Canvas learning mastery export to Nuventive TracDat input

As the Nuventive team has noted, Nuventive Improve does not retrieve data from the Instructure Canvas learning management system directly. While this was deeply disappointing to me personally as this means double data entry will have to done by faculty to get data into TracDat, my knowledge of Canvas and spreadsheets provided me with a path to reduce the amount of manual counting on my fingers work my courses would involve.

Note that this process works as shown below if a faculty member is using course learning outcomes from the institutional bank of course learning outcomes in Canvas as outlined in a video on how to import course learning outcomes for use in rubrics and question banks.

Start from the Gradebook and use the drop down menu to switch to Learning Mastery.

The values seen in the columns are the outcomes results, on a five point scale calculating using a decaying average.

In a rubric outcome achievement is deemed to be optimal, sufficient, suboptimal, or no evidence that the outcome was achieved. An explanation for the adoption of this scale was given in March 2021. The next step will be to export the learning mastery data for the desired section of the course. 

On the right side in the Learning Mastery screen select the section for which you want to export data. Once the desired section is selected, tap on Export report to download a comma separated values spreadsheet. 

Open the comma separated values spreadsheet. Check for the existence of the Test Student and delete that row. There may or may not be a Test Student present in the section and the sort order considers "Student" to be a last name and alphabetizes this entry into the rest of the students with a last name starting with "S". In other words, Test Student is no longer the last entry.

Then insert a column as seen in column B above and manually enter the gender. One could conceivably copy this from the MyShark extended export of the class list, but I find manual entry here just as fast. Column B and the outcome result columns D and F seen above will be copied to another spreadsheet for purposes of generating the input for TracDat.

Click to generate a copy this spreadsheet for your own use

In the above spreadsheet the gender column has been copied to B13. The results from column D in the export report has been copied to C13. The results for a second and third course learning outcome have been copied into D and E above. Note that row 7 is a "recommended" results statement for input to TracDat at the institution. Some modifications were made for source clarity in downstream reports. 

The formulas driving the TracDat input spreadsheet can be seen above. The choice to count outcome achievement as being greater than a three is based on the logic that the result values are averages. An average above a three means that the student must have obtained some ratings of four, which is deemed a sufficient level of mastery. Obviously this value could be changed to set the success rate higher. 

The text blocks in row seven of the spreadsheet can then be copied into the appropriate field in Nuventive Improve and the values therein manually propagated down into the appropriate fields below. This is still a manual, after action, post-term activity and does not provide real time data during the term, but then Nuventive is not designed nor intended to function that way. 

Closing thoughts

While Canvas can used with business intelligence software such as Datastudio to produce near real time data on learning during the term without requiring double data entry by faculty, these dashboards do not handle qualitative data nor are they well suited to storing data across multiple terms.

Dashboard

I now see the two systems as complementary and necessary. The use of institutionally banked course learning outcomes in Canvas and dashboards in Datastudio provide during the term data to identify and address areas of weakness in learning while Nuventive Improve can provide longer term storage, aggregation, and reporting on learning across multiple terms. Nuventive Improve provides a history of assessment, while the Datastudio dashboards provides during term assessment data. I wish there was a way around the double data entry required of faculty, but I cannot see a way around this at this time. The above process of exporting learning mastery data and using spreadsheet functions to assemble the TracDat input would not be necessary in a course with only a few students. In larger courses, the system could reduce data entry errors.