Date:         Sat, 31 Oct 2009 11:46:34 -0500
Reply-To:     Joe Matise <snoopy369@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Joe Matise <snoopy369@GMAIL.COM>
Subject:      Re: Help with SAS code
In-Reply-To:  <88674200-58C8-4F2D-82A7-1D8743073065@alumni.stanford.org>
Content-Type: text/plain; charset=ISO-8859-1

I think I fall somewhere in between on this one.  When I initially read the
posting, I immediately felt somewhat like Gerhard felt; it looked to me like
a request by someone who doesn't know SAS at all and wants someone to
implement a solution, effectively an unpaid consultant.  I however didn't
respond, both because I generally avoid getting into the middle of such
things, and because I wasn't sure what I felt was correct - just an initial
feeling.

Certainly SAS-L regularly helps people with 'individual problems', myself
included.  And for relatively concrete questions, that's exactly how it
should work; someone doesn't know how to do (something), and asks how.

However, SAS-L is not a destination to obtain unpaid consultants [except
when it is :) ]; and as such, the (something) needs to be fairly concrete
and specific, if you want specific help.  And as such, I think the problem
with the initial post was quite simply that he didn't provide code or data
such as to indicate what he wanted to adjust; with the code and data, it
sounds more like every other request we get on the list from a new-ish
poster.  The request for off-board contact for the code was a bit
off-putting, as it not only requires additional work on the part of the
responders [a very bad idea, because increasing the necessary energy to help
is like increasing the activation energy of a reaction - makes it much less
likely to get help, out of proportion to the extra energy required], but
also (as Gerhard notes) takes it away from being a 'community' help, and
turns it into exactly as he puts it, an unpaid consultancy.  It also wasn't
clear to me that the OP knew anything about SAS, or had a desire to actually
learn how to code in SAS; and while that's not a problem per se, it again
makes me much less interested in helping and continue to wonder if the OP
really just needs a SAS consultant.

That said, I also don't think we should discourage folks from posting, just
because they don't know how to post what we need; then we'd get very few
postings indeed.  I think it's much better to respond to the OP of anyone
who doesn't respond with <data> and <code> when applicable and <desired
results> with a form letter of
{We are more than willing to help.   Please provide a sample (real or faked)
of your in-data, sample results desired, and where applicable any SAS code
that applies to your current problem.  Specific questions ("How do I
transpose this data so that my variables are organized by state", eg) will
be met with specific answers("Do this: PROC TRANSPOSE data=test; by
state;..."), while more general questions ("How do I design a reporting tool
to display all of my students' grades") will usually be met with more
general answers ("Use PROC MEANS to summarize your data, and then use PROC
REPORT to display the resulting dataset.").}

Very much wish we could have an initial page like that on the google pages,
or an automated response from the listserv... would avoid things like this
(which are what, 60% of posts, one way or the other?)

-Joe

On Fri, Oct 30, 2009 at 10:48 PM, Jack Hamilton <jfh@stanfordalumni.org>wrote:

