Chapter 5 - Measures Flashcards
Operationalization and Suffrage
Should it be understood nominally? As in you have it, or you don’t? Or perhaps ordinally, like a continuum?
Measurement Invalidity
A conceptual definition that does not match the operationalized definition.
Concept
Abstraction Representation
Variables
Concrete manifestations (of abstractions)
Indicators
Operational measure of a variable.
3 Main Ways to Measure Variables
Nominal
Ordinal
Interval
Nominal Measurements
Categorical Cannot be ranked Anything with yes or no answer Forced choice Least precise Ex.: religious affiliation, gender
Ordinal Measurement
Order
Relative to each other, but don’t know difference between categories
Ex.: Low, mid, high education; disagree, neutral, agree
Sorta precise
Interval Measurement
Rankings have precise (known) distances between categories
‘Hard numbers’
Most precise
Ex.: Age cohort: 50-59, 60-69, 70-79, etc.
Vagueness and Operationalization
An expression is vague when there is no precise boundary between the cases to which it correctly applies and the cases to which it does not.
Miniscule differences may not seem to matter.
Two Types of Measurement Error
Random Error: Inaccuracies caused by factors that are not systematic and/or intential; found in all samples because the full range of possible respondents cannot be included; amount of random errors can be estimated in a probability sample
Non-random (systematic) error: Systematic error that, for example, occurs because people lie; voter turnout is often affected by this error, as many people say they voted even though they didn’t.
3 Ways to Critically Assess Measures
Reliability
Validity
Cultural Sensitivity
Reliable Measuring
How do I know that the test, scale, instrument I use works every time I use it?
3 Ways to Ensure Reliable Design
Increase Precision Use multiple measures (indexes) Use careful design - Measure on thing at a time - Exhaustive categories - Mutually exclusive categories
Two Types of Validity
Face Validity (do the items seem to measure the concept) Construct Validity (assess if instrument really measures the construct and not another)
3 Ways to Design Valid Measures
- Ensure clear conceptual definition
- Increase number of measures
- Increase abstraction
3 Considerations for Culturally Sensitivity
- Linguistic Equivalence
- Conceptual evidence
- Metric equivalence
Values
Categories that capture the variance between observable characteristics of phenomena; for effective coding to be effective, values should be collectively exhaustive and mutually exclusive
Complex multiple indicators
A combination of variables such as the aspects of a person’s political participation; into an index or scale; used to provide a more complete description or measurement of a concept
Indicator
A variable that used to represent the presence of a quality
Index
A complex measure that combines responses from more than one question in the creation of a new measure
Closed v Open Questions
Open = can provide own answer Closed = pick from a selection
Response Sets
The group of response categories that comprise all responses to a particular question.
To be avoided cause yay/nay-sayers
Exhaustive
Inclusion of all possible causal factors in an explanatory model; also inclusion of all possible response options in a measure
7 Characteristics of Good Question Design
- Use neutral language.
- Be clear.
- Avoid response sets.
- Keep response categories mutually exclusive and exhaustive.
- Select the highest reliable level of measurement.
- Pay close attention to question order.
- Minimize defensive reactions.
Operationalization
The process of moving from a conceptual definition (abstraction) to a measure or a set of measures (concrete) that enables a researcher to empirically observe the construct in one’s particular research project
Simple v Complex Concepts
Simply will require less measures (age - years of life - years born) compared to complex concepts (Social status - education, income, occupation - highest level of education obtained, etc.)
5 Ways to Assess Measurements (I think 3 are like… idk more important or something)
- Vagueness
- Measurement Error
- Reliability
- Validity
- Cultural Sensitivity
Vagueness / Precise Thresholds
An expression is vague when there is no precise boundary between the cases to which it correctly applies and the cases to which it does not
