Stat 601, Fall 2000, Class 5

What you need to know from last time

Hypothesis tests on means

One sample t-test; testing a single population mean, p.136

Two sample t-test; assuming equal variances, p.138

Two sample t-test; NOT assuming equal variances, p.146

Paired t-test; p.161,166

Assumptions

What makes the Paired t-test so useful?

How to use p-values to make the decision

The relationship between hypothesis tests and confidence intervals

Today's class - putting ideas together

Dealing with confounding variables

Review example - introducing correlation

Dealing with confounding variables

Marginal association - the relationship between two variables ignoring other possible explanatory variables

Partial association - the relationship between two variables having taken into account other explanatory variables

Association does NOT imply causation

Example Salary.jmp

Review example - FinMark.jmp

Points to note

Log transforms to linearize exponential growth

Returns remove most time trend, reveal volatility

Relative volatility between T-bills and VW-return

Good normally properties but a bit fat tailed

Return/variance tradeoff in portfolios