Intro

·      there will be plenty of programming every lecture

·      homeworks will be turned, graded, returned on a weekly basis: they consist of programming and statistical analyses

·      there will be a written final. No programming will be required for that.

·      There will be some class notes on each topic that participants are encouraged to complete by themselves

·      Group work will be encouraged but is not required

 

contents

 

1.      intro R:  intro intro, intro R (1 wochen)

2.      discrete random variables, contingency tables

3.      intro stats: binomial, uniform, exponential, normal density, t-distribution, chi2, F, exponential (2 wochen)

4.      data analysis: box plots, quantile plots,  empirical distributions (ecdf), running histograms, nparm dichte schätzer (1-2  wochen)

5.      tests: p-values, 1,2 – sample-t-test, mean-median. Ranks, Wilcoxon 1,2-sample rank tests, Levene test, Kolmogorov-Smirnov, k-sample tests, anova, kruskal-wallis (3 wochen)

6.      linear Regression: simple linear regression, ols – lad regression, multiple linear regression, categorical regressor, anova (2 wochen)

7.      binary Regression: logit, probit regression, multinomial regression, ordered regression (2 wochen)

8.      Curve estimation: lin regression, shifted slopes, intercepts, regression splines, weighted lin regression,  local regression (2 wochen)

 

 

 

Programm:  last change: 06/07/14@19.00

Lec

Thema

Xtra

Ex

1-2

Intro2R: importieren (read.table, scan), xedit, Typen, Numeric, vector, string, factor, list,  Data selection,  tables, person-period, person-level

Refcard

rats fes smoke crime maps robjects

ex1

3-4

Discrete random variables:

2dim-table, independence, odds ratio, 3dim-table, Simpson’s paradox

 

ex2

sol2ex2

 

introStats: contUnif, normal, z-tests, p-values, confints,

t, chi2, F, exponential, binomial, poisson, discUnif, hyperGeo

 

ßprogramme im skript

ex3

 

 

 

introDatAn: ecdf, symplot, qplot,

normal plot, density estimation  

ecdf  quantiles  fesDens

runningHist

bwSelect kernels

ex4

 

5-7

Tests: binomial test, p-value,  proportion, 1-sample t-test, sign test, signed rank-sum test,  matched pairs, KS-test,

2-sample t-test, rank-sum test, KS-test, k-sample, multiple testing, Bonferroni, anova, kruskal-walllis

pValues

1Rank

MCWilcox

matchedPairs

Glivenk KS_1

twoIndepSampleInf

2Rank KS_2

k-IndepSampleTest

  Krusk

ex5

 

ex6

 

 

 

13-14

15-16

17-18

Regression(simple)

Lineare modelle: residuen, cook’s distance, Vorzeichen fehler predictions, gof-measures testingSelecting, categorical regressors, an(c)ova,

randomAnova

 diagnost  regPredict regRobust

car varTrans SAT

satData gofR testingSelecting

ancova anova1

studentData

statesData

randomAnova1

randomAnova2

faraway’s book

 

ex7

 

ex8

 

ex9

 

 

 

…one

 

 

week

 

off…

19-20

20-21

BinaryRegression Interpretation,

 logit, probit, odds, gof(Rsquared),

 classification ROC, deviances, LR-test,

ordRegIntro ordinalRegression

 

logitSimple probit

coefDet&LR ORC LRtest  binomReg Orings  OringsData

contingency  SimpsonParadox

MrozData

ordReg

 

ex10

 

 

ex11updated

 

 

22-24

CurvEstimation regression splines, smoothing splines, roughness-penalty

 

ßprogramme im skript

ex12updated

 

 

Review

 

 

 

 

 

klausUrSol1term

 

 

 

 

 

 

 

 

 

 

 

 

 

N.B. Für die durchFührung dieses programms übernehmen wir keine gewähr, d.h. abwandlungen, vertiefungen, kürzungen sind möglich

 

Literature:  lmgtfy.com

Kneip: statistikB

Kneip: http://www.uni-bonn.de/~kutikal/SS09/Nparm/Mainkurven.pdf