Scoring by Questions: Adding Feedback, Scoring State and Modal Enhancements

1 Initial Problem

Compared to the view 'Scoring by Participant', the view 'Scoring by Question' is missing a possibility to add manual feedback by the corrector. Furthermore the state of a corrected question for the displayed users could not be marked as 'finished'.

2 Conceptual Summary

To solve this problem some new user interface elements should be added to the table of the 'Scoring by Question' view. The corrector should be able to add the manual feedback for each user listed in the context of the chosen question. Furthermore he/she should be able to mark a record as 'finished'. If done, ILIAS should store and present the 'by whom/and when' information.

3 User Interface Modifications

3.1 List of Affected Views

Test > Manual Scoring > Scoring by Question

3.2 User Interface Details

New Filter Elements:

• New element to filter table items according to a particular state: Finalized Evaluation with options: All Users / Evaluated Users / Not Yet Evaluated User

• Feedback (Type: Textarea/TinyMCE, editable if evaluation is not finalized)
• Finalized Evaluation (Type: Editable Checkbox, optional column)
• Finalized by (Type: Text with the name of the corrector, optional column)
• Finalized on (Type: Text with a datetime presentation, optional column)

• Manual Feedback
• Finalized Evaluation

According to other table presentations (e.g. sticky threads in forum objects), rows with a given feedback (for the user/question tuple) should be hightlighted.

3.3 New User Interface Concepts

None

4 Technical Information

New database fields for table 'tst_manual_fb':

• finalized_evaluation (int, 1)
• finalized_by_usr_id (int, 4)
• finalized_tstamp (int, 4)

5 Contact

• Author of the Request: Vollbach, Guido [gvollbach] / Jansen, Michael [mjansen]
• Maintainer: Heyser, Björn [bheyser]
• Implementation of the feature is done by: Vollbach, Guido [gvollbach]

6 Funding

The feature has been already implemented for a customer GIT fork. The source code could be provided as a Pull Request on GitHub or we could directly commit/push this changes to the trunk.

7 Discussion

Heyser, Björn [bheyser], 22. February 2018: I want to point on some critical aspects:

• The filter "Finalize Evaluation" should be named more clearly
• The corresponding table column should be a static column
• Adding the information of the maximum grantable points to the table is quite senseful, but perhaps it can be combined with the column for the points input to save horizontal space (<textInput> of <maximumPoints>)
• The link show Answer should be changed as the modal does not show only the answer any more, but also a feedback input
• Having the TinyMCE (when enabled) in the table in every row might be a performance problem

Tödt, Alexandra [atoedt]2018-02-23:
Labels

• In ILIAS "Feedback" is what ILIAS provides automatically OR what peers provide. Tutors / teachers give Evaluations. This should be amended in both subtabs "Scoring by Question" and "Scoring by Participant". But very definately the same thing should not be called "Evaluation" and "Feedback" on the same table. 
• What is labelled as "Finalized  Evaluation" in the subtab "Scoring by Question" is labbeld "Marked as 'Scored Participant'" in the subtab "Scoring by Participant". 
• "Points" / "Maximum Points"  in "Scoring by Question" does not match "Points for the Answer" and "Maximum Points for Question" in the subtab "Scoring by Participant". 

Notification

• In the subtab "Scoring by Participant" a notification can be sent. It would be very desirable to offer this option here, too. 

Educational

• Scoring a solution that is not visible to the Scorer is a bad practice. The interface must offer the solution provided by the participant on top of the evaluation statement. 

Please consider using the Reporting Panel

Jobst, Christoph [cjobst], 26.2.2018:

Since this is a finished feature, asking for more might be inappropriate, but I'd like to point out some things worth considering.

1. The manual scoring, the evaluator and the comment should always be saved to the tests log.
2. Grading should at least have an option to be anonymous (without making the whole test anonymous).
3. From an higher education perspective the system still lacks the two examiner principle (Zwei-Prüfer-Prinzip), which is part (at least in germany) of every Higher Education Act (Hochschulgesetz). See this Feature-Request. I can see some synergy here. We are currently trying to get a quote to implement the linked feature as UI-Hook-Plugin. Maybe we can work something out to put this request into 5.4 too with some conceptual alterations? 
4. A filter for the correction pass (!= test pass) might be useful too.

JourFixe, ILIAS [jourfixe], 26 FEB 2018: We highly appreciate this feature request and schedule it for 5.4 with the following changes:

• Scoring should take place in modal. Therefore all necessary inputs should be available there.
• No TinyMCE (or multiline text input) in tables.
• Change label 'Finalized Evaluation' to 'Scoring completed' (and 'Completed at', 'Completed by').
• Do not show feedback and scoring to participant until 'Scoring completed' is set.
• No background colour in table. Identifying completed scoring easily is possible by filter.
• Save 'Completed at' and 'Completed by' only when scoring completed is set to checked for a participant.

JourFixe, ILIAS [jourfixe], 12 NOV 2018 : We schedule this feature now for ILIAS 6.0. Please update the feature request according to the last requirements of the Jour Fixe.

8 Implementation

Implemented as described above.

Test Cases

Test cases completed at 2019-05-03 by Jansen, Michael [mjansen]

• C25166: Bewertung nach Frage: Struktur der Datentabelle
• C13533: Manuelle Bewertung
• C25167: Bewertung nach Frage: Entfernen der „Endgültig bewertet”-Markierung
• C25168: Darstellung des Kommentars der manuellen Bewertung für den Lernenden
• C25169: Bewertung nach Frage: Filter „Endgültig bewertet”

Approval

Approved at 2019-05-29 by H. Markus, A. Seibel, A. Zahn / Uni Göttingen.