Quiz 2 Flashcards
Why is measurement necessary
to empirically test utility of a scientific explanation. (Note that different measurements can yield different conclusions)
Measure
Quantitative representation of a concept
Operationalization
The process of transforming concepts into measurable indicators
Nominal Measures
Values assigned represent different categories, without order or hierarchy
Ex: ethnicity, religion, place of birth
Ordinal Measures
An order among the categories (possible values)
Distance between the categories is not necessary equal
Ex: income categories in surveys
Interval Measures
Values are continuous (any value in a range)
The distance between values is real, known, and meaningful
Ex: age, rainfall (in inches), income
Ratio level Measures
Includes the values of categories, the order of the categories, and the intervals between them
Also includes the relative amounts of
Have a defined absolute 0 value, so fractions/ratios
Validity
Degree to which the measure matches the concept it is thought to measure
Reliability
The extent to which a procedure or measure in repeated test produces consistent results
Measurement Error
Difference between the true value of an object and the observed value
Random Error
Has no systematic direction or cause.
Doesn’t affect the average, only the variability around the average
Systematic Measurement Error
Results in systematically over or under-measuring the value if a concept
DOES affect the average and is therefore much more problematic
Measures of Central Tendency
Mean, median, and mode
Dispersion (measures of spread)
St. Dev, interquartile range, range
Frequencies
How often a value occurs in the data. Can be a raw count or a proportion
Range
Works best for ordinal and interval variables
Mean works best for
Interval/ratio
Median works better for
Ordinal variables
Mode works best for
Nominal and ordinal variables
IQR works best for
Ordinal and interval variables
Variance
The average squared differences between values and the mean. Works for interval variables
Standard Deviation
The square root of the variance. Works best for interval variables
Steps to specify and test relationships between variables
Propose relationship
Develop a causal theory
Frame hypothesis
Collect data
What makes a good theory
- Causal
General
Consistent with existing knowledge
Internally consistent
Explains multiple outcomes
Makes falsifiable claims
Bolet et al main finding
Just Transition Agreement for phasing out coal was electorally successful, driven by unions’ support of the JTA
Tai Li Desmarais Main Findings
Republican legislators share misinformation at higher rates, with certain states emerging as misinformation hotspots.
Legislators in professional legislatures are less likely to spread misinformation