What does “Total Activity” mean?

[bwalters]

How is the “Total Activity” time that’s shown on the people page in Canvas calculated.  Is it based on total interaction time on pages?  How does “hang time” between interactions get taken into account?  If at all.

[Stef_retired · Accepted solution]

 @bwalters , this is explained in How do I use the People page as an instructor? . Under the View Users section, you'll see the legend for #7, Total Activity:

• Total Activity [7]—allows you to see how long students interact within a course and is associated with page views. Total activity time is displayed in hours:minutes:seconds. If a user has not yet reached an hour of activity, total activity time is displayed as minutes:seconds. Total activity requires a two-minute page view minimum for performance. Total Activity does not capture page views for videos that do not include intermediate page requests, such as a half-hour recorded lecture.

[bwalters]

stefaniesanders​ Ah! I didn't think to look there... 😕 Thank you.

[bwalters]

Stefanie, I have one item that I wonder if you can clarify...

I’m still not sure what the hang time means, for example, if somebody is on a page for 20 minutes and then moves onto another page does it count the whole 20 minutes as activity time?  What’s the point where this cut is off?

This bit in that sentence also has me stumped “Total activity requires a two-minute page view minimum for performance” - what does this mean?

[Stef_retired]

 @bwalters ​, what I think "Total activity requires a two-minute page view minimum for performance" means is that the Total Activity column won't populate until the user has been in Canvas for at least two minutes. And I'm not clear on the significance of "hang time" in relation to these stats. If a student is on a page for 20 minutes, as far as I know, 20 minutes of activity is recorded in Total Activity; after all, how is Canvas to know if the student is reading the page, or has walked away to get a snack, or is in fact still sitting at the computer while simultaneously texting on her phone?

Personally, I think course analytics gives you a better handle on student engagement: How do I view Course Analytics?

Even so, I'm not a guru on the subject of Total Activity, so I've looped in someone who is a guru to get more information, confirmation, and/or refutation of what I've just surmised.

[erinhmcmillan]

The short answer is there is no good way to analyze student activity in a computer. In our upcoming release we'll be updating our People page lesson to clarify the two minute statement as:

• Total Activity records any time spent viewing course content that exceeds two minutes. If the time between a new activity and the last completed activity is under ten minutes, all time between these two events will also be included.

You can also use How do I view the course access report for an individual student? to view specifics on course participation.

Thanks,

Erin

[bwalters]

erinhallmark​ thank you for the reply. That makes lots os sense. I think that is what we are looking for.

[richard-jerz]

Erin, I am a little new to Canvas so I apologize if my question seems elementary. I am trying to understand the "equation" used to calculate Total Activity, so I will ask some questions about your definition.

"Total Activity records any time spent viewing course content that exceeds two minutes."  Computers do not know what a person is viewing, they only know that the person clicked on a webpage and have not clicked on something on this webpage that removes it.  Of course, if someone clicks on a web page and goes to the bathroom, they are not viewing anything.  So your word "viewing" needs to be defined.  So are you saying that "viewing" is defined to be that a student has clicked on a webpage, the webpage remains active for at least two minutes without the student clicking on something that that puts it away in less than two minutes?

If a student clicks on a webpage and leaves it active on their computer, this webpage will be considered "viewed" for more than two minutes.  Right?  Or must they click on something else within 10 minutes for the first webpage to be viewed?  Also, if the student clicks on a webpage in their dorm room, leaves it active, then goes to a computer lab and opens Canvas, might "double counting" occur?

Also, you say "the last completed activity."  What is the definition of "completed activity."

Is a detailed definition (and equation) of Total Activity somewhere in the Canvas docs?  I sometimes don't like to use the term "equation" because it sounds scientific, but computers are programmed, and there has to be an equation somewhere.  I just would like to know what it is.

Thanks.

[InstContentTeam]

 @richard-jerz ​,

Thanks for your comment. Let me see if I can help answer your questions.

1. You are certainly correct that Canvas doesn't monitor if a user is physically viewing a page. The time in Total Activity can only count when a page is viewable on a browser.

2. A second page does not need to be viewed in order for the first activity to count. Once a page has been viewed for two minutes, that time will be added to the Total Activity.

3. When the definition points out "last completed activity", it refers to the last time a page was viewed for two minutes or longer. For example, if a student views a page in Course #1 for longer than two minutes, navigates to Course #2, returns to the Course #1 in under ten minutes from the time they left, and views a page on Course #1 for longer than two minutes, the time inbetween the two activities will be added to the Total Activity.

