PSYCH 100A Lecture Notes - Lecture 5: Dependent And Independent Variables, Internal Validity
Thursday, January 16, 2020
Psych100A Week2
Lecture3-1!
-Distribution triad!
•Real — distribution of scores in a sample!
•Imaginary — distribution of a statistics (sampling variability) !
-Hypothetical cuz we can repeat it a million times!
•Hidden — population / DGP!
-Type 1 Errors!
•Internal validity!
-Different situation than in the real world!
-Modeling Variation!
•Outcome = explanatory variable + other stuff !
•Outcome = other stuff !
•Model: function that provides a predicted score for each observation!
•A “null” or “empty” model to predict a distribution!
-“Empty” because Sample mean is not an explanation!
•The variance of scores around the mean equals the variance of model errors
(residuals) in the empty model!
•The mean doesn’t explain the variance, it reveals it!
1
Document Summary
Distribution triad: real distribution of scores in a sample, imaginary distribution of a statistics (sampling variability) Hypothetical cuz we can repeat it a million times: hidden population / dgp. Di erent situation than in the real world. Modeling variation: outcome = explanatory variable + other stu , outcome = other stu , model: function that provides a predicted score for each observation, a null or empty model to predict a distribution. Empty because sample mean is not an explanation: the variance of scores around the mean equals the variance of model errors (residuals) in the empty model, the mean doesn"t explain the variance, it reveals it.