Stat II Jargon
Central Limit Theorem, Type I and Type II error, P-value, Biased estimation, mean
squared error.
Ch. 5:
Bivariate data, Define error in regression,
Problem of common response, Problem of confounding,
What criterion is used in regression (Hint: least squares
minimizes error sum of squares)
Slope of a regression, Intercept of a regression,
What are the limits in which a simple correlation
coefficient must lie? ANS(-1 to 1)
What test should one use for the overall regression model?
What test for individual regression coefficients?
What is R-square and Adjusted R-square?
What is the command for regression of y on x1 and x2?
ANS: lm(y~x1+x2)
Give two R commands, which run above regression and print the main regression
results.
reg1=lm(y~x1+x2)
summary(reg1)
In the above command how to ensure that the regression line does go through the
origin (no intercept)?
ANS: lm(y~x1+x2-1)
What is the R command to compute the correlation coefficient between x1 and x2
and test it? (ANS: cor.test(x1,x2) )
What is the R command to test the null that the mean of y is zero?
(ANS: t.test(y) )
What is the R command for analysis of variance?
(ANS: anova(lm(y~X1+x2)) )
What is the R package needed to do basic descriptive statistics conveniently?
(ANS: fBasics)
What are the R commands for computing the mean and standard deviation of y?
mean(y)
sd(y)
What are the R commands for:
1) Reading data ANS read.table(file="filename")
2) Running regressions including multiple regressions ANS: lm(y~X1+x2)
3) Descriptive stats (which package do you use and how?) basicStats(y)
5) Plots and plot headings plot(x,y, main="Heading")
6) Confidence intervals ANS: confint(lm(y~X1+x2))
7) Analysis of variance. ANS: anova(lm(y~X1+x2))
8) Finding the residuals of a regression ANS: resid(lm(y~X1+x2))
9) Printing the results of regression to the screen.ANS: summary(lm(y~X1+x2))
10) Computing the portion of test stats which comes from z, t and F tables. qnorm,
qt, qf
11) Computing the portion of p-value statistics which comes from z, t and F
tables.pnorm, pt, fp
#Finding CRITICAL VALUES from t table and F table using qt and qf
alp=0.05 #the alpha value
alpby2=alp/2 #defines alpha by 2 for typical t test
dft=length(y)-3 #defines degrees of freedom for t
qt(alpby2,df=dft)#quantile of t distribution
qf(alp, df1=2, df2=dft, lower.tail=F)#quantile of F distribution
#Finding p-values from t table and F table using pt and pf
#pvalues are already given by R
#for example t value in summary table is -2.306 for the pie-sales example
pt(-2.306,df=dft, lower.tail=T)*2
# we multiply tail area by 2 to get the p value= 0.03976325
#doubling is needed because t test is two-sided
Part 2 of the final will have 3 questions
1) Use excel to do anova worth 2%
2) Given a payoff table, find maximax, maximin, minimax regret, max expected value and minimum expected regret worth 3%
3) 6 questions from jargon worth 3%
see
http://www.fordham.edu/economics/vinod/jargon2.txt
file at my website for possible list for the third question.