thinking like a psychologist part 33 1 / 3. roll call essay #2 (missing several) apa mastery test...
TRANSCRIPT
Agenda
Roll call Essay #2 (missing several) APA Mastery Test – see me if
questions Review Control, Research Methods, and
more Intuitions Test Probability, Chance, and the Popular
Image of Psychology
Review
Does correlation prove causation? Why or why not?
What is the principle of faslifiability? What are the problems with
testimonial evidence?
Control
Necessary to make causal inferences and rule out alternative explanations
When a researcher holds everything in 2 (or more) different situations constant except for a particular variable
Requires that we separate and individually control variables that may naturally occur together
Then if the outcome changes, the only explanation is the variable that changed… Outcome = Dependent Variable Manipulated variable = Independent Variable
Control (cont’d)Variables Quiet Condition Noisy Condition
Noise Level (IV) Low High
IQ (EV) Average Average
Room Temperature (EV) 70° 82°
Sex of Subjects (EV) 60% Female 60% Female
Task Difficulty (EV) Moderate Moderate
Time of Day (EV) Morning Afternoon
Control (cont’d)
The Control Group A group of participants that receives no (or
alternative) treatment Why important?
Examples of importance Clever Hans Facilitated Communication
Separation of Variables Must rule out alternative explanations; therefore
must create artificial situations where variables that naturally occur together are teased apart.
Video – Prisoners of Silence
Intro to Research Methods
Research Methods
There are 6 basic categories of scientific method that virtually all research falls intoResearch
Non-Experimental Experimental
Naturalistic Observation Quasi-Experimental
Survey
Relational Research
Field Study Experimental
Research Methods –Naturalistic Observation
Addresses most basic scientific question: “What is out there?”
Requires operational definition of events to be observed
Observer must be unobtrusive, and design must be nonreactive
Research Methods –Field-Based Research
Like naturalistic observation, conducted in real-world settings
Goal is to establish natural relations among events
Observer must be unobtrusive, but methods are intentionally reactive
Research Methods –Survey Research
Appropriate to the study of private behaviors
Two primary styles: Interviews (structured/unstructured) Questionnaires
(structured/unstructured)
Research Methods –Relational (Correlational) Research
Goal to verify systematic (usually linear) relations among events
Strengths/directions of relations generally expressed in form of
correlation coefficient (rxy)
Research Methods –True Experiment
Goal: to establish a cause-effect relationship among events Does low-fat diet cause decrease in
cancer risk? Does exposure to violent video games
cause increase in violent behaviors? Does spaced study cause increase in
memory accuracy and retention? Do genetic variations cause sexual
preference?
Research Methods –True Experiment
Requires:random assignment of participants to at least two
equivalent conditionsmanipulation of one factor (independent variable,
or IV) in one condition (experimental), leaving it unchanged in other condition (control)
measurement of one other factor in both conditions (factor called dependent variable, or DV; measurement instrument called dependent measure, or DM)
Research Methods –True Experiment
Concludes:if groups are NOT equivalent with
respect to DV, and if the difference between the groups is so big it
probably did not happen by chance, thenmanipulation of the IV caused the difference in the
DV
Research Methods –Quasi-Experiment
Goal also to establish cause-effect relations among events
Required when random assignment is not possible, becausemust use pre-existing groups, orIV impossible to manipulate directly, orIV unethical/illegal to manipulate directly
Research Methods Review
Name 6 categories of scientific research
Which method of research can be used to establish cause and effect relationships?
Gradual Advancement
Converging Evidence
Gradual Synthesis Model Great Leap
Remember, the goal of science is really to get less and less WRONG.
Multiple Causation
Most phenomena are caused by multiple (not just one event)
Examples? There’s usually no “magic bullet” or
“missing key”