These are chat archives for FreeCodeCamp/DataScience

15th
Sep 2015
M Used
@mildused
Sep 15 2015 00:18
I'm still confused on what to do...
dc
@dcsan
Sep 15 2015 05:24
i'm confused as welll..
i think this part is the data:
they don't send any message after 15 days of their last participation, we got that 721 of the total participants (1726) in the Gitter FCC/Help room wer considered active when observing activity during 2014-12-30 and 2015-08-16. Similarly, if we assume that a participant is a frequent one if he/she has at least 5 messages at different days, that number is drastically lower: only 308 people were considered as frequent visitors, visiting Gitter Help room to an average rate of 42% of the all days between their first and last message, being the longest stack 83 days long.
but its hard to parse that paragraph into a conclusion - for me at least.
are you saying there are ~300 regular daily active users?
there are many ways to defined "retention rate" and one is that a user is retained if they log in at least ONCE in the period.
they are in the D7 RR cohort if they login once that week
for a comparison games aim for a 30% D7 Retention Rate
Rex Schrader
@SaintPeter
Sep 15 2015 05:28
I would like to compare Gitter usage with FCC website login and/or waypoint completion.
I'm assuming that only a subset of active users use the chat rooms
dc
@dcsan
Sep 15 2015 05:51
yeh i was hoping someone could add google analytics to camperbot, then we'd have all that data, broken down by room even.
evaristoc
@evaristoc
Sep 15 2015 06:17
This message was deleted
evaristoc
@evaristoc
Sep 15 2015 06:34

Hi people! (@dcsan and @SaintPeter)

  • @dcsan: you are right, but not having access to login I am measuring those who send at least one message per day. The measure is a daily one.
  • I am measuring active as sending at least a message at any day.
  • I also give 15 days (arbitrary, based on experience) of no sending message to classify them as no-active in the room (they could still be active in FCC, be active in another room, or be passive users, we don't know yet) and defining the frequent use as sending messages during 5 days or more. You can say this is not a retention measure as you mentioned above, but an activity measure.

It seems that:

  • Many people came during that period just to post one or more messages during less of 5 different days (no frequent) and became no-active in that room afterwards.
  • It seems that there was a small fraction of the group of people who came back and were actively participating in the Help room (possibly the helpers?).

Anyway, as I mentioned, there are a lot of "invisibles". I am personally not sure if an event-driven capture of data from the bot using google analytics will supply more info about no-actives, but certainly will let us know if those who are active are using the bot. I am not sure though... I think both analyses would give us a better picture about how some people use Gitter, but at the moment there is no info about all people using FCC.

evaristoc
@evaristoc
Sep 15 2015 06:51
@mildused: would you like to help to translate some python code into javascript / node? I am thinking about using functional programming, fs library, etc and try to take advantage of other features at node when translating that code...
@techstonia: welcome!
evaristoc
@evaristoc
Sep 15 2015 08:25
AND!! @dcsan and @SaintPeter: I think that Gitter people are providing data in the API about those who are reading the messages. I have to see if that information can be used to make some analyses about potential passive Gitter users. I don't think it will be precise as I have noticed that it takes some time for Gitter to mark messages as "read" (which I think triggers the event capture...) but we can give it a try...
I will explore the attribute and let you know next week...