4. According to our Canvas support team, Total Activity should not count twice if a user has logged into Canvas on computers in two locations.

5. The definition for Total Activity can be found in How do I use the People page as an instructor?​, which is the same definition Erin shared above. I will certainly take a look at this and see if there is anything we can update to clarify this.

Cody

[richard-jerz]

Canvas states "Total Activity records any time spent viewing course content that exceeds two minutes. If the time between a new activity and the last completed activity is under ten minutes, all time between these two events will also be included."

I have tried to verify this, and have not had success.  For example, a "refresh" of a browser window is considered "activity".  Oh, really!  I have also logged out, logged back in, an the clock appears to keep running continuously.  Hmm, I guess Canvas considers this as activity, logging out and then logging back in.

Right now, I tried this little experiment:

Logged into Canvas, went to course 1, noted my activity time in Course 1.

Immediately switched to Course 2, noted my activity time in Course 2.

Stayed in Course 2 doing absolutely nothing.  Then, 5 minutes later refreshed my browser.

Course 2 reported 5 more minutes of activity.  Well, doing nothing in Canvas is not activity.  Far from it.  I guess refreshing the browser by Canvas is considered activity... this is all I did.

Then, I immediately switched back to course 1.  Guess what, it had added 15 minutes to course 1 activity time, even though I only spent 5 minutes doing this experiment and did nothing in Course 1.  For 5 minutes I was staring at Course 2, not Course 1.  Even worse is that Canvas thinks I spent 20 minutes in Canvas, if I were to add up times in each course.

The point is that this continues to be a bogus calculation and cannot be accurately repeated or verified.  As such, it is very misleading to those who think that they can track student's "activity" using Canvas.  This is why I suggest that it be removed.

My conclusion is that there must be people who like seeing numbers, regardless of how meaningless or incorrect, than seeing the reality that "activity" in Canvas cannot be tracked.

[Renee_Carney]

Hi  @richard-jerz ​

I was just reading through the conversation here, and it looks like it has been a valuable one.  Thank you for brining your questions, and, Doc team, thank you for providing clear answers!  Rick, I'm sorry you didn't get the answers you wanted, but our docs team provided the best information they have based on how the feature works today. 

If you look through the Find Answers​ and the Canvas Feature Ideas​ spaces of the community, you will see that this idea and conversation is not new - it comes up frequently.  In-fact, I was the original poster of a related idea in our very first community.  You can find it at " modifiedtitle="true" title="Student Activity Report: Student Time Log.   (please add your vote and use cases there).

We all know that it is virtually impossible to track time on task in the virtual environment.  In my opinion, that is where instructional design strategies come into play; utilizing strategies that require engagement for advancement in the course.  Unfortunately, as I stated in the original idea, there are still outside constraints that require time tracking.  In that case, we do offer Canvas Data.​  You might want to explore that feature to draw out raw data that you are seeking.

[richard-jerz]

Renee, thanks for your post.

I feel sorry that there are administrators who lack a basic understanding of what LMSs can and cannot track, and who think Canvas can track students' activity time.  The fact is that no LMS (not picking on Canvas) can track a student's activity.

I also feel sorry that Canvas is supporting this incorrect idea and is providing this "Total Activity" statistic to appease administrators that don't understand the above stated fact.

For administrators who want a student's activity logs, the student should simply write their activity time on a sheet of paper and submit it.  If administrators don't trust what the student writes down, it is a shame that they trust bogus data from the Canvas LMS.

[dwillmore]

This conversation needs to be picked up again.   I have a very hard time explaining to faculty, designers, administrators that the 'Total Activity Time'  is worse than not absolute, it is meaningless.   It would be more helpful for all if it were removed from the people page.   You can record gets and puts but nothing else indicates active use on a web page at all.

Basically, we are recording time on a page between 2 and 10 minutes.   If you are a page for more than 10 minutes it is not recorded (maybe only 10 minutes is recorded?).  Correct me if I am wrong, please.   Again, this time does not track activity, just time on page, but that is even inaccurate by design. An assumption was made that you could determine an active user if they click from page to page from 2-10 minutes excluding some pages.   When faculty look at the people page and see 02:48:46 they think the student has spent almost 3 hours in the course and make judgements based upon that, or worse, administration sees that an instructor has spent 05:00:00 Total Activity Time they conclude that the instructor is doing nothing in the course.

