statistical method that examines the
relationship between one dependent variable and multiple
independent variables, allowing researchers to assess the unique
contribution of each predictor while controlling for others.
Measuring More Than Two Variables

Potential third variables?


Using Statistics to Control for Third Variables

“Control for” (to account for other variables' effects to isolate the
unique impact of the main variable of interest on the outcome)
Using Statistics to Control for Third Variables
Using Statistics to Control for Third Variables
Using Statistics to Control for Third Variables
Using Statistics to Control for Third Variables
Regression Results Indicate If a Third Variable
Affects the Relationship
• Criterion variables (e.g., pregnancy risk)
• Predictor variables (e.g., sexual TV content)
• Use beta to test for third variables
Using Beta to Test for Third Variables: Part 1
β (beta) represents the standardized coefficient in a regression
analysis; it indicates the direction and strength of the relationship.
Using Beta to Test for Third Variables: Part 2
• Interpreting beta (both p-value and confidence interval are used
to determine statistical significance)
Using Beta to Test for Third Variables: Part 3
• Using p-value
Using Beta to Test for Third Variables: Part 4
• Using confidence interval
Using Beta to Test for Third Variables: Part 5

• Can only compare beta strengths


within a single regression table
Using Beta to Test for Third Variables: Part 6
• What if beta is not significant?

• Example:
• Predictor variable: family meal frequency
• Criterion variable: academic success
Using Beta to Test for Third Variables: Part 7
Adding More Predictors to a Regression
Adding More Predictors to a Regression
Interpreting a Regression Table
• Predicting political conservatism at age 18

Predictor Beta p
Child’s gender (0 = male, 1 = female) -0.11 .01
Child’s ethnicity 0.17 <.01
Cognitive skills at 54 months -0.05 .30
Socioeconomic status at 54 months -0.09 .07
Authoritarian parenting attitudes at 1 0.16 <.01
month
Egalitarian parenting attitudes at 1 month -0.13 <.01
Maternal sensitivity from 1 to 54 months 0.02 .77
Interpreting a Regression Table: Question #1
What does the beta for cognitive
Predicting political conservatism at age 18 skills at 54 months mean?

Predictor Beta p A. Having better cognitive skills at


Child’s gender (0 = male, 1 = female) -0.11 .01 54 months is not associated with
Child’s ethnicity 0.17 <.01 political conservatism at age 18.
Cognitive skills at 54 months -0.05 .30
Socioeconomic status at 54 months -0.09 .07 B. Better cognitive skills at 54
Authoritarian parenting attitudes at 1 0.16 <.01 months is not associated with
month political conservatism at age 18
Egalitarian parenting attitudes at 1 month -0.13 <.01 when controlling for gender,
Maternal sensitivity from 1 to 54 months 0.02 .77 ethnicity, SES, and parenting
qualities.
Interpreting a Regression Table: Question #2
What does the beta for egalitarian
Predicting political conservatism at age 18 parenting mean?

Predictor Beta p A. Egalitarian parenting predicts


Child’s gender (0 = male, 1 = female) -0.11 .01 lower levels of political
Child’s ethnicity 0.17 <.01 conservatism at age 18.
Cognitive skills at 54 months -0.05 .30
Socioeconomic status at 54 months -0.09 .07 B. Egalitarian parenting predicts
Authoritarian parenting attitudes at 1 0.16 <.01 lower levels of political
month conservatism at age 18 when
Egalitarian parenting attitudes at 1 month -0.13 <.01 controlling for gender, ethnicity, SES,
Maternal sensitivity from 1 to 54 months 0.02 .77 authoritarian parenting, and
maternal sensitivity.
Interpreting a Regression Table: Question 3

Predicting political conservatism at age 18

Predictor Beta p
A friend looks at these data and says,
Child’s gender (0 = male, 1 = female) -0.11 .01 “The only reason authoritarian
Child’s ethnicity 0.17 <.01
parenting is associated with adult
conservatism is income—richer
Cognitive skills at 54 months -0.05 .30 parents are probably more
Socioeconomic status at 54 months -0.09 .07 authoritarian, and richer parents are
Authoritarian parenting attitudes at 1 0.16 <.01 going to have more conservative
month kids.”
Egalitarian parenting attitudes at 1 month -0.13 <.01
Maternal sensitivity from 1 to 54 months 0.02 .77

What do you say back?


Interpreting a Regression Table: Question 4

Predicting political conservatism at age 18


A different friend looks at these data
Predictor Beta p and says, “The only reason
Child’s gender (0 = male, 1 = female) -0.11 .01 authoritarian parenting is associated
Child’s ethnicity 0.17 <.01 with adult conservatism is because of
Cognitive skills at 54 months -0.05 .30 parental conservatism—conservative
Socioeconomic status at 54 months -0.09 .07
parents are probably more
Authoritarian parenting attitudes at 1 0.16 <.01
authoritarian, and conservative parents
month are going to have more conservative
Egalitarian parenting attitudes at 1 month -0.13 <.01 children.”
Maternal sensitivity from 1 to 54 months 0.02 .77
What do you say back?

You might also like