Model Comparison Approach: Model 2

The Model Comparison App is focused on linear modeling inferential statistics and describe the computation of linear estimators based on the data (i.e., the model estimation), the test of the fit of the model to the data (i.e., interpreting the parameters estimates and respective error) and the generalization of the the results to the population (i.e., performing the statistical inference). The app is developed for two different statistical models with different levels of complexity. Model 2 uses one categorical independent variable and a metric dependent variable and offers a more advanced discussion using linear regressions with dummies, t-test, and factorial ANOVA. Using these research problems we provide a description of the commonly used statistical techniques and the demonstration that all belong to the broad class of general linear analysis.

The app was developed with a learning design. Each page of the app builds on the previous page to ensure a gradual buildup of knowledge. This ensures that the user obtains the necessary understanding of simple concepts that allow greater comprehension of more complex information. The pages in the app are organized in a consistent manner and each page includes an introduction/explanation of what the user should expect from the page, a section with the main content including tables and graphs, this is followed by a short quiz regarding the content learned in the page and finally the user is presented with a summary of the key things that they should take away from this section. The app content is interactive and allows the user to choose what they want to visualize, which is accompanied by dynamic changes in the description of content depending on what the user is looking at. The last page of both apps includes a glossary of key concepts that were introduced alongside with further recommended reading.


TquanT is co-funded by the Erasmus+ Programme of the European Commission.

This server is powered by RStudio's Shiny Server Open Source.

© 2018 University of Lisbon

Legal Disclaimer | Privacy Information