DISCUSSION FORUM
SUBJECTS
Forum: Suggestions / Ideas
Timer for courses
347 views
Would it be possible to add a timer that counts the exact time spent practising or learning a course? Should be counting minutes/hours per day/week/month.
I would find it really useful to how much time I invest daily for a specific language course for example. It would also add a new layer to the Progress report which for now only focuses on XP earned. I am not even sure how to interpret the XP earned and don't really find much value in this info since it depends on how quick you were to answer.
347 views
13 comments
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
I agree that time spent would be useful to have and should, perhaps, supersede XP as the default metric. It's a good idea.
But, it is a matter of deciding how to count the time spent.
How would you expect to see the time being counted for it to be useful?
e.g.
1) Total sum of "session" time: From session begin time, to session end time. That would give time spent learning, but it wouldn't include time between sessions. Problem is, we don't save that data at the moment. So would only work after we add it. We also need to consider what happens if they start a session , then leave the computer, and come back to it an hour later...
2) Total time between login / log out (but what about switching courses, and users who leave themselves logged in...)
3) Total sum of time from when question loads to question answered (that's what we do record, but I don't think it's good enough for your purposes)
Option 1 seems most sensible to me - e.g. to the effect of: "Total time spent actually answering questions in this course" - it would give the total time spent learning a given course, it's just it wouldn't account for time spent online doing anything else outside of the learning area.
We are going to consider how we can measure it accurately (instead of an approximation) whilst obeying privacy laws etc, let me know if you have any thoughts on the above.
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
Could you please tell me the name of a course you have used, and an estimate of how many hours you think (approx.) you have spent learning it in total?
Option 1 is the one that would give me the data I need, but it would be really helpful to know the duration of sessions per day as well as per week or month.
For the situation in which a session is started and the user leaves the computer, maybe a timeout would help? For example, if the user is not active (answering questions) for let's say 120 seconds, then the timer should record this as the end of the session and resume timer (session started) when the user starts answering questions again?
For your second request, please take as an example my Italian course (not sure if a link would help since it is a private one). I am trying to spend at least 10 minutes daily so I would say that the total for last month (Sept 2020) is of about 6 hours. But there were days in which I spent more time (for example 120 minutes on Sep 19, 35 on Sep 20 and so on).
If I take this as the base time, and allowing a bit more time for the sessions in which I am practising more since the course creation I would say I spent 20 to 22 hours learning and practising on this course.
Hope this helps!
Fender Tremolo
Level: 24
Streak: 0
Using MEMORY to learn Spanish. Pienso que mi habilidad actual es promedio...
Registered:
Level: 24
Streak: 0
Administrator
For interest's sake, we originally designed to have "time spent learning" as the main value of graphs, and an "activity detector" (it detected AFK, mouse click to another browser tab discounted etc) in place originally for this exact purpose. It worked great - we can restore that, but we have to make it privacy-opt-in only this time.
Total time played is a definitely an important metric to have for people spending real time with the application, we will sort this. Just need to do it right rather than provide a wild approximation
I will leave support to reply on your points, thanks for the valuable feedback
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
That's really interesting. Actually, I think we have this data to a reasonable level of reliability already, based on session start time and time spent answering questions. I don't think we need to add an activity detector - we can but I think its overkill - we can also make allowances for exceptionally long questions and just normalise them.
Sorry for the tech talk above. To the point:
Thanks Andreea for providing that useful info. This is what our database guy came out with for your time in September for the Italian course:
Total hours: 7.25
Hours Date Answer submitted
0.23 2020-09-01 199
0.09 2020-09-02 103
0.27 2020-09-03 229
0.23 2020-09-04 198
0.13 2020-09-07 141
0.14 2020-09-08 155
0.14 2020-09-09 136
0.23 2020-09-10 234
0.14 2020-09-11 135
0.13 2020-09-12 144
0.13 2020-09-13 130
0.14 2020-09-14 142
0.13 2020-09-15 132
0.14 2020-09-16 137
0.16 2020-09-17 141
1.58 2020-09-19 1511
0.47 2020-09-20 531
0.13 2020-09-21 153
0.20 2020-09-22 214
0.14 2020-09-23 143
0.14 2020-09-24 127
0.16 2020-09-25 88
0.91 2020-09-26 900
0.24 2020-09-27 271
0.39 2020-09-28 405
0.26 2020-09-29 274
0.21 2020-09-30 230
Hope you don't mind me posting this publicly... if so will delete it. But does that look accurate to you for the Italian course?
