Levels of Measurement
What are the four levels of measurement?
Nominal, Ordinal, Interval, Ratio
Name the three measures of central tendency.
Mean, Median, Mode
What does a correlation coefficient (r) tell us?
The strength and direction of a relationship between two variables.
Define reliability.
The consistency of a measurement.
Define validity.
The accuracy of a test in measuring what it claims to measure.
What type of variable is “gender”?
Nominal
Which measure is most affected by outliers?
Mean
What is the range of possible correlation values?
From -1.00 to +1.00
A stress test given twice, one month apart, checks what type of reliability?
Test-Retest Reliability
A nursing exam missing key skills lacks what type of validity?
Content Validity
What level of measurement is used for ranking satisfaction (1st, 2nd, 3rd)?
Ordinal
The mean and median differ greatly. What does this suggest?
The data may be skewed or have outliers.
What is the difference between correlation and causation?
Correlation shows a relationship; causation means one variable directly affects another.
Two raters give nearly identical scores. What type of reliability?
Interrater Reliability
A new cholesterol test matches results from a gold-standard lab test. What type?
Criterion (Concurrent) Validity
What makes ratio data different from interval data?
Ratio has a true zero (e.g., weight, age).
What are the three main measures of variability?
Range, Variance, Standard Deviation
If r = -0.85, describe the strength and direction.
Strong negative correlation
Items on a scale measuring the same construct are highly correlated. What type?
Internal Consistency
A depression scale correlates with clinical symptoms. What type?
Construct Validity
A dataset includes income levels ($20K, $40K, $60K, etc.) but has one extreme outlier ($2 million). Which measure of central tendency should the researcher use and why?
Median, because it’s less affected by outliers and gives a more accurate center for skewed data.
If s = 0, what does that mean about the data?
There is no variability. All scores are identical.
If r = 0.7, what is r²? What does it mean?
r² = 0.49, meaning 49% of variance in one variable is explained by the other.
Name one way to improve reliability.
Provide clear instructions, eliminate ambiguous items, or increase number of items.
What is the relationship between reliability and validity?
A test can be reliable but not valid, but it must be reliable to be valid.