Measure Treasures
These are the four different types of measurement scales.
What are nominal, ordinal, interval, and ratio?
A causal claim must satisfy these three criteria.
What are covariance, temporal precedence, and internal validity?
Range, variance, and standard deviation all describe this property of a dataset.
What is variability / spread?
You reject the null when p falls below this preset value.
What is alpha?
This test compares one sample mean against a known population value.
What is a one-sample t-test?
Psychologists use these three broad methods (types of measures) to operationalize conceptual variables.
What are self-report, observational, and physiological measures?
This kind of validity concerns whether a study’s results generalize beyond the sample.
What is external validity?
This rule states that about 68, 95, and 99.7% of scores fall within 1, 2, and 3 SDs of the mean in THE normal distribution.
What is the empirical rule?
This error means concluding there’s an effect when there really isn’t one.
What is a Type I error?
This test compares two separate groups made up of different people.
What is an independent-samples t-test?
This is the type of reliability you check when the same people take a measure twice over time.
What is test-retest reliability?
Random sampling protects this validity (1), while random assignment protects this other one (2). (two answers here)
What are external validity and internal validity?
This standardized score tells you how many standard deviations a value sits from the mean.
What is a z-score?
This is the standard deviation of a distribution of sample means.
What is the standard error?
This test compares the same people measured twice, or matched pairs.
What is a paired / dependent-samples t-test?
This big-picture form of validity asks whether a measure truly captures the concept it claims to.
What is construct validity?
This unmeasured variable can be the real reason two things appear to correlate.
What is a third / confounding variable?
This type of value drags the mean toward it but barely moves the median.
What is an outlier?
This range gives plausible values for a population parameter, often reported at 95%.
What is a confidence interval?
This statistic, r, captures the direction and strength of a linear relationship.
What is the correlation coefficient?