2. Measurement allows researchers to make comparisons.
3. Discrete data are known as categorical or nominal data and describe the presence or absence of some characteristic or attribute.
4. For categorical data, each variable is comprised of two or more classes or categories that should be mutually exclusive, exhaustive, and equivalent.
5. Continuous level data can be one of three types: ordinal, interval, or ratio.
6. Ordinal data rank the elements in some logical order, but without knowing the relative difference between ranks.
7. Interval data are more sophisticated in that they represent a specific numerical score and the distance between points is assumed to be equal.
8. Ratio data are the most sophisticated data type; they have the characteristics of interval data and a true zero.
9. Issues of validity and reliability are associated with all types of measurement.
10. Data are valid to the extent that they measure what you want them to measure.
11. Face validity exists when the measurement reflects what we want it to.
12. Content validity exists when the measurement reflects all possible aspects of the construct of interest.
13. Criterion-related validity exists when one measurement can be linked to some other external measurement.
14. Construct validity exists when measurement reflects its theoretical foundations.
15. Reliability is the degree to which measurement is dependable or consistent; it is expressed as a matter of degree.
16. Internal reliability is achieved when multiple items purportedly measuring the same variable are highly related.
17. Test-retest reliability is achieved when measurements at two different times remain stable.
18. Split-half reliability is achieved when two forms of one measuring instrument capture similar results.
19. Measurement of data must be both valid and reliable.
20. Validity and reliability are threatened by the choices researchers make about how they collect data, and whom or what they choose as their sample, as well as alternative explanations that are plausible.
21. Regardless of how data are collected, they must be collected and reported accurately, ethically, and responsibly.
