PSYC 490 Lecture Notes - Lecture 1: Observational Error, Internal Consistency
Document Summary
There"s no such thing as a perfect (error-free) measurement. Thus, a person"s final score on a measure = true score + error. Systematic error: measures are influenced by other conceptual variables not part of the conceptual variable under study. Formal definition: the extent to which a measure is free from random error. Test/retest: extent to which scores on the same measure, administered at two different times, correlate with each other, susceptible to practice effects, fatigue, & other threats. Alternative-forms: extent to which scores on two different measures of the same construct, administered at two different times, correlate with each other, requires independent confirmation of validity. Design and analysis: measure as a whole will be more reliable than any individual item. Split-half reliability: correlate a person"s score on one half of the items with his or her score on the other half of the items.