0
Views
“Data analysis using IBM SPSS Statistics” - course RUB 42,000. from MSU, training (2 months), date December 3, 2023.
Miscellaneous / / December 06, 2023
In a very condensed form, this course is part of the popular distance learning course “How to do scientific research: methodology, tools, methods” of the E. Foundation Open University. Gaidar (about 2 thousand. listeners per year). The Faculty of Economics of Moscow State University provides students with the opportunity to use an equipped computer class with SPSS installed study in detail the methods of working with data face-to-face with a teacher, work with the program with your own “hands” SPSS. It is possible to work not only with databases proposed by the teacher, but also with student data; the teacher will advise on what methods and how to use to analyze your data.
This course has been tested at the Open University of the E. Foundation. Gaidar.
Doctor of Economics, Professor of the Faculty of Economics of Moscow State University, specialist in quantitative research in social sphere, leader of more than 30 research projects, has experience teaching analytical courses at National Research University HSE, REU im. V.G. Plekhanov.
Email: [email protected]
1 The essence and main directions of sample surveys of the population. Possibilities of using special PPPs for processing sample survey data
Methods for collecting quantitative information. Sample studies. Sample socio-demographic surveys in Russia. Basic statistical software packages for social research. Functions of special software (Statistica, SPSS) in processing data from sample studies. Structure, SPSS modules. Areas of data processing. Data preparation. Entering and saving data. Measurement scales (quantitative, ordinal, nominal). Properties of scales and their permissible transformations. Types of data categorization.
2 Data preparation. Data selection and modification
Selection of observations. Sorting observations. Dividing observations into groups. Data modification. Calculation of new variables. Calculation of new variables according to certain conditions. Formulation of conditions. Data aggregation. Rank transformations. Case weights. Reasons and mechanisms for generating data gaps. Possibility of ignoring omissions. Methods for filling missing values. Methods for identifying anomalous values. Application of robust assessment procedures. Multiple Response Analysis
3 Descriptive statistics. Contingency tables
The role of statistics in processing the results of sample surveys. Micro and metadata. Areas of application and limits of applicability of mathematical and statistical methods. Summary of observations. Descriptive statistics. Univariate distributions. Variation indicators. Dispersion, variation range, mean absolute deviation, quantile ranges. Construction of contingency tables. Graphic representation of contingency tables.
4 Parametric and non-parametric tests
Analysis of the relationship between characteristics. Independence of variables. Basic characteristics of communication. Nonparametric and parametric tests. Independence test (goodness-of-fit test χ2). Comparison of two and several samples (dependent and independent). t-test. Statistical tests for contingency tables. Correlation coefficients (for nominal and ranking scales). Measures of the closeness of the relationship between variables. The simplest measures of connection closeness (for dichotomous variables). Relationship measures for tables with ordinal data. Kendal t-measures and their properties. Somers' d-measures. Goodman-Kruskal measure and its properties. Analysis of variance
5 Correlation and regression analysis
The essence and objectives of correlation analysis. Scatterplots. Paired correlation coefficients. Measuring the degree of closeness of a statistical relationship, “cleared” of the influence of extraneous characteristics using partial correlation coefficients. Checking the significance of the relationship between signs. Confidence intervals for correlation coefficients. Multiple correlation coefficient. Determination coefficient. Two-dimensional regression analysis model: linear and nonlinear regression models. Growth curves in forecasting problems, “dummy” variables and their applications. Multiple linear regression model. Nonlinear regression (binary logistic regression, multinomial logistic regression, ordinal regression, probit analysis, curve fitting).
6 Dimensionality reduction methods
Statistical approach in the principal component method. Calculation of principal components and their graphical interpretation. Information content of the reduced feature space. Principal components regression. The role and place of nonparametric methods in structural modeling. Hierarchical cluster analysis. Metrics of feature space. Principles of measuring distance between groups of objects. Algorithms for fast cluster analysis, k-means method. Two-stage cluster analysis. Building a goal tree
Subscribe
YOU MIGHT ALSO LIKE
