Chapter Five Measurement Research Using Numbers Introducing Communication
Chapter Five Measurement: Research Using Numbers Introducing Communication Research 2 e © 2014 SAGE Publications
Key Concepts • NOIR – nominal interval ordinal ratio • Reliability • Validity • Scales • Numerals - labels such as zip codes • Numbers - can be computed Introducing Communication Research 2 e © 2014 SAGE Publications
“N-O-I-R” Measures can be • Nominal - labels • Ordinal - rank order • Interval - numeric scale • Ratio - numeric scale with a zero point. Introducing Communication Research 2 e © 2014 SAGE Publications
“N-O-I-R” Examples • Nominal - press, radio, television • Ordinal - freshman, sophomore, junior • Interval - age (years): __ 0– 4 __ 5– 9 __ 10– 14 • Ratio - age in years ____ Introducing Communication Research 2 e © 2014 SAGE Publications
Validity & Reliability Measures must have – • Validity – ▫ measure what they are supposed to measure. and • Reliability – ▫ perform consistently. Introducing Communication Research 2 e © 2014 SAGE Publications
Assessing Reliability of Measures • Test - retest. • Split half. • Intercoder or observer reliability. Introducing Communication Research 2 e © 2014 SAGE Publications
Assessing Validity of Measures • Content or ‘face’ - “looks OK. ” • Expert or panel - “looks OK’” to experts. • Construct - theoretically OK. • Convergent - agrees with similar measures. • Divergent - low agreement with different measures. • Predictive - predicts “real world” outcomes. Introducing Communication Research 2 e © 2014 SAGE Publications
Likert Scale Strongly Agree Neutral Strongly Disagree 1. Hybrid vehicles are powerful. ____ _____ 2. Hybrid vehicles reduce dependency on foreign oil. _____ Introducing Communication Research 2 e © 2014 SAGE Publications _____
Semantic Differential Scale The XKJ Bogbuster 300 is Powerful. Expensive Beautiful - Introducing Communication Research 2 e © 2014 SAGE Publications - - Weak Cheap Ugly
Proprietary Measures • Tests and measures developed as commercial products. • Copyright. • May not be used without permission. Examples ▫ SATs ▫ GREs ▫ vocational tests Introducing Communication Research 2 e © 2014 SAGE Publications
Chapter Summary • Measures must have – ▫ Validity and reliability. • Measures exist at four levels – ▫ nominal — ordinal — interval — ratio. • Two common scales – ▫ Likert ▫ semantic differential. Introducing Communication Research 2 e © 2014 SAGE Publications
Vocabulary Review • • • • Coding Concurrent validity Construct validity Content validity Convergent validity Correlation Criterion validity Divergent validity Established measures reliability Expert validity Face validity Internal reliability Interitem reliability Intercoder or • • • • observer reliability Interval Likert scale Measurement Nominal Numbers Numerals Ordinal Panel validity Phrenologist Predictive validity Rank order questions Ratio Reification Reliability Introducing Communication Research 2 e © 2014 SAGE Publications • • • Reliability coefficient Scale Semantic differential “Split half” technique Test-retest Validity
- Slides: 12