Regression
These are prediction errors. On a graph it is the vertical distance from the point to the line. It’s represented by the equation y-ŷ
What are residuals?
True or False: Both x and y are categorical on a contingency table
True
Numerical value that can be averaged and used for a minimum or maximum
Quantitative variable
Experimental vs. Observational
Experimental - researchers control some variables
Observational - researcher only observes x and y
What‘s a response variable
The variable that’s being affected
What is R^2 and what is it’s purpose?
Coefficient of Determination
= (r)^2
categorical relationship between x and y
percentage of variability in y explained by linear regression on x
Name a graph that holds categorical data
Bar charts
Pie charts
What does SOCS stand for?
Shape
Outlier
Center
Spread
Explain Confounding effects
The effects of multiple variables on a response can’t be distinguished
Explain Sample vs. Population
Sample - subjects for data
Population - the general subjects
Explain correlation and how to interpret ”r”
. Summarizes direction and strength .
. r is between -1 and 1
. -1 is strong negative correlation
. 1 is strong positive correlation
. 0 is weak correlation
What is RR (what it stands for and what it is)
Relative Risk
Ratio of two proportions
True or False: Mean is best to use to determine spread and center when there are outliers or skewness. If false, say why and correct answer
False, Median and IQR are best
Name three random sampling techniques
Cluster sampling, stratified sampling, multistage sampling, SRS
Explain Parameter vs. Statistics
Parameter - numerical summary of population
Statistics - numerical summary of sample
What does DOFS stand for
Direction, Outliers, Form, and Strength
What do Contingency Tables do?
. Summarize Categorical data
. Display frequency
. Computer percentage to determine association
What’s the IQR
Q3-Q1
What’s the margin of error equation?
1/sqrt(n)
What’s extrapolation
Predicting outside the observed range
What are the three steps you much take when you encounter an outlier?
1. Verify data
2. Check if it belongs to the data set
3. If yes, do regression with and without the point. If they’re very different, collect more data
Simpson’s Paradox
Association can be reversed if there’s a third variable
What’s the variance equation
Summation (x - mean)2/(n-1)
What are the steps of Simple Random Sampling (SRS)
. List all elements in a population and number them
. Let the computer select at random
What’s undercoverage
Sample lacks part of the population