It may be possible to make a guess at the activity in the course at the high times and low times recorded, but even that is highly suspect.   Looking at the type, length, and quality of communication content and if you are assessing properly the grades.

This is my opinion on what I think is a very misleading metric.

[richard-jerz]

David, thanks for supporting my original position.  Apparently some folks, maybe administrators, like to show how active students are in Canvas so they like this statistic, and they like how it errors on the high side.  But as you have also noted, it is meaningless. In my experimenting, it appears that a student's activity time in a second course will be increased even when they were only in a first course (meaning the first course activity time gets added to other courses.)  I have also been unable to verify that the Canvas logic can be always reproduced.

A much more accurate statistic would be "total clicks in Canvas by a student."  But many people want "time."  Total clicks could be multiplied by any "time per click" value that an administrator wants to use, such as 1-minute per click, or 10-minutes per click.  But then the burden would shift away from Canvas' assumptions to the administrator's assumptions, and it appears that administrators would rather not take responsibility for it so they seem to prefer having Canvas keep this in.

This poorly conceived statistics makes me question "any statistic" within Canvas.  It makes me question the Canvas developer's way of thinking.  My position, like yours, is that it should be removed.

[kona]

 @richard-jerz ‌, as an (evil) administrator I tell faculty to ignore the total time spent and that it's junk. Since I also teach (statistics) I took total time spent for each student and overall course grade and found pretty much no correlation between the two (if I remember correctly, and this was a few semesters ago, r was between .20-.25). More time does not equal better grades or more learning. Yet, as an administrator I also know that sometimes we have have to justify things at a level higher than ourselves. This could be at the Institutional level, State level, or Federal level. If someone tells us we have to assign a specific number to something or else lose funding then guess what, we'll try to find a number to assign to it <-- and yes, this doesn't always mean it's the right number or even a good number, just the only number that's available.

What I've found useful, but once again not statistically significant, is whether students are even accessing the course materials. For this I use the following canvancement‌ - Obtaining and using Access Report data for an entire course‌

Something else we use at our College that's much more accurate for figuring out if students are doing what they should be doing are the metrics assigned by dropout detective (by  AspirEDU‌). Dropout Detective ranks students from most at risk to least at risk based on a number of factors including: missing assignments, late assignments, course grades, how often the student accessing the course and  materials, etc. We've been using it since 2013 and it is normally spot on in determining how a student is doing.

I hope this has provided a different perspective and some insight.

[Chris_Munzo]

Dropout Detective has never measured "activity time" or "page views" as a factor in our risk assessment, for the reasons mentioned in many of these comments.  Just because something can be measured, it doesn't mean that measurement has significance.  Submission of gradeable items, course access and grades (or "marks" for you,  @bwalters ) have value, for sure.  

[richard-jerz]

Kona, so if you tell students to ignore this, it might be better if the statistic wasn't shown so you don't have to waste your time explaining that it is "junk."

I didn't mean to imply that all administrators do this, and that administrators who do do this are "evil."  What I meant to say is something like you said, that there are some administrators at some levels who do not understand that the LMS cannot provide accurate "time" information, and that "time" is not by itself correlated to learning.  You said it better.    And I also understand the issue of creating information for funding (or accreditation) reasons.

If one has a good grade book, and has a system to penalize students who turn in late work, then the grade book itself becomes one good indicator of student success.  This is what I see most often.  If a student has missed, let's say, 60% of the first 10 assignments, they are at risk of failing the course.  And yes, when I see this, I look at a report that shows which materials they have accessed.  Then I can say to the student "You are not doing your assignments, and I can see that you are also not looking at the links that provide information about the assignments."  Then I might say "Do you have "time" for this course?"  But maybe what I am really saying is "Do you have "energy" for this course?" Well clearly it does take some "time" to be involved.  But a good student who takes less time to do well should not be penalized because they didn't spend more time in Canvas.

[kona]

This times one million!

But a good student who takes less time to do well should not be penalized because they didn't spend more time in Canvas.

Did you mean students here or faculty? I only ask because at least for us students can't see total time spent in course - at least not from the People page.

...if you tell students to ignore this, it might be better if the statistic wasn't shown so you don't have to waste your time explaining that it is "junk."

And the evil part was a joke. I'm married to a full time faculty member and he always says he's married to the enemy... an evil administrator. 😉

[richard-jerz]

Good questions, and points, Kona.

If an instructor is somehow going to award points for "time", why should they be thinking of penalizing students who do well in a course but spend less time in it?  If this (meaning "time") is part of the course grading metics, it might be unfair that students do not know how much time they have spent.  So maybe students should be able to see their time.  Why not?

