Statistical Methods in Economy
Section outline
-
The material in this Moodle course is part of the course Statistical Methods in Economy, teached by the Faculty of Economics of the University of South Bohemia in České Budějovice, the Czech Republic. It introduces the reader to several statistical topics appearing in quantitative economical analysis. It is accompanied with computational examples based on the free statistical software R and RStudio.
This learning aid was created as a part of the project Rozvoj JU - ESF, reg. number CZ.02.2.69/0.0/0.0/16_015/0002348, financed by the European subsidy program Operační program výzkum, vývoj a vzdělávání období 2014-2020.
Tento výstup lze užít v souladu s licenčními podmínkami Creative Commons BY-SA 4.0 International
(http://creativecommons.org/licenses/by-sa/4.0/legalcode).
-
First steps with R programming environment: data types, direct data editation, data import, indexing. Basic statistical procedures and methods, graphs.
-
Descriptive statistics. Distributions: cumulative distribution function, quantile function, pseudo-random generation. Hypotheses testing: t-tests (one-sample, paired, two-sample/Welsh), F-test to compare variances, one-way ANOVA. Graphical visualization.
-
-
Exercise and Supplemental Material
-
-
Tests of normality (Shapiro-Wilk), QQ-plots. Non-parametric variants of t-tests, sign test, Kolmogorov–Smirnov test.
-
-
-
-
Standard non-parametric statistical tests: Mann–Whitney (Wilcoxon) rank sum test, Wilcoxon signed rank test, sign test, Kolmogorov test, Kruskal-Wallis test, Friedmann test.
-
-
-
-
Modus, median, percentiles, sample counts + measures of variability (variational ratio, range, variance), entropy, třídění. Graphical diagnostics: barplots, piecharts, histograms, empirical distribution function, Box-Whisker plots, contingency table, lattice graph.
-
-
-
-
Introduction to multivariate data analysis. Notion of data similarity (coefficient of association, correlation, metrics: Euclidian simple and weighted, Mahalanobis, city-block, Minkowski).
-
-
-
-
Cluster analysis, dendogram. Single and complete linkage, pair-group average linkage, Ward–Wishart's error sum of squares method.
-
-
-
-
k-means method, partition around method.
-
Probabilistic regression models (models with binary response). Logistic regression, logit transformation, maximum likelihood method.