Health Sciences 2801A/B Study Guide - Quiz Guide: Observational Error, Standard Deviation, Umber

• School: Western University
• Department: Health Sciences
• Course: Health Sciences 2801A/B
• Professor: Christopher Lee

Document Summary

Zero (0) - indicators are completely unrelated and represent nothing other than systematic error (all systematic error score) One (1) - indicators are completely related and represent the construct without any unsystematic error (all true score) If = 1, a person"s true score equals his or her observed score (in the real world, this does not happen) A con dence interval is centred around an adjusted observed score (xa) The width of a con dence interval depends on . The desired level of con dence z-score. The amount of unsystematic error (does the measurement instrument have a number of unsystematic errors?) A z of 1. 96 is used to calculate a 95% con dence interval. * 5% of the time the true score could fall out of this interval. A z of 1. 64 is used to calculate a 90% con dence interval. Used to estimate that alpha would be if the umber of indicators were increased (or decreased)