[Design] Collecting User Data from Dogfooders

[Mimi Yin]

Ashkan, this summer's PPD intern has created an eventLogger parcel in  
Chandler that will log user interaction events and send them to us  
for user analysis.

If you've been paying even lackadaisical attention to the design  
list, you know very well that there are often contentious design  
issues on the list. Now we have an additional tool at our disposal to  
better understand user behavior and interaction patterns.

The eventLogger is still in its infancy and does not have access to  
all of the data we'd like to collect.

It logs most, but not all click interactions. It doesn't log editing,  
although sometimes we can infer edits when we see that item titles  
and collection names have changed.

If you're dogfooding Chandler today, even if it's only for  
calendaring, please considering upgrading to today's build: http:// 
builds.osafoundation.org/and turning on Start Logging in the Test  
menu to send us data! (See below for screenshot.)

Some of the questions we would like to start collecting data for in  
the Alpha 4 and 5 timeframe (in other words...when the Dashboard  
becomes experimentally usable) include:

What are people using Chandler for? Keeping track of their own  
schedule? Scheduling and Invitations? Task management? Organizing  
ideas and static information, aka notes?

What data do we need to collect in order to answer these questions:

1. Total time in each App Area / Collection
2. Total # of interaction events in each App Area / Collection
3. Breakdown of interaction events in each App Area / Collection
    * # of New items created
    * # of Edits - Can we track it on a per field basis? per Stamp?  
Never share button in the mark-up bar?
    * # of Navigation events - e.g. Mini-cal interactions
4. Lengths of continuous sessions in each App Area / Collection
5. # of continuous sessions in each App Area / Collection
6. # of events per continuous session in each App Area / Collection

Within each App area, we'd like to know:
1. # of collection context switches total
2. # of collection context switches per continuous session
3. Amount of time spent in each collection
4. Number of interaction events in each collection
5. Breakdown of interaction events in each collection

Can we figure out the inverse? For each collection, how much time is  
spent in each App area?

1. How often does stamping happen?
2. Breakdown of stamps: # of Event/Tasks versus Event/Messages
3. How often does stamping happen in the Detail View versus the Table?

=====


Thanks!

Mimi

-------------- next part --------------
Skipped content of type multipart/related