Unformatted text preview: of variables
• Issues of reliability and validity
validity Determining Cause & Effect:
The Experimental Approach
Variable •Random assignment + Control of
extraneous variables = Internal Validity
•Representative sample +
Psychological realism = External Validity
External The Experimental Approach - An
Example: Rosenthal & Jacobsen
Expectancy Change in
IQ Score •“Bloomer” label assigned randomly + students
/parents unaware of label = High Internal Validity
•Sample of varied ethnicity, ability etc. + Real
classroom, real IQ tests = High External Validity
External Did Rosenthal & Jacobsen
prove SFP to be true?
• Could this outcome be due to chance
variation and not the effect of the IV?
What is the likelihood that this outcome is
due to chance?
If it’s < 5% (i.e., p < .05), we decide that the
result is statistically significant
What factors affect significance?
• Magnitude of effect
• Variability in data
• Sample size We NEVER prove anything in
psychological research because We study samples, not populations
We draw inferences about populations
using data obtained from samples
Therefore, any result can be due to chance
It’s always a question of probability, never
of Statistical vs.
Practical We have Statistical Significance whe...
View Full Document