[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SEUL: draft survey: v.0.001

Roger Dingledine wrote:
> O btw, it looks like the 'lust' acronym has already been taken:
> http://www.ch4549.org/lust/
> :)

Damn.  I quit. <g>

> In message <358DBE1B.C19522DA@ix.netcom.com>, kmself@ix.netcom.com writes:

> >> 2. I have used this operating system for how long? [exclusive: <1month,
> >> <6months, <1year, >1year]
> >
> >Scale:  < 6 mos, 6 mos < 1 yr, 1 yr < 5 yr, 5 yr < 10 yr, 10 yr +
> We kind of have a problem here. Nobody on the linux side gets to check
> the 10+ button. We're going to have to be careful about interpreting
> results.

OTOH, this becomes a screen for dubious responses.

> >> 4. profession: exclusive: (options?)
> >
> >Do you want to ask "profession" or "position" or "industry"?  There are
> >some standard classification schemes, you might want to check US Census
> >bureau.  When asked, "profession" is usually a fill-in-the-blank type
> >response.
> Oo. 'Position' might allow us to distinguish the decision-makers from the
> non-decision-makers. But then again, probably not (but it's a neat thought).

See also the SAS User Linux Interest Profile for categories.  This also
starts to answer some of your followup questions:

                    <option> President/CEO
                    <option> CIO/CFO
                    <option> Other Executive
                    <option> IS Directory/Manager
                    <option> Other Manager
                    <option> Programmer/Analyst
                    <option> Other Technical
                    <option> Other

...I was mostly interested in programmers and managers.  We would
probably want to expand "Other".

Industrial Sector:

                    <option> Academic
                    <option> Financial
                    <option> Insurance
                    <option> Government
                    <option> Healthcare
                    <option> Industrial Manufacture
                    <option> Independent SAS Consultant
                    <option> Internet Services
                    <option> Non-profit
                    <option> Semiconductor
                    <option> Pharmaceutical/Biotech
                    <option> Professional Services
                    <option> Retail/Merchandise
                    <option> Software
                    <option> Other

> >General categories -- some of these are terms we *don't* want to use,
> >but as ideas:
> > - Unemployed/not working
> > - Retired
> > - Student
> > - Home Maker
> > - Laborer
> > - Skilled laborer (incl. factory) ?
> what's the difference between these two? we should be clearer here.

I don't know.  Some of these are vague.  See my above responses for what
I think is a better approach.  I don't think we need gross levels of
detail, we don't want to over abstract either.

> > - Journeyman (carpenter, plumber, glazier)
> > - General office
> secretary? what is this?
> > - Business Professional
> > - Computer professional (operator, analyst, programmer, sysop)
> > - Business management
> > - Computing management
> should we divide these up further? the goal of the user profile
> section is to figure out what sort of person is answering the rest of
> the survey, so we can know how to interpret their answers. it's very
> relevant if it's a sysadmin or a grunt programmer filling out the survey.
> > - Business executive
> > - Computer executive (CFO)
> > - "Content provider" (we don't want to use this term -- artist, writer,
> >other creative professional)
> 'liberal arts' if we ask it in terms of education rather than profession.
> perhaps we should ask both, given that most liberal arts people aren't doing
> liberal arts right now...(i'm sure there are other examples of that)
> > - Professional (doctor, lawyer, veterinarian, etc.)
> >
> >
> >> 5. income (optional): exclusive: <$10k, 10k-20k, 20k-30k, 30k-50k, 50k-100k,
> > >100k
> >>   (should we do 'household income'? what's standard here?)
> >>   (should we care that we're being US-centric?)
> >
> >I'd do "household", in broad ranges.  < 20, 20 - 59, 60 - 119, 120+
> >Basically: poverty (or starving student), lower middle, middle, made
> >it.  Any measure has to be denominated in some currency.  Simply specify
> >US dollars or equivalent.
> makes sense. it's also a 4-choice answer, which is what we're trying to
> use when we can.
> >> 6. education level, and specialization (if applicable)
> >
> >Some secondary, secondary, some college/2 year college, baccelaureate,
> >advanced degree, technical training, professional degree.
> what's the comparison between advanced degree (presumably that's a Master's)
> and 'technical training'?

Technical might be "dental assistant", or other, usually vocational
training.  Think "Heald Business School" or "De Vreis (?) Institute of
Technology".  Usu 1-2 year programs.  Notorious as the largest source of
student loan defaults, FWIW.

> >> 7. missing anything?
> >
> >Yes.
> uh, thanks. ;)

No problem <g>
> >How about, general areas, fill in later:
> >
> > - General use:
> >   I use my computer for: games, internet/email, schoolwork, small/home
> >office (SOHO), office applications, dedicated/business systems,
> >development.
> graphics, number crunching/computational algebra, databases are other
> things to consider (stolen from the german survey). also, "computer(s)"
> is more appropriate.
> It would be nice to figure out which OS they used for each thing they
> did. I don't see an elegant way of doing that yet, though.
> > - Rank the following in terms of importance (low, moderately low,
> >moderately high, high):
> We might need to provide something that will help guide them to know
> exactly what we're asking, else we risk being ambiguous. importance to
> productivity? importance to personal happiness? general impression of
> what the user thinks bill gates deems important? (maybe i'm being
> paranoid)

