The course provides principles of design and decision-making to select an appropriate methodology for data collection in regard to specific research questions. Then, it provides the principles of data processing, hypothesis testing, and statistical modeling. The course contains topics: basic statistical parameters and variability in a population, formulation of research and statistical hypotheses, parametric and non-parametric methods of hypothesis testing: tests of goodness of fit, contingency tables, comparison of two and more samples (t-test, ANOVA, Man-Whitney test, Kruskal-Wallis ANOVA, etc.), regression, correlation, general linear and non-linear models, multivariate analyses (ordinal and classification methods). A special focus is paid to practical training in selected statistical software, data interpretation and presentation of results. Seminars (training tutorials) are focused on two aspects: 1) training of own calculations of basic statistical parameters and decision making for selection of correct statistical tests for data analyses; 2) training of application of statistical tests in available statistical software, either STATISTICA software or R software, or other available ones. Another focus is paid to the interpretation and presentation of results obtained by statistical analyses. Seminars will be designed basically as a single-student project to tackle a series of tests in the form of a comprehensive model study embracing all aspects of research work.
Lectures:
- Formulation of research objectives, designing studies, types of data, basic statistical indicators and variability measures;
- Probability, data distribution, and data transformations, hypotheses testing, categorical data: goodness-of-fit tests
- Parametric and Nonparametric methods: two-sample comparisons, analysis of variance, assumptions and use
- Regression methods (simple-, multiple-, non-linear), correlation and general linear models (GLMs)
- Multivariate methods (ordination, classification)
- Ecological modelling: spatial data and predictive models
Seminars:
- Introduction to student project
- Tutorial: distribution testing, transformation of data & goodness-of-fit tests
- Tutorial: parametric tests
- Tutorial: non-parametric tests
- Tutorial: general linear and non-linear models,
- Tutorial: multivariate analyses
Students will be required to prepare (self-study) for their theoretical lectures utilising the literature sources, and will be provided with practice exercises through which they will need to work in preparation for the tutorials. Recorded lectures and demonstrations of statistical tests utilising various statistical software will also be utilised. Doctoral students will provide support as tutors, assisting in the demonstration, application, and practise of statistical exercises.
