Learning objectives
The course aims to provide students with the theoretical and applicative tools to independently understand and develop the most important statistical techniques that make up the applications of the General Linear Model (GLM) and its extension (Generalized Linear Model) most frequently applied in psychobiological research and of cognitive neuroscience. Guidelines will also be provided for the correct drafting of the statistical techniques used in the production of texts that illustrate research results (thesis, conference communication, scientific articles), according to the APA recommendations. In detail, the training objectives are:
1. Knowledge and understanding. Students will need to achieve a good understanding of inferential statistics and understand how it is used in basic and applied research.
2. Ability to apply knowledge and understanding. Students should be able to use the R environment to develop and interpret inferential analyzes on simple and complex linear models.
3. Autonomy of judgment. Students will have to develop critical skills and independent judgment with respect to the production and description of data in texts that illustrate research results.
4. Communication skills. Students must be able to communicate the results of inferential analysis of data, both in numerical and graphical form, placing them in the context of research hypotheses.
5. Learning skills. Students will have to be able to learn new techniques for conducting inferential analyzes, in particular in the R environment, and their report in scientific texts.
Prerequisites
Students must had succefully passed Tecniche di Analisi di Dati I, before to take the exam of Tecniche di Analisi di Dati II.
Course unit content
Statistical inference: NHST versus model fitting. General Linear Model: features and assumptions. Models of relations among quantitative variables: zero and first order correlations, simple and multiple linear regressions, mixed models. Models of relations among cathegorical and continuous variables: ANOVAs, ANCOVA. Models of relations among cathegorical variables (Generalized Linear Model): logistic regressions, multilevel inear models.
Full programme
Statistical inference: NHST versus model fitting. General Linear Model: features and assumptions. Models of relations among quantitative variables: zero and first order correlations, linear regression, mixed models. Models of relations among cathegorical and continuous variables: ANOVAs. Models of relations among cathegorical variables (Generalized Linear Model): Poisson and logistic regressions, non parametric tests.
Bibliography
263/5000
All training materials are made available on the Elly page of the course: https://elly2020.medicina.unipr.it/
The student can download the course notes, slides, video recordings, exercises.
Furthermore, the following texts are recommended:
Micciolo, R., Espa, G., Canal, L. (2013). Ricerca con R – metodi di inferenza statistica. Apogeo edizioni (capp. 1, 2, 5).
Gallucci, M., Leone, L. , Berlingeri, M. (2017). Modelli statistici per le scienze sociali ( 2 Edizione). Pearson. (pp. 19-246; 281-288).
Task Force on Statistical Inference – American Psychological Association (1999). Follow up report: Statistical methods in psychology journals. (pp. 1-11). http://www.apa.org/science/leadership/bsa/statistical/tfsi-followup-report.pdf
Teaching methods
The course will be held through lectures and group exercises (statistical package R) to Students either in the classroom (“in presenza”) or in synchronous-streaming (“in telepresenza”) on the Teams platform. Therefore, the opportunity of Student/Teacher interaction will be preserved both face to face and remotely, by the simultaneous use of the Teams platform.
Lectures will be supported by slide presentations, which will be available to students on the Elly platform (https://elly.medicina.unipr.it).
Assessment methods and criteria
During the lessons there are two intermediate moments of verification, in which attending and non-attending students can freely participate. the tests will concern the theoretical areas addressed in class, and are structured in two open theory questions and a practical exercise, on environment R, divided into three parts: one relating to descriptive statistics, the other two relating to inferential hypothesis tests. The exercise involves the analysis of data that will be made available (through Elly) no later than 48 hours before the test. The analyzes will include a first part relating to descriptive statistics on the data and two subsequent parts in which inferential hypothesis tests will be required. In addition to the correct execution of the statistics, the ability to adequately interpret and comment on the outputs of the analyzes will be considered an essential part for the purpose of sufficiency. The evaluation out of thirty will be as follows:
first theory question: 0-8 points; second theory question: 0-8 points
exercise: 0-14 points, divided as follows:
first part 0-4 points;
second part 0-5 points;
third part 0-5 points. Only for those who have completed all the on-going checks, the average evaluation obtained is considered valid for the final outcome. For those who have not completed the ongoing tests or who have not reported a fully sufficient assessment, the final assessment of learning will consist of a written exam completely similar to the intermediate tests, except that the required contents will cover the entire program: two open theory questions on the entire program, and an exercise in the R environment. The exercise involves the analysis of data that will be made available (through Elly) no later than 48 hours before the test. The analyzes will include a first part relating to descriptive statistics on the data and two subsequent parts in which inferential hypothesis tests will be required. In addition to the correct execution of the statistics, the ability to adequately interpret and comment on the outputs of the analyzes will be considered an essential part for the purpose of sufficiency. The evaluation out of thirty will be as follows:
first theory question: 0-8 points; second theory question: 0-8 points
exercise: 0-14 points, divided as follows:
first part 0-4 points;
second part 0-5 points;
third part 0-5 points.
If due to the persistence of the health emergency it is necessary to adopt the remote mode for the exams, we will proceed with the written test conducted at a distance (via Teams and Elly).
The student has the right to ask that the exam be supplemented by an oral test, provided that the outcome of the written test is overall sufficient. The oral exam is structured in a similar way to the written exam: questions related to the contents of the entire program and a short exercise in R environment on data used for the exercises during the course. If it is necessary to adopt the remote modality, the optional oral exam will also be done remotely, using the Teams and Elly platforms.
Students with SLD / BSE must first contact Le Eli-che: support for students with disabilities, D.S.A., B.E.S. (https://sea.unipr.it/it/servizi/le-eli-che-supporto-studenti-con-disabilita-dsa-bes)
Other information
The execution of the proposed exercises is strongly recommended; it is recommended to contact the teacher to verify their correctness.
