Class Notes (1,100,000)
CA (650,000)
UTSC (30,000)
Psychology (9,000)
PSYB01H3 (200)
Lecture 7

PSYB01H3 Lecture Notes - Lecture 7: Confounding, Construct Validity, Signify


Department
Psychology
Course Code
PSYB01H3
Professor
Connie Boudens
Lecture
7

This preview shows page 1. to view the full 5 pages of the document.
PSYB04 – Foundations in Psychological Research
CHAPTER 9: Multivariate Correlational Research
Introduction
- Because correlation is not causation, what are the options? Researchers have developed some techniques that
enable them to test for cause. The best of these is experimentation: Instead of measuring both variables,
researchers manipulate one variable and measure the other.
- Research often begins with a simple bivariate correlation, but since bivariate correlations cannot establish
causation, researchers use other techniques that help them get closer to making a causal claim.
Reviewing the Three Causal Criteria
- *Multivariate Designs: involves more than 2 dependent /measured variables
- The 3 criteria for establishing causation are covariance, temporal precedence, and internal validity
Establishing Temporal Precedence with Longitudinal Designs
-Longitudinal Design: an observational research method which the data collected, concerns the same subjects
repeatedly over a period of time
- This design can provide evidence for temporal precedence by measuring the same variables in the same people at
several points in time. Often, longitudinal research is used in developmental psychology to study changes in a trait
or an ability as a person grows older. In addition, this type of design is adapted to test causal claims.
Interpreting Results from Longitudinal Designs; Since there are more than two variables involved, a multivariate
design gives several individual correlations, referred to as cross-sectional correlations, autocorrelations, and cross-lag
correlations.
1) Cross-Sectional Correlations
-cross-sectional correlations: test to see whether two variables, measured at the same point in time, are correlated
-Example: First look at the correlations of the variables when measured at the same time. In third grade, preference
for TV violence is correlated with aggression; in thirteenth grade, these two variables do not appear to be
correlated. In these cross-sectional correlations, there is no way to know which of the variables came first in time.
2) Autocorrelations
- The next step was to evaluate the associations of each variable with itself across time
-autocorrelations: determine the correlation of one variable with itself, measured on two different occasions
-Example: Although TV violence preferences are not stable over time, aggression levels appear to be somewhat
stable. The third-grade measurements came before the thirteenth-grade measurements.
3) Cross-Lag Correlations
- cross-sectional correlations and autocorrelations are generally not the researchers’ primary interest, but cross-lag
correlations are the ones researchers are interested in
-cross-lag correlations: show whether earlier measure of one variable is associated with the later measure of other
variable
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version

Only page 1 are available for preview. Some parts have been intentionally blurred.

PSYB04 – Foundations in Psychological Research
- inspecting the cross-lag correlations in a longitudinal design, we can investigate how people change over time and
therefore establish temporal precedence
-Example: the cross-lag correlations in this study suggest that viewing violent TV shows causes aggression,
because a preference for TV violence in third grade predicts later aggression, but aggression in third grade does
not predict later preferences for TV violence.
Longitudinal Studies and the Three Criteria for Causation
Longitudinal designs can provide some evidence for a causal relationship by means of the three criteria for causation:
1) Covariance; significant relationships in longitudinal designs help establish covariance. When two variables
are significantly correlated there is covariance.
2) Temporal Precedence; longitudinal design can help researchers make inferences about temporal precedence.
Because each variable is measured in at least two different points in time, they know which one came first. By
comparing the relative strength of the two cross-lag correlations, the researchers can see which path is
stronger. If one of them is stronger, the researchers move a little closer to determining which variable comes
first, causing the other.
3) Internal Validity; when conducted simply, that is by measuring only the four key variables, longitudinal
studies do not help rule out third variables.
Why Not Just Do an Experiment?
- an experiment is the best way to confirm or disconfirm causal claims, however the problem is that in many cases
people cannot be randomly assigned to a variable
- one reason is assigning preferences eg. people either like/dislike horror films so it’s hard to manipulate this
variable
- second reason would be ethics eg. children are under a study whom are forced to watch horror films for 10 yrs. in
order to collect data about aggression and how it affects people over certain time period
Ruling Out Third Variables with Multiple-Regression Analysis
*Multiple/Multivariate Regression: a statistical tool used to derive value of a criterion from several other
independent/predictor variables
Measuring More than Two Variables
- by measuring all possible variables instead of just two (with the goal of testing the interrelationships among them
all), a multivariate correlational study was conducted
oconducting a multivariate design, researchers can evaluate whether a relationship between two key variables still
holds when they control for another variable
Regression Results Indicate If a Third Variable Affects a Relationship
Criterion Variables and Predictor Variables
find more resources at oneclass.com
find more resources at oneclass.com
You're Reading a Preview

Unlock to view full version