PS295 Lecture Notes - Partial Correlation, Cognitive Dissonance, Social Desirability Bias
Week 7 – Lecture 11
PS295 / June 18, 2018
Review:
-cannot use causal terms in correlational research
-correlational research only used to calculate whether there is a relationship, not the direction
-causation research is done through experimental research
Correlation coefficient can be distorted b:
-offline outliers
-inline outlier
r squared is also known as:
-the coefficient of determination
Effect of x on y is influenced by z, studied through:
-partial correlation analysis
Descriptive Research:
-conducted to describe how participants think, feel and behave
Correlational Research:
-conducted to study the relationship between variables
-no manipulation of variables
-only measurement
Experimental research:
-conducted to study the cause and effect relationship between variables
Cognitive dissonance theory:
-people are distressed when their beliefs are inconsistent with each other
-motivated to reduce this distress by making their cognitions consistent
-found that women who were put through severe initiation reported liking members more than women
put through mild initiation
Web Based Experimental Research:
-online methods used to conduct surveys, correlational studies, and experimental studies
Pros:
-access to larger samples
-access to more diverse samples (age, race, ethnicity, education)
-easier to obtain samples with specific characteristics
-generally require less effort, time, and money
-less susceptible to social desirability biases and experimenter expectancy effects
Cons:
-difficulty controlling nature of sample (ex participating more than once)
-cannot control environment in which study is completed
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Correlational research only used to calculate whether there is a relationship, not the direction. Inline outlier r squared is also known as: Effect of x on y is influenced by z, studied through: Conducted to describe how participants think, feel and behave. Conducted to study the cause and effect relationship between variables. People are distressed when their beliefs are inconsistent with each other. Motivated to reduce this distress by making their cognitions consistent. Found that women who were put through severe initiation reported liking members more than women put through mild initiation. Online methods used to conduct surveys, correlational studies, and experimental studies. Access to more diverse samples (age, race, ethnicity, education) Less susceptible to social desirability biases and experimenter expectancy effects. Difficulty controlling nature of sample (ex participating more than once) Cannot control environment in which study is completed. Limited to certain study designs: research varies/manipulates at least one independent variable (iv) and measures at least one.