CRIM 220 Lecture Notes - Lecture 4: Theoretical Definition, Operationalization, Observational Error
13 Jun 2016
School
Department
Course
Professor
Document Summary
Conclusion: whether 2 variables are statistically related to each other. Internal: whether the relationship between the 2 variables is causal, not due to a 3rd variable. Construct: whether our variables properly represent out constructs. Works/symbols used to present these mental images. Specifying precisely what we mean when we use particular terms. Conceptional definition: working definition assigned to a construct. Ivs: # hours spent studying - dvs: midterm exam score. Ivs: time spent in class on facebook - dvs: course grade. Able to classify every observation in terms of 1 of the attributes. Be able to classify every observation in terms for one and only one attribute. Qualitative labels that do not represent anything except differences among groups. E. g race, gender, prov of birth: ordinal measures. Attributes are rank-ordered on a continuum; represent relatively more or less of the variable. No info about distance betw. the values: internal measures. Distance (i. e interval) that separates the attributes has meaning.