Ambiguity is a problem w/ the SAS User survey.  I ask users to rate
importance of issues to their decisions, but some factors might foster a
decision because they are perceived as better on the alternate platform,
or worse, on the existing one.  It's a distinction not drawn in the
survey.  A phrase or sentence describing each issue might be helpful. 
OTOH, anything which requires more than this to describe, probably
doesn't belong on the survey.  If we really want the question answered,
but it's a difficult concept, we can work at simplifying it, breaking it
up into components, or bending our rules and adding a longer

> But I can't come up with a good rewording for it. I just know that
> at each question, I would ask myself "what do they mean 'important'?"
> Something more concrete would help me out there.
> I wonder if it would be useful to split between "how crucial is it to"
> and "how useful would it be to". Most required features can be kludged
> into existence, but there are easy ways to kludge them and hard ways.

This could be a basis for re-scaling the response categories:

 unnecessary, of little use, useful, necessary

...though this starts to make measurements in two dimensions.  OTHOH,
how many things are useful but not necessary?
> >   - local networking (connecting to other computers at my office, home,
> >or other location).
> >   - wide area networking (connecting two or more locations)
> are there any examples of this besides businesses creating wan-company-
> networks? I think we might be self-selecting the answerers of this
> question. (But then, we might surprise ourselves. And it could
> also be valuable to actually document that assumption that only
> businesses care about this is correct.)

Both good points.  My guess is that most people won't know or understand
these terms.  We may find that there's some cyber-grannies out there
setting up family WANs.  God knows (so let's ask Linus...).
> >   - internet connection (direct or dial-up internet connection)
> >   - privacy (ability to keep other users from reading my files)
> >   - security (ability to prevent unauthorized people from using my
> >system(s).
> >   - stability (computer and applications run without crashing or
> >requiring restart)
> >   - multi-user (several people can use machine, possibly
> >simultaneously)
> these should be separate questions. having the computer keep separate
> information about the different people that use the machine is very
> different from letting multiple people do stuff at the same time.
> ("How can that work? I only have one keyboard.....")

Right.  Take my suggestions as suggestions.  Most will probably
subdivide into further questions.  Some may collapse into one another.
> >   - applications (I need to run specific applications on my computer)
> >   - uses (I need to use my computer for specific tasks.  Specific
> >applications don't matter as long as they fill the need).
> >   - backups, administration
> easy to use, powerful, fast, documented...
> >   - adding/removing software
> in a clean and elegant way. everybody has a cp and an rm...
> >   - upgrades
> easy to get upgrades, or easy to apply upgrades...(is price an
> issue here too?)
> >   - compatibility with existing systems
> >   - compatibility with prior versions
> >   - ease of use (this *MUST* be split up into "what constitutes EOU"
> >elements).
> this is, unfortunately, not my specialty. i'll try to throw this
> question to a couple of graphics friends of mine.

"EOU" also doesn't fit the response cat very well.  Specific EOU
features would be a better fit.
> >   - GUI
> spell it out, graphical user interface (see below)
> >   - CLI (command line)
> This is a good question to ask, but nobody knows that acronym. Asking
> "command line interface" directly is probably better.
> >   - documentation
> similar to support, is there a difference between delivered (printed)
> documentation and obtainable (electronic) documentation? are drivers
> a separate category?
> >   - support, vendor
> >   - support, internet
> >   - source code (probably not a big one, but it would be interesting to
> >ask)
> >   - cost
> is it relevant if the business pays for it rather than the user?
> or is this going to be expressed by just asking about cost?

Ok, let's ask how the computer was aquired:  purchase, gift, provided by
work/school, used at work/school, stolen, other....

> >   - corporate reputation (vague -- good or bad reputation is
> >important.  may not want to ask)
> I'd like to ask it. We could do two separate questions, how important
> is good corporate reputation in your decision to use an OS, how
> important is bad corporate reputation in your decision to not use an
> OS. But wait -- what is debian's "corporate reputation"? Is there
> a broader wording here that doesn't confuse people?

I phrased this "vendor satisfaction" on SAS survey.  Could break this
into tech support, sales & ordering, general impression.

Note that responses here scale differently from most of above.  Maybe
"importance of vendor satisfaction, ..." would fit better.
> >Oh, hell.  I'm all out of ideas.  Feel free to add.
> These are some good ideas. They should be integrated with the
> current draft, which pete has put online (tho it's currently the same
> as the one he posted here last week). I figure the thing to do is to
> throw out ideas for a couple days, then refine them into presentable
> ideas, then finally refine them into the actual words we're going to use.
> final thoughts:
> "is it important to support internationalization?"
> along that line, we might ask for country in the user profile section.

IMO, not important to the individual (I want it to talk my language). 
Very important to ISV (we want to support multiple languages easily). 
Maybe "support in my native language".  Eg:  man pages, texinfo pages.
> Also, consider OCR support (software), as well as speech recognition and
> other specialty applications...

Starts to fall under SW support.  These would be categories of SW.  I
doubt Linus is going to put OCR or speech at ring 0.
> --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    

 12:11pm  up 9 days,  9:40,  8 users,  load average: 1.01, 1.05, 1.07