PSYCH 209 Lecture Notes - Lecture 9: Verbosity, Criterion Validity, Likert Scale
4/16/18 Lecture Notes
Review: Types of Construct Validity
• Face validity
o Taking your measure and showing it to an expert and seeing if the measure shows what it's
supposed to
• Content validity
• Criterion validity
o Whether or not your measure correlates with concrete outcomes
• Convergent validity
o How well a measure correlates with things that it should
• Discriminant validity
o How well a measure does not correlate with things that it shouldn't
Behavior Observations
• Suppose you want to understand some psychological variable better. (Ex: Memory, prejudice, self-
esteem)
• What are some reasons to choose to observe and measure behavior rather than using another
method of measuring variables (Ex: self report)?
"Clever Hans"
Behavior Observations: Issues to Consider
• Observer Effects
o Presence of an observer can affect the way participants behave
o Solutions
• Masked (blind) design; unobtrusive observations
• Observer Bias
o Seeing what we expect to see
o Perception can be biased in the direction of our beliefs
o Solutions
• Minimize prior expectations + specific operationalization
• Ideas for addressing these issues?
o One-Way Mirror
• Observer can see participants, but participants can't see observer
o Double-Blind study
• Observers aren't aware what the study is for, but is just watching for certain behaviors
Behavior Observation: Holding the Door
Hold the Door: An Observational Study
• The maximum distance between people would be 9 feet for holding the door for a stranger
• Opening the door and standing by the door until the person passes through
Surveys: Construct Validity
• Open-ended
• Forced-choice
• Likert Scale
find more resources at oneclass.com
find more resources at oneclass.com