Statistical analysis of data
This training will give participants the tools they need to conduct a quantitative analysis of survey data. It is an introduction to descriptive statistical methods and bi-variate statistical tests. The techniques and appropriate situations for the application of these techniques will be presented and illustrated with practical examples. Exercise sessions will be used to perform the analyses seen in class and to interpret the results obtained using R and SAS software. At the end of this training, participants will know when and how to use basic statistical techniques to perform a proper analysis of their own data.
This training allows participants to understand data modeling and processing methods in order to extract useful information to validate a hypothesis or to assist in decision making in a company. The objective of the training is to understand how to "extract the phenomena, laws, and knowledge contained in the data that we cannot directly apprehend".The training allows participants to:
- Synthesize the information contained in data.
- Understand the theoretical and practical means to exploit information from multidimensional statistical databases using multivariate statistical analysis methods.
- Master techniques for describing, reducing, classifying and clarifying data, identifying relationships, similarities or differences between variables or groups of variables.
- Choose the appropriate method for the construction of a typology, according to the nature of their data, and interpret the results using several statistical software programs.
- The Data Analysis training course explains traditional data analysis methods (Principal Component Analysis (PCA), Correspondence Factor Analysis (CFA), Multiple Correspondence Factor Analysis (MCA), Discriminant Analysis (DA), and Hierarchical Classification Methods (HCA).
The main elements of this training include data processing and ways of presenting results for decision-making. The content of the training will include:
- Descriptive statistics with categorical and continuous variables; the relationship between a continuous variable and a categorical variable
- Hypothesis testing: t-test for independent samples or paired samples, analysis of variance with one factor: one-way ANOVA
- Relationship between two variables: cross-tabulations, Chi-square test and measures of association, McNemar’s test, test of hypotheses on proportions, correlations.
- Special topics: weighting, sample size, Bonferroni adjustment and writing the results.
- Multiple linear regressions: presentation of models, estimation and interpretation of coefficients, centered and reduced (standardized) variables, dichotomous and categorical independent variables.
- Model validation: residual analysis, variable selection, multicollinearity
This training is intended for all analysts or other professionals seeking to acquire the necessary skills that will help conduct or understand the analysis of your company’s data.
- Level : Beginner
- Duration : Three days