PSYCH 270 Lecture Notes - Lecture 7: Confounding, Null Hypothesis, Statistical Significance
Document Summary
Behavior is complex no simple answers. Scientists make mistakes just like everyone else! Control (i. e. treatment) is the element that distinguishes psychopathology from normal psychology. Internal validity: low when results are confounded. External validity: low when results cannot be generalized. Confounding variable = something that influences all other variables behind the scenes . Need to develop a testable hypothesis - a tentative explanation about how or why something happens. E. g. full moon is associated with violent behaviors. E. g. violent behaviors = number of homicides. Distinguishable variables in a hypothesis (if x , then y ) Use of random assignment to reduce confounding in experiments. Distal = farther away in vicinity or timeline. Experimenter effects (self-fulfilling prophecy) - you get what you look for. Placebo groups: give one group a sugar pill as opposed to real medication. Use double-blind: neither the experimenter nor the participant(s) knows who is in what condition. Null hypothesis: independent variable has no influence on the.