Kinesiology 2032A/B Lecture Notes - Lecture 14: Categorical Variable, Criterion Validity, Biomedical Model

• School: Western University
• Department: Kinesiology
• Course: Kinesiology 2032A/B
• Professor: Lindsay Nagamatsu

Document Summary

Yes: you just need a common scale, sweetness, vitamin content, pleasantness of aroma, firmness of flesh, this is a measurement issue, not really different from measuring clinical variables. Independent vs. dependent: predictor vs. criterion, continuous vs. discrete. Levels of measurement: nominal, categorical data with no implicit ordering, unequal distance between points, ordinal, categorical data with implicit ordering, unequal distance between points. Interval: continuous (equal distance between points, no meaningful zero, ratio, continuous (equal distance between points, meaningful zero. True scores and errors: any observed score (x) is comprised of two distinct components, true score (t, error component (e, x = t + e, the difference between t and x is measurement error. Overview of error management: reliability, validity, construct validity. Internal validity: external validity, statistical conclusion validity, utility, efficiency, generality. Reliability and validity: reliability imposes an upper limit on validity. Is the measure stable across: time, circumstance.