Statistical Significance
Variables & Functional Forms
Rules-of-thumb
Classical Assumptions
Random Questions (Ha!)
100

This test is used for multiple hypotheses

What is an F-test?

100

This variable can take on a value of 0 or 1

What is a dummy (or indicator) variable?

100

A VIF exceeding this level is usually taken as a sign that multicollinearity is severe

What is 5?

100

The error term is assumed to have this distribution

What is the normal distribution?

100

This is the first step when doing empirical research

What is review the literature?

200
The p-value below which a variable is considered to be statistically significant

What is P < 0.05 (so 5%)

200

This is the default function form

What is linear?

200

The t-stat "rule-of-thumb" for determining if a variable is statistically significant

What is a t-statistic > 2?

200

This term is used to describe an error term that does not have a constant variance

What is heteroskedasticity?

200

This Stata option will allow you to control for heteroskedasticity

What is "robust"? (so getting robust standard errors)

300

The appropriate hypothesis test if you are unsure about a variable's sign

What is a two-sided test?

300

The estimates from this functional form are interpreted as elasticities

What is double-log?

300

A correlation coefficient exceeding this level is an indication of multicollinearity

What is 0.80?

300

This term is used to describe error terms that are correlated with each other

What is serial correlation?

300

This variable is (usually) the product of a continuous variable and a dummy variable

What is an interaction term?

400

Serial correlation will have this effect on t-stats

What is increase?

400

These are the four specification criteria

What are theory, t-test, R(bar)squared, and bias?

400

A Durbin-Watson value that is close to this value is a sign of positive serial correlation

What is 0?
400

This occurs when all dummy variable categories are included in a regression equation

What is perfect multicollinearity?

400
This condition must be met to estimate an equation in a simultaneous system

What is the order condition?

500

This error occurs when a true null hypothesis is rejected

What is a Type I error?

500

Estimates from this functional form, when multiplied by 100, can be interpreted as growth rates

What is a semilog (lnY)?

500

Multiplying a logit coefficient by this amount gives an estimate of the marginal effect

What is 0.25?

500

This term is used to describe the OLS estimator when Classical Assumptions I through VI are met

What is BLUE (best linear unbiased estimator)?

500

This type of data set combines time series and cross sectional data

What is a panel data set?

M
e
n
u