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. Lecture 7 Outline of Last Lecture I. Correlational Study II. Limits of Corelational Method III. Experimental Method IV. Four Validities Outline of Current Lecture I. Frequency Claims II. Association Claims III. Causal Claims IV. Internal Validity in Experiments V. Internal Validity vs. External Validity Current Lecture Which of the four validities are most relevant for each of the three claims? Frequency Claims • construct validity ◦ how well the variable was measured or manipulated -> accuracy of measurement ■ i.e. how many young adults have depression? ■ construct validity -> measure the depression as accurately as possible ■ have a good operational definition of what depression is ◦ you have to measure the variable reliably ■ if someone comes up depression the first time through, they should come up as depressed a second time ◦ measures can be reliable, but not valid ◦ a measure that is valid must be reliable • external validity ◦ the extent to which the results of a study can be generalized to other situations and to other people PSYC 3980 1st Edition■ i.e. assessing depression in a sample of 200 teenagers -> to what extent could the results of this study be generalized to another sample of teenagers? ◦ a representative sample is necessary Association Claims • construct validity ◦ asking how both of the variables were measured in the study ■ to what extent is stress linked to insomnia? -> measuring both stress and insomnia as variables • external validity ◦ what extent can results be generalized to other situations? ■ is helping behavior related to the number of bystanders present? -> look at where the study is conducted in terms of the situation ■ to what extent would you see the results in a different situation? • statistical validity ◦ what is the probability that the results could have been obtained by chance if there really is no relationship? ■ strength of the correlation coefficient ■ Type I (say there is a true correlation when there is not a true correlation between the variables) and Type II errors (say there is not a so relation between the two variables when there is a relationship not due to chance) ■ is the correlation coefficient statistical significant? Causal Claims • ALL FOUR VALIDITIES ARE IMPORTANT • the two variables must covary ◦ as A changes, B changes • temporal precedence ◦ if A causes B, then A must come before B ◦ manipulated variable • internal validity ◦ how well a study provides a test of a causal relationship ■ since A was the only difference, there are no other possible explanations for why you see the change in B ◦ this is the only claim that sees this kind of validity Internal Validity in Experiments • accomplished by controlling all extraneous variables and by randomly assigning people to different experimental conditions ◦ extraneous factors -> confounding variables ◦ random assignment to condition -> process ensuring that all participants have an equal chance of taking part in any condition of an experiment ■ this is the most important part of the experimental method ◦ example -> does sleep deprivation make you irritable?■ randomly assign people to experimental conditions of sleep deprivation and non sleep deprivation Internal vs. External Validity • representative sample • realism is needed to generalize to other situations • typically -> higher the external validity, the lower the internal