-- Edit: Also, total time spent learning this entire course for you comes out as 28.1 hours
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
We decided to jump on this while the idea is fresh. When you next login, checkout the graph area and let us know your thoughts.
We tested it across 25 user accounts and think the timings are quite accurate.
Thanks for the idea, we are open to other suggestions on this whilst we have the resource available to work on it.
RE about the data - no worries for posting it, I don't mind at all. I compared this briefly with the info I have in my habit tracking app and it seems to be accurate, including the days when I had no activity on this course.
RE the graph added to the courses - this is perfect!! 😃 I will check how it behaves over the next couple of days and will let you know if I notice any issues or have any ideas for improvement.
Quick question - will this data be included in the Progress Report from now on?
I will keep the post open for a while, just in case other users might have some suggestions on this.
Thank you for implementing this, it is greatly appreciated!
Some things I noticed so far.
1. I practised about 15 minutes today for the same Italian course, in the graph/table, there are 12 minutes registered. I assume that the loading time between one session and another is not counted, but could you please confirm this? This is my understanding from your comments above.
2. If I move both sliders on Today (From to Today), shouldn't the values for today's session be registered there? I am referring to the Total XP, Total Answers, and Total time spent. If I only have this selected, I have 0 for all.
Maybe this relates to issue 4 since I moved the From slider to 6-Oct and values are displayed correctly (From 2020-10-06 to Today). Maybe I shouldn't be able to select From to Today at all by moving both the From and To sliders on the same spot?
3. If I get the slider to From 2020-10-05 to Today, I am not sure why but I have 17 minutes registered, although yesterday I practised 14 mins and today 12. Only 6-Oct is shown in the graph in this case and if I move the pointer over it, I get 12 minutes displayed so not sure why the additional minutes in the Total values.
4. Any chance another option to select the time period can be implemented? The slider works OK but it's a bit difficult if you want to see your timing for let's say, the last 3 or 4 days. It's even more difficult on small screens than on my desktop.
5. If I use the Fullscreen Graphs option, the graph itself is quite zoomed in, at least on my screen. The table looks OK. Could this be made to adjust to the screen resolution? I also quickly checked this on my iPhone, on both Chrome and Safari and the button doesn't seem to work at all.
Apologies if some of my comments above are confusing, being able to add screenshots to posts would have helped a lot today. Let me know if clarification is needed on any of these.
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
Hi, great , and thanks again for all the extra feedback. Please keep this open and we will use it as a working TODO list.
I will go through each of these issues - all should be rectifiable.
1 - Loading time between sessions is not counted. Any time spent on session complete screens won't be counted at the moment. It would be ideally. Will think about how its possible.
2 - this appears to be a bug - will fix this. I didn't test setting both dates to today only actually, good point.
I think that ties in with 4 - a drop down date selector will be better than the slider (or as well as a slider)
3. We need to do more testing on this. I think there is a problem with what the system considers "today"
5. Agree, full screen is not great - will rework the UI in a stage 2.
Currently, we're working on trying to solve the time zone problem as that ties in with both this and user streaks. It's a bit complicated and a big oversight on our part which we need to correct.
I will let you know if we need more info, but that's plenty for us to go on for now 😃 thanks.
This is actually really good and the system built for it will help us with a new system in progress: giving the user an option to set daily targets (they were going to be just XP, but now we can do time when we get this right).
I am working on these issues. Please note I had to quick-fix the issue today to stop the graph slowing down other users on the site which may have broken some functionality temporarily, it will have a few imperfections until I can finish.
Hey team, from what I can see most issues here were already fixed so just wanted to check if we can close this?
The only other thing that is missing is the ability to select multiple answer types. Not particularly useful for me since I always go for typing, but maybe would be useful for other users? Let me know if you want me to add a separate post for this so that we can close this one.
Thanks a mil for implementing this! It's way easier now to keep track of the time spent on various courses and the daily target feature is a plus as well 😃
Memory Support
Level: 8
Streak: 0
This is a shared account for Memory developers and support
Registered:
Level: 8
Streak: 0
Administrator
Hi, that's great. Glad it worked out well so far. Regarding "stacking" of modes, I've made a note of this as a possible to-do item. It's quite a big change though, wish we thought of that in the design stage now 😃
You can close this one.
This post is closed