MARK 302 Lecture Notes - Lecture 6: Observational Error, Test Validity, Internal Consistency
Document Summary
Chapter 6: the concept of measurement and attitude scales. Ordinal: there is an order relationship between data (i. e. first, second, third; small, medium, large; best liked, worst liked) Interval: there is an order relationship between data and the difference between data points is significant (i. e. scale of 1 to 10, age, annual income: zero has no value, it represents the absence of whatever is being measured. Ratio: there is an order relationship between data, the difference between data points is significant and zero is meaningful (i. e. temperature) Where m = measurement, a = complete accuracy, e = errors. Reliability: the degree to which measures are free from random error and provide consistent data. Ways to test validity: convergent validity: the degree of correlation among different instruments that are used to measure to same construct, discriminant validity: a measure of the lack of association among constructs that are supposed to be different.