Measurement Scales
Nominal - define and provide an example
does not have a relative value; categorical only
gender, race, type of ____
What type of validity describes whether or not the measure fully captures what it's supposed to?
Content Validity
Characteristics of experimental design
random assignment, control and experimental groups with manipulation of independent variable, high internal validity
Anything less that 0.05
Conceptual variable vs. Operational variable
conceptual - what you mean (trust in the government)
operational - what you measure (on a scale from 1-5 how much do you trust the government)
Ordinal - define and provide an example
ranked without a set distance between values
strongly disagree --> strongly agree, education level
What type of validity describes whether or not the measure relates to other variables as it should (in theory)?
Construct Validity
Characteristics of quasi-experimental design
no random assignment (often pre-existing groups), control and experimental groups with manipulation of independent variable, moderate internal validity
What is a null hypothesis (H0)?
a statement that there is no effect or no relationship between the variables being studied.
Ex. H1 - people who pass comps have higher salaries
H0 - there is no relationship between passing comps and salary
NOT - people who pass comps have lower salaries
an alternative explanation for relationships between X and Y
Interval - define and provide an example
ordered with equal intervals, with no true 0
years, temperature (0° doesn't mean no temperature)
What type of validity describes whether or not the independent variable caused the outcome?
Internal Validity
Characteristics of non-experimental design
observational, no manipulation of independent variable, no random assignment, no control group, high external validity, low internal validity
What is a p-value?
probability that your results happened by chance; can range from 0 to 1; the smaller the value, the more likely the results are evidence of your experiment, not an accident
What are descriptive statistics?
a description of the measure of central tendency (mean, median, and mode), variability (interquartile range, percentiles, standard deviation), and distribution (skew)
Ratio - define and provide an example
ordered with equal intervals, with a true 0
income, age
What type of validity describes whether or not the measure could be generalized beyond the study?
Which type is the gold standard for research design and why?
Experimental - because it establishes causal relationships by controlling for outside variables and using a random assignment
What is an alpha value?
the rule you set before you start the experiment - the cutoff for what is good enough to reject your null hypothesis; can range from 0 to 1; the lower the value, the stricter the experiment
What is a time-series design?
collects data on the same variable at multiple points in time at regular intervals to observe patterns
Ex. voter turnout every year
What is measurement reliability?
the consistency, stability, and dependability of the measurement instrument; if you measure the same the thing under the same conditions, you should get the same result
Types of measurement error
Random - happens by chance, makes it less reliable but not due to bias; ex. data entry mistake
Systematic - consistent bias error, threatens validity; ex. leading questions
What is a cross-sectional design?
collects data at a single point in time to examine relationships between variables
Ex. education level and voter turnout
What are inferential statistics?
the measurements we use to make assumptions about the subjects (p-value, alpha value, Z-score)
What is a panel design?
collects data from the same subjects at multiple points in time to study individual change
Ex. political affiliation at 18 and at 28