Re: laptop/edu-users

["Karsten M. Self" <kmself@ix.netcom.com>]

Roger Dingledine wrote:
> 
> In message <35ABBBBF.C8A03DD@ix.netcom.com>, kmself@ix.netcom.com writes:
> >Looks like we're shaping this section up.
> >
> >> [_] Check here if you would like to use a laptop.
> >>
> >> And then the laptop question will put in another section for people
> >> who check it. (Note that I ask "would like to use", not "currently use".)
> >
> >I like.  Language:
> >
> > "Check if you use or plan to use a laptop"
> >
> >  ...covers all bases.
> 
> Ok, so I'm going to try to blow this laptop issue wide open again. Sorry.
> 
> I don't think we should have this checkbox on the front page of the
> survey. In my opinion, this is pretty much the same as having little
> checkboxes for "Check if you use or plan to use a modem", "Check if you
> use or plan to use a printer", "Check if you use or plan to use
> mathematical modelling software", and even "Check if you use or plan to
> use a monitor". All of these issues have certain user groups who have
> specific needs and specific concerns. Is there really much of a difference
> between asking somebody if they prefer interlaced or non-interlaced
> monitor resolutions, and asking them how important the weight of their
> laptop is, from the perspective of the operating system?

I'll argue that there is a difference for the following reasons:

 - "Laptops" (and other portables) as a class of machines collectively
possess a number of hardware features which strongly differentiate them
from any other hardware configuration.  You want to ask certain types of
questions, you don't want (or need) to ask others).

 - I suspect the majority of laptops are used in business for what
someone (virtanen?) called "Road Warriors".  Typically sales and
marketing staff, needing email, fax, file transfer, and presentation
tools.  Other uses exist, but this is the big one.  So there's a strong
uniformity of end use.

 - Specific software/OS features working around
hardware/weight/power/networking limitations are needed.  Again, a
number of capabilities, much similarity across the field, and little in
common with other HW configurations.

 - I don't think we want to worry about specifics of hardware -- that's
outside the scope of the OS.  We do want to worry about supporting
special HW configurations or needs within the OS.

 - I'm not arguing this because I'm a laptop user -- having nags waiting
for me at home and work is bad enough -- I don't need to lug my
addictions around with me.  I just think it's a significant interest
area.

> But I think that having a bunch of "check here if you want feature foo"
> checkboxes on the first page of the survey, and then using those checks
> to build a survey once we know what the user already cares about, is
> extremely backward. At the least, we will skew our survey. More likely,
> we'll also end up asking the user each question twice ("Should I ask you
> about foo?" "Ok, now that you're expressed interest in foo, how important
> do you think foo is?")

Agreed, would not be good.  The 'Laptop' checkoff could lead to seperate
questions on hardware/hardware support, later in the survey.  This might
be a better way to approach the issue.  A non-portable HW section could
be the alternative -- again, focus on HW support in OS, not HW
components request (maybe?).

I don't see other HW questions as having the same level of impact and
importance.  Monitor, modem, networking support, etc., are not specific
to HW or use.  Portables introduce their own can o' worms.

<snip>

> Actually, that brings up another interesting point. Instead of just
> collecting information and saying "thank you", why not compile the
> information they've given about their interests and goals, and give
> them a specialized set of recommendations once they've finished the
> survey? This might answer virtanen's question earlier about "How can
> we motivate people to take the survey?" -- we motivate them because
> they get automated personalized advice once they're done. This has
> other issues tho, like people experimenting with our survey to see
> what advice we give for which user type, when they don't actually
> belong to that user type. (And it's also more work. Whee.)

Mixed feelings.  Recommendations can be hard.  Remember the financial
advice website which generated the same mix of a company's investment
products for Bill Gates and the welfare mother?  This was maybe a year
ago.  Big stink.  Has to be done right, with thought.  Tough.  

Showing compiled stats can be interesting as well.  Motivation is
generally tricky.
 
> --Roger

-- 
Karsten M. Self (kmself@ix.netcom.com)

    What part of "gestalt" don't you understand?
    Welchen Teil von "gestalt" verstehen Sie nicht?

web:       http://www.netcom.com/~kmself
SAS/Linux: http://www.netcom.com/~kmself/SAS/SAS4Linux.html    

 10:41am  up 32 days,  8:10,  2 users,  load average: 1.11, 1.08, 1.02