Okay, from what I can see, you are not an evil administrator.  

I know how this goes.  If someone said to me "If you cannot tell me how much time a student works in your course, you are fired!" then I would love this "activity time" statistic.  And I would love that I could shift the "blame" for this poorly created statistic to Canvas and away from me.  But I would still see it as unethical to be faking data, even though it appeases my administrators.

[dwillmore]

Kona, don't let it leak out that we are not all EVIL.  We need to maintain our rep.

[kona]

LOL

[dwillmore]

We use that here as well.   I wonder who that man is that keeps coming up with these gems?  @James ‌

[James]

Sorry  @dwillmore ‌, I don't think I can help on this one. I'm still trying to figure out who put the bomp in the bomp bah bomp bah bomp and made my baby fall in love with me.

[dwillmore]

LOL, if you figure this out give me the secret as I am still trying to figure out why my better half sticks around.

[kona]

As James' other half I'd say it's his brains... and that we make pretty babies. 🙂

[dwillmore]

Man those are cute kids, but I think it is the pure happiness on their faces that is an even greater accomplishment.

Ok, now it is an official facebook thread.....sorry.

[kona]

LOL... I think we can be forgiven this time! It's the sense of community (and friendship) that makes the Community such an amazing place!

[tdelillo]

^^^ This ^^^

[aolsonpacheco]

Hi Kona,

Thanks for your post. I've been researching how to track daily student activity (measured by time) for a program that needs to report this in order to receive funding. While I agree with all the reasons stated why time-tracking not the right metric, we need to find a solution. Do you know of any use cases that have found a workaround or that have successfully convinced the powers that be to change that metric?

I know about the Access Report canvancement, but am wondering if there is anything else out there for fulfilling funding requirements that track "positive attendance" with daily time in / time out logs for noncredit courses in particular. Any pointers would be greatly appreciated!

[BrendaBerres]

Ha. Thank you. This is what I needed to know. I have a student with 55 hours in the course in the first week. I was highly skeptical. It's clear that the time being tracked isn't actual "activity" in course. I'm a little more certain of the students showing 18 minutes (or 2 minutes . . . ) than the one with 55 hours. 😕 Thanks again for sharing your experiment.

[AlishaAlleman]

Currently there is not a way for Canvas to show users (students) the time they have spent in a course, correct? If not, is this a feature that will be added in a future upgrade? We have courses that have an 80 hr requirement and want to accurately track these hours, thus the reason for the question. As an instructor, we can see Total Activity time but the student cannot so we are manually transferring this data and it is very time consuming. Looking for a better solution. 

[Chris_Hofer]

@AlishaAlleman ...

Even though you can view the time logged in to Canvas via the "People" page to get an idea of how long students have been logged in to your course, is that really an accurate number?  For example, a student could log in to a course, do some work for 20 minutes (responding to a discussion post, completing a quiz, etc.) and then walk away from the computer for an hour...still being logged in to Canvas.  Canvas is still tracking that time even though the student really isn't doing any work in the course.

Just curious to know how you are measuring and recording time actually logged for your students.

[bwalters]

Thanks stefaniesanders​​ that's really helpful. I agree, I think analytics gives a much better view of student behaviour and engagement.

[lesley_cryderma]

Is it still true that activity that takes place within the Mobile app is not documented in these numbers at all?  This is becoming a bigger and bigger issue as students are using the mobile apps more and more.  How do we know (when and how much) if a student is using the app?

[Stef_retired]

 @LesleyCryderman ​, Canvas Data captures mobile app activity: Identifying Mobile Users in Canvas Data

[mp97]

Trying to figure out if Canvas records last activity every time a computer is woken up from sleep mode. Student says he was not on Canvas and yet it shows "last activity" during an exam.

how do I see what pages were accessed?

[AnneBach]

I also have this question, but specific to how far back does that total activity go? Is it for the past week, for example, or is total activity since the beginning of the course? Can we change this setting overall to show how much students have been active, say, for the past 7 days? 

Thanks!

[James]

@AnneBach 

The total activity is the total for the entire course since the user's first activity and you cannot change or filter the duration.

If you want to track usage for the last week, then one solution is to set up your own database where you use the Canvas REST API to download the enrollment record for each student on a daily basis and save it the total activity (I would also recommend saving the last activity). Then you can query your database to see how much time they've spent in the last week or month.