# Fundamentals of Measurement

## 29 terms · Vocabulary from Chapter 4 of Research Methods

### nomological net

the theoretical network of construct to construct associations derived from relevant theory and stated at an abstract level

### true score

function of the construct we're trying to measure

### systematic error

influences from the other constructs besides desired one

### random error

non-systemic, ever changing influences

### observed score

true score + systematic error + random error

### social desirability response bias

general tendency to over-report one's desirable behaviors and characteristics

### correlation coefficient

statistical index of the strength of association between two variables; "To what extent do two variables measure the same thing?"

### reliability

the extent to which a measure is free of error; unreliable score can't have validity; precisiin
(e.g. students taking a test in a language they don't speak, guessing; all differences in scores attributable to chance)

### validity

the extent to which a measure reflects only the desired construct w/o contamination from other systemically varying constructs; accuracy

### Constructs

abstractions that soc. scientists discuss in theories e.g. status, power, intelligence; any one construct can be measured in many ways -- variety of representations for the idea; any one variable gives an indication about the construct, but not the whole thing

### Variables

representations of constructs; any construct has many different variables; partial representations of constructs that we work with because they are measurable

### Operational Definitions

specifies how to measure a variable so that we can assign someone a score as high, medium, or low score for a construct; steps and procedures a researcher follows to get a measurement -- numerical or categorical

### Definitional Operationism

flawed assumption that the operational definition is the construct (e.g. intelligence is what an intelligence test measures)

### Test-Retest Reliability

correlation of the scores on the same measure administered on 2 separate occasions; provides estimate of measure's reliability

### internal consistency reliability

alternative estimate of reliability; assume that all the questions or items on the measure are measuring the same construct, so lack of correlation among specific items serves as the basis for an estimate of the influence of random error

### split-half reliability

set of items in measure split in half, then compared for internal consistency reliability; problem- exact estimate of reliability depends on how items are split up --> arbitrary

### coefficient alpha

preferred measure of internal consistency reliability; estimate derived from correlations of each item w/ each other, not arbitrary; ranges from 0 to 1, 0 completely unreliable, 1 perfectly reliable

### Inter-Rater Reliability

Multiple observers, can measure reliability by measuring agreement btw observers; can treat observer ratings like items on a questionnaire and use coefficient alpha

### Kappa

measure of agreement /inter-rater reliability when observer ratings are yes/no (binary)

### face validity

judges look at measuring technique and decide whether in their opinion it measures what the testers claim

### convergent validity

overlap between alternative measures that are intended to tap the same construct but have different sources of systemic error; how much they overlap --> how much they measure the actual construct

### discriminant validity

valid measure shouldn't correlate with measures that are supposed to tap basically different constructs (measure shouldn't correlate with other measures that pick up expected sources of systemic error )

### multitrait-multimethod matrix (MMTM)

a table of correlation coefficients, let you evaluate convergent and divergent validity of a construct at the same time; need at least two constructs measured with at least two methods to make one

### traits

the underlying construct the measurement is supposed to tap

### methods

the mode of measurement

### reliability coefficients

correlations between scores that reflect the same trait and the same method

### convergent validity coefficients

correlations between scores that reflect the same trait measured by different methods

### discriminant validity coefficient

correlation between different traits measured by the same method

### nonsense coefficient

correlation between different traits measured by different methods