Developer Testing

Developer Tests and Outcomes

Pass/fail has already been implemented and delivered at course level.  It has also already been delivered at programme level for undergraduate HSS programmes.  The scope of the work under test is to implement pass/fail at programme level for postgraduate taught HSS programmes.  Tests were carried out using the programmes PTMSCCOMGL1F and PTMSCISLAM1F. All results for courses CLLC11003(2012-3_SS1_SEM1) and IMES11037(2012-3_SV1_SEM2) display as pass/fail results.  As these courses all carry credits the result calculated by SMART is "unknown - credit-bearing pass/fail".  This is by design and has been explained to users by the business analyst during the delivery of previous pass/fail functionality. Credit profiles work as per other PG sheets with the additional feature that a pass on a pass/fail course should always be included in the "passed >=50%" column.  Spot-checks on students 1239037 and 1226790 (1st programme), 0910651 and 1147542 (2nd programme) confirmed that the credit profile calcs were correct for the given course enrolments and results.  This was checked on both the dissertation and graduation sheets for each programme. In terms of results, only student 1160756 on the 2nd programme was able to have an interim recommendation calculated due to them not having taken a credit-bearing pass/fail course.  The result was correctly evaluated as "fail" given the enrolment results.  The corresponding graduation sheet correctly flagged this student as having exited at progression stage and would therefore not be eligible for classification. In order to check other results in DEV the pass/fail courses needed to be manually reconfigured to carry no credits.  On the progression sheets all recommendations were checked against the reported percentage and credit profile and correctly recommended all of the available options: "proceed to dissertation", "proceed plus eligible for distinction", "exit with PG diploma", "fail".  As the 1st programme was the only one with enough students to cover all of the recommendation options, only this programme is used for subsequent tests.  On the classification sheet all students who exited at progression stage were correctly identified as such.  All classification recommendations were correct based on the enrolment results and the previous progression recommendation.  All recommendations are calculated according to pre-existing SCE PG business rules which have been confirmed as correct for HSS PG by the business analyst. Manual overriding of recommendations was tested using 1226790.  Overriding, and reverting back to the calculated value, worked correctly on both the dissertation and the graduation sheets.

PDF versions of the spreadsheets for both the dissertation and the graduation sheets look consistent with the onscreen versions.  On the PDF reports overridden recommendations are clearly flagged by the text "(AMENDED)" next to the recommendation.  CSV versions of the spreadsheets for both the dissertation and the graduation sheets look consistent with the onscreen versions.  On the CSV reports overridden recommendations are also flagged by text "(AMENDED)" next to the recommendation.  As with other HSS delivered reports, enhancements have been made such that the PDFs and CSVs will load immediately instead of having to go through the request and alert windows. The final check is that the classification outcomes for these HSS PG programmes is picked up correctly by the "PG classification return to registry" reports.  The onscreen reports for both programmes were checked against the return to registry reports.  All classifications appear to be represented correctly in the return to registry sheet.

Also included in this work is a fix to part of a production issue logged in I130429-0344.  If a course-level result in SMART is "unknown" then all of the % columns should be empty.  This was tested using the T2_gcarter course on Dev for student 1000012.  The PDFs amd CSVs also reflect the change.

Testing of TEL003-56

As documented in the JIRA issue, this was tested on both a UG and PG programme.  The calculated averages were checked before and after forcing a course result to be "Abs".  The test was based on checking that the post-change average was based on SMART evaluating "Abs" as zero instead of -1.

Regression testing

This phase 3 work - Retested the above developer tests using programme PTMSCCOMGL1F.

These jiras were also retested, as per note on testing contained in the comment logs for each jira:

• TEL003-30 - Retested using T2_gcarter and student 1000012.
• TEL003-31 - Retested using T2_gcarter.
• TEL003-32 - Retested using T2_gcarter.
• TEL003-33 - Retested using BILG08008 and T2_gcarter.
• TEL003-35 - Retested using T2_gcarter.
• TEL003-36 - Retested using T2_gcarter.
• TEL003-52 - Retested using T2_gcarter.

Confirmation of Standards Adherence