# Ch 5

## 23 terms

### Numerals

labels for phenomena (street addresses, brand name, a new commuter jet, etc)

### Numbers

assign value and relativity to phenomena

### Nominal Measurement

labeling or classification, purely qualitative (male/female)

### Ordinal Measurement

indicate some level of progression (freshman, sophmore, junior, senior)

### Interval Measurement

intervals between values are the same; numeric scale (distance between 1-2 = distance between 84-85)

### Ratio Measurement

numeric scale with a zero point

### Reliability

measure of the extent to which a test performs consistently

### Test-retest

common test for reliability: test is given to group of people and then repeated with the same group at a later time

### Correlation

statistical procedure for measuring the strength of association between two or more variables

### Observer Reliability

measure of the extent to which observers are in agreement in their coding of observations

### Validity

measure of whether a test measures what is is supposed to measure

"looks ok"

### Expert/Panel Validity

"looks ok" to experts

### Convergent Validity

agrees with similar measures

### Divergent Validity

measure of a construct is unrelated to measures of unrelated concepts

### Criterion Validity

relates your measures to other specific measures in two ways

### Concurrent Validity

demonstrated when a measure correlates highly with other measures designed to measure the same construct

### Predictive Validity

predicts real world outcomes

### Construct Validity

theoretically "ok"

### Likert Scale

interval scale on which respondents record their reaction to a statement by checking their level of agreement between "strongly agree" and "strongly disagree"

### Semantic Differential Scale

scale anchored at opposite ends by opposing words such as "strong-weak" or "hot-cold"

### Independent Variable

variable whose changes in values result in changes in another (dependent) variable

### Dependent Variable

variable whose value changes as a result of changes in another (independent) variable