Unformatted text preview: • Investigate • Control • Confounds vs. Noise o Confound – varies systematically with IV o Noise – can water-down results but does not mislead you o Noise does not have to happen to everyone, whereas confounds have to happen to everyone in the experiment People are Different (another potential threat to validity • Experimental Control • Selection Bias o Technology Survey • Non-response Bias o Should participation in class discussions be graded? People who do not participate won’t voice their opinions People Change • History • Maturation – many disorders appear differently depending on age • Regression Towards the Mean – extreme scores on either end go towards the mean over time...
View Full Document
- Spring '08
- external validity, internal validity, Low External Validity, o Correlational Research o Experimental Research Validity