Copyright (c) 2021 Institut fuer Lern-Innovation, Friedrich-Alexander-Universitaet Erlangen-Nuernberg, GPLv3, see LICENSE
- Authors: Fred Neumann fred.neumann@ili.fau.de, Jesus Copado jesus.copado@ili.fau.de
- Forum: http://www.ilias.de/docu/goto_docu_frm_3474_4315.html
- Bug Reports: http://www.ilias.de/mantis (Choose project "ILIAS plugins" and filter by category "ExtendedTestStatistics")
When you download the Plugin as ZIP file from GitHub, please rename the extracted directory to ExtendedTestStatistics (remove the branch suffix, e.g. -master).
- Copy the ExtendedTestStatistics directory to your ILIAS installation at the followin path
(create subdirectories, if neccessary):
Customizing/global/plugins/Services/UIComponent/UserInterfaceHook - Execute
composer duin the ILIAS main directory - Go to Administration > Plugins
- Choose action "Install" for the ExtendedTestStatistics plugin
- Choose action "Activate" for the ExtendedTestStatistics plugin
- Choose action "Configure"
- Select which evaluations should be presented to all users with access to the test statistics, which should only be shown to platform admins and which should be hidden at all.
- Set the calculation parameters and quality thresholds for single evaluations.
This plugin replaces the sub tab "Aggregated Test Results" of the "Statistics" tab in a test with two new sub tabs "Aggregated Test Results" and "Aggregated Question Results"
Both tabs show tables with statistical figures related to the test or its question. The values that are calculated by standard ILIAS are always shown. Other values are shown based on the setting in the plugin configuration.
Additional test evaluations:
- Coefficient of Internal Consistency (Cronbach's alpha)
- Mean Score
- Median Score
- Standard Deviation
Additional question evaluations:
- Discrimination index
- Facility Index
- Standard deviation
- List of chosen options for single and multiple choice
- Percentage of Correct Answers (shown as diagram on the question overview)
The evaluation titles have tooltips with further explanations. Some values have read/yellow/green signs to indicate their quality. Values displayed in italics are considered as 'uncertain' because they are calculated in a random test. Nevertheless they may provide useful hints. Values may also have comments appearing as tooltips.
The toolbar above the tables allows to select of they are calculated for the scored pass (standard), the best pass or the last pass of each participant.
The aggregated results can be exported to CSV or MS Excel files. A CSV file contains only one table with either test or questions overview and includes no comments or quality signs. An MS Excel file contains both tables on separate sheets with an extra sheet for the legend. It includes all comments and color bachgrounds for the differend quality signs. An MS Excel file with details contains an extra sheet for any detailed evaluation.
Since version 1.1.0 the generation of a diagram is supported. An evaluation has to provide a details table that is used to generate the diagram. One column of this table is used for the labels and one or some other columns are used as the data series of the diagram. The diagram is shown to the details screen if the evaluations supports this.
-
The extended statistics are currently not calculated for tests with setting 'Question Queue - All Questions from a Question Pool'. This, however, is also the case for the standard statistics of ILIAS.
-
The calculation of extended evaluations takes care of performance, e.g. the additional question evaluations are calculated on a separate page only for the questions that are shown. But the basic values from ILIAS are calculated for all, which takes time for large tests. This is now improved by caching.
You may extend the plugin with your own evaluations. See the following add-on as an example: https://github.com/ilifau/ExampleEvaluations
See CHANGELOG