[Events] Crafting an exit survey for campus events

[Asheesh Laroia]

Excerpts from Kevin Carillo's message of Thu Apr 12 16:25:32 -0400 2012:
> * In the time since the workshop, have you participated in an open source
> project? Answer with one of the following:
> 
>   ( ) No increase in participation
>   ( ) Increased participation as an observer (reading mailing list, or
> reading bug reports)
>   ( ) Increased participation as a visible contributor (writing to the
> mailing list, or submitting bug reports)
> 
> There are several issues with this question. First, the question is about participation and the answers are about increase in participation. If you want to capture participation increase then perhaps ask:
>  In the time since the workshop, has your participation in one or several open source project(s) increased?
> ( ) Decrease in participation
> ( ) No increase in participation
> ( ) Increase in participation 
> 
> You may be more specific if you want to assess increase level (to what exent have your participation .... increased?). I would be tempted to go for this option instead.
> 
> Second, I would personally suggest to be more detailed as it would be interesting to know the kind of behaviours you are generating (readling mailing lists, reading bur reports, wriiting mesages in mailing lists, writing bub reports, writing code...)
> This aspect measures your overall effectiveness in generating new open source contributors.
> 
> I would thus use separate questions instead of having all in one. 
> 
>  In the time since the workshop, has your participation in one or several open source project(s) increased?
> 
> ... as an observer (reading mailing list, or reading bug reports)
>  ( ) Decrease in participation
>  ( ) No increase in participation
>  ( ) Increase in participation 
> 
>  ... as a visible contributor (writing to the mailing list, or submitting bug reports)
>  ( ) Decrease in participation
>  ( ) No increase in participation
>  ( ) Increase in participation 
> 
> * If a friend or colleague heard about an OpenHatch-led Open Source Comes
> to Campus workshop, how likely is it that you would recommend them to
> attend?
> 
>      Extremely likely to recommend against
> 
>      Moderately likely to recommend against
> 
>      Slightly likely to recommend against
> 
>      Neither likely to recommend nor recommend against
> 
>      Slightly likely to recommend
> 
>      Moderately likely to recommend
> 
>      Extremely likely to recommend
> 
> 7-points is perhaps too much here and the wording of answers like ""Extremely likely to recommend against" could lead to confusion for some respondents (and this is a question that is important of OH, you don't want to have errors because of unclear choices).  
> Your choices actually capture two ideas: recommending and recommending against (which are two different things). You may consider focusing on the one that you are the most interested in or even have two separate questions if you prefer. 
> 
> Personally, I would go for:
> 
> If a friend or colleague heard about an OpenHatch-led Open Source Comes
> to Campus workshop, how likely is it that you would recommend them to
> attend?
> 
> "very likely," "likely," "neither likely nor unlikely," "unlikely," and "very unlikely." 
> That should be enough. 
> 
> All this will help. Good luck .

Thank you, Kevin! I will make all the adjustmentions you mentioned -- in particular,
I'll go the separate-questions route for the "observer" vs. "visible contributor" stuff.

I'm slightly scared to find out the answers, but that's science. (:

(...Asheesh fiddles with Google Docs...)

I've made this form:

https://docs.google.com/spreadsheet/viewform?formkey=dGNtSUk1STVlYzlwekZRX0pQS0F2MVE6MQ

and I'll distribute it later today to UMD attendees (by sending them a direct
email, so that I can pre-fill Name and Email) unless there's more feedback on
the form's contents. I'll wait about 8 hours for feedback, then send; if your
thoughts arrive after I can use them for the UMD one, then I'll adjust future
surveys to match people's suggested changes.

Also, I wish there were some convenient text-based representation for Google Docs
surveys, or other surveys like this, that I could store and then auto-create a
new Doc for each event. Maybe I should look into LimeSurvey or a Google Docs API.

-- Asheesh.