PSYCH 312 1st Edition Lecture 5 Outline of Last Lecture I. Designing an experiment II. Variablesa. IVb. DVIII. PredictionsIV. Treatment effectOutline of Current Lecture V. Internal validity VI. External validity VII. Initial steps for protecting validity a. Random selection b. Random assignment VIII. Threats to validity IX. Extraneous variables X. Strong inferencea. Music exampleCurrent Lecture-Internal validity oWant experiments to be internally valid-Test what we think they test-Observe changes in DV, they are due to the IV-External validity oWe want experiments to be externally valid-We want to be able to generalized our finding to the real world -Not so artificial that results no relevance outside the testing arrangement/setting-Initial steps for protecting validity oRandom selection -Participants have an equal opportunity of participating in studyoRandom assignment -All participants have equal likelihood of being assigned to either the control or treatment conditions-Internal & external validity oGoal These notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.-To maximize internal & external validity -Increasing one will decrease the other-Threats to validity oConfounds/confounding variables-Any variables other than the IV that…-Could influence the DV-Vary systematically with the IV-Differ systematically across conditions/grps-Threaten internal validity -Observed differences in DV across conditions may be due to these variables and/or IV-Random assignment helps to control confounding variables related to participants -Helps insure all groups are equivalent on all relevant variables at the beginning of the study so that only difference during the experiment is the manipulation of the IVoExtraneous Variables -Variables other than the IV that.. -Could potentially affect the DV-Do not vary systematically with the IV-Threaten to internal validity -Increase variability in performance (DV) & makes it difficult to see potential effect of IV on DV-Random assignment will help equate the variability across conditions but will not elimate it -Threaten to external validity -B/c sample may behave diff from general population-Procedure for examining hypothesis: strong inferenceo4 steps of strong inference-Devise a set of hypothesis (null/alt)-Devise an experiment to test these hypotheses so that diff possible outcomes exclude 1 or more hypothesis -Carry out well controlled experiment & note results -Return to step 1 with further refinements-Strong inference: Music example oDevise a set of hypotheses (null & alt)-HO: (Null) Music of any kind (IV) does not decrease problem-solving (DV) relative to when no music is played.-H1: Any kind of music (IV) decreases problem-solving (DV) relative to when “no music” is played .-H2: Only “rock” music (IV) decreases problem- solving (DV) relative to when no music is playedoDevise an experiment to test alternative hypotheses so that alternative outcomes (if observed) can exclude 1 or more hypothesis.-Need at least 3 conditions -No Music-Rock Music-Classical-DV = time to accurately complete a problem-solving task.oCarry out experiment & note results -Find support for H2-Rock only decreases performance relative to other 2 conditions which arenot diff from one anotheroReturn to step 1 with further refinements of the supported hypotheses -Example refinement -Does it matter if rock music contains lyrics -Repeat process-Changing IV or adding IV-Re-run