Day 2 Class Outline BUSN212 Wednesday 11/16/201: Week 1.
- Correlation analysis
(section 7 chapter 11. Problem set 1).
- Note key to problem set #1 will be posted to Moodle by Friday.
- Check Moodle for suggested problems from the book you can
try.
Class outline (probably won't cover all of this today!):
CORRELATION (Section 7 of chapter 11. Problem set #1.)
- Rho, the population
correlation coefficient.
- Measures the strength and the direction of the
LINEAR relationship between TWO variables.
- Given by the formula Covariance(X,Y)/standard
dev(X)*standard dev(Y)
- Always between -1 and +1.
- Estimated by the statistic r.
- Computational Formula for r.
- Finding and interpretting
r.
- Examples. Problems 1-3 of problem set #1.
(Example 1 first)
- Does NOT pick up nonlinear relationships (tire
pressure example)
- Sensitive to extreme values (outliers)
- Fall 2009 Principles of Macroeconomics.
- Do the results above suggest that how students at the very start of the
term perform on the first quiz (given Tuesday of week 3) is a good predictor
of their scores on the final exam? Why or why not?
- Correlation (1) calculation (2) interpretation.
- Examples. Problems 1-3 of problem set 1.
- NOTE: (1) Correlation does NOT pick up "nonlinear"
relationships (tire pressure example)
and (2) is
also sensitive to extreme values or what are called "outliers" (click here)
- Statistical inference. t tests
What from an
estimate of rho can we claim about the actual correlation between two variables.
Use t
tests to answer this question. (Assume that both x and y are normally distributed. Shapiro-Wilk goodness of
fit test.)
- READING ASSIGNMENT FOR FRIDAY: "Simple" regression
model. Sections 1-3 chapter 11.
.