> I don't agree with Gerhard Hellriegel's statement that the list is not
> intended to provide solutions to individual problems.
>
> My view is that the list members can answer questions however they see
> fit.  Sometimes very detailed solutions to very particular problems
> have been provided; it just depends on who's reading, how much time
> they have to look at questions, and whether the question is of
> interest to them.
>
> You are more likely to get an answer if you provide a clear statement
> of your problem, with sample data, and at least a sketch of the
> desired output.  If you are trying to implement a proprietary
> algorithm, make sure to explain exactly what you want; don't think
> that we will be able to guess.
>
> So I don't think your question was out of line, but you didn't provide
> enough information for anyone who's not a minder-reader to be able to
> answer it.
>
>
> --
> Jack Hamilton
> jfh@alumni.stanford.org
> Caelum non animum mutant qui trans mare currunt.
>
> On Oct 29, 2009, at 11:21 am, Daniel R Flowers wrote:
>
>  I apologize for the lack of clarity in the initial email.  As I
>> mentioned, I am new to the list and was not aware of the specific
>> goals and composition of the list you referenced based on what I had
>> read from the list homepage.  There is code on our end that exists
>> and we are able to edit and develop.  As we didn't want anyone to do
>> our work, we had tried to make this a more general concept question
>> (although beginning) then something very specific.  I apologize that
>> this was out of the scope of the listserv and will be happy to
>> reference other more introductory sources.  Thank you for your time
>> and I apologize for the inconvenience.
>>
>>
>> ----- Original Message -----
>> From: "Gerhard Hellriegel" <gerhard.hellriegel@T-ONLINE.DE>
>> To: SAS-L@LISTSERV.UGA.EDU
>> Sent: Thursday, October 29, 2009 1:16:26 PM (GMT-0500) America/
>> New_York
>> Subject: Re: Help with SAS code
>>
>> don't know how the others think about that, but many of the people
>> which
>> share their knowledge here are consultants. They get their money for
>> writing applications for customers. It is great that they provide
>> help for
>> some sophisticated problems (also for not sophisticated, if there are
>> some), but it is not the goal to do other's work! The only goal is
>> to help
>> other's over some obstacles, which prevent them to do their work.
>> If I read something like "If the current SAS code would be helpful,
>> feel
>> free to contact me on an individual basis" I don't have positive
>> feelings!
>> One thing is, the code is to big to provide it here. Another is, there
>> seem to be no ideas about concrete problems, only "it does not do
>> what we
>> want". A (for me) very important thing is: this list has NOT the
>> goal to
>> provide INDIVIDUAL problem solutions, but should give everybody the
>> chance
>> to participate.
>>
>> I have two "solutions": 1. look for a SAS-base course. 2. make a
>> contract
>> with one of the SAS consultants to help you with concrete
>> applications and
>> maybe give you enough informations to do it yourselve in future, if
>> you
>> have time enough to do so.
>>
>> To be not missunderstood: if there is a question like "we have that
>> kind
>> of data (example) and want to summarize them with that classification
>> (example) and expect that result (example), but get the following
>> (example). We tried it with the following code (code-part). How
>> could we
>> do that better?"
>> or
>> "we have that code and get the following error which we don't
>> understand... - Why?"
>>
>> all that is ok and you'll get the best answers here, also for
>> beginner's
>> questions.
>>
>> But not like "we want a report with that content, write it for us!"
>>
>> All only my personal opinion, maybe others think different.
>>
>> Gerhard
>>
>>
>>
>>
>>
>>
>>
>> On Thu, 29 Oct 2009 11:43:38 -0400, Daniel R Flowers
>> <dflowers@SAINTMARYS.EDU> wrote:
>>
>>  Hello,
>>>
>>> This is my first post to the group.  Hello!  My director and I are
>>>
>> relatively new to the world of SAS code as we are more heavy users of
>> SPSS.  A SAS code existed in the office before our time that generated
>> course evaluation reports for individual faculty members and
>> departments.
>> Essentially, we have code that generates individual reports by faculty
>> member and by department.  Within the individual faculty reports, we
>> show
>> totals by response and provide an overall comparison to the college
>> with
>> averages of all courses offered at the college (as we have a standard
>> evaluation form).  We would like to be able to add a departmental
>> comparison line (so in essence we would have the break-out for the
>> individual faculty, a comparison to the overall college average, and a
>> comparison to the departmental average).
>>
>>>
>>> Note: The SAS file is pulling from a text file generated from a
>>> scantron
>>>
>> of the evaluation form.
>>
>>>
>>> We have not yet been able to generate code to do this.  Any help
>>> would be
>>>
>> much appreciate.  If the current SAS code would be helpful, feel
>> free to
>> contact me on an individual basis.  I very much appreciate any advice.
>>
>>>
>>> Daniel
>>>
>>