Date: Mon, 24 Nov 1997 21:43:20 -0800
Reply-To: AA <hope41@PROAXIS.COM>
Sender: "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From: AA <hope41@PROAXIS.COM>
Organization: All USENET -- http://www.Supernews.com
Subject: Re: SPSS output
Content-Type: text/plain; charset=us-ascii
My two cents.
1. My impression is that SPSS and SAS are going away from the education market
into the commercial market. The Pivot tables have numerous formats, but I've
found one that is appropriate for APA style requirements. Moreover, for folks
write journal articles getting the data ready is what takes the time and typing
tables, while tedious, is a tiny portion of the total work effort. I suppose
business community want 10 variations on color combination defaults for a 2x2
table. I imagine SPSS hired a consultant to select the color combinations. They
have to spend the money some way.
2. What distinguishes SPSS from SAS is SPSS's approach of offering a useless
package (useless for researchers), and then having tiny add on procedures
the missing values module Dick mentioned) cost a huge amount for the amount of
they involve. Price is determined by how much they can get before we go to
stat package rather than by the cost of developing the production. SPSS's
for missing values, apart from the menu system, is less complex than the free
software coming out of Penn States Human Development Department. What about
charging extra for a power analysis module that will handle much of Cohen's
suppose they will charge extra for the replicate samping module when that is
available. For anybody who wants to do analysis on large scale survey research,
they must have some software for missing values, some software for replicate
sampling, and some software for power analysis. Unless we want to be ripped for
dollars and educational institution doesn't have, we have to use all kinds of
specialized software. This may be okay for us old xxxs, but it is a pain for
students. I hope SPSS makes enough money to continue investing in software
development, but I would hope they try to do this by offering an integrated
America's auto industry has paid a price for charging extra for adequate tires,
engines, paint, seats, etc. SPSS's efforts to segment the market and gouge
segment reminds me of the old days in the American auto industry.
Dick Campbell wrote:
> I have paid close attention to the various discussions of the new SPSS
> output, ascii conversions thereof and related topics. I would like to
> put in my two cents worth. I hereby certify that these comments come from
> an old (fill in the blank) who used SPSS on an IBM 360, who has taught
> hundreds of students to use the product (including some folks who are now
> working for SPSS) and who was computing at a time when an IBM 1410 was
> considered hot stuff. I happily admit to being out of date when it comes to
> modern computing. I don't know what "object oriented" means, I don't deal
> with OLE, OCDBCBD (or whatever that stuff is) and I am a very old dog sick
> of new tricks. All I want to do is get the numbers on to a sheet of paper in
> the most efficient way possible. So, could someone please answer the
> 1. What is the advantage of putting simple regression output in a pivot
> table? Yes, I know what a pivot table is, and I can see its purpose
> in complex tabs. But regression? Anyone pivoted an ANOVA table lately?
> 2. If Ver. 8 is going to have ASCII output, will the various parts of
> regression output (and related procedures) still be scattered all over
> the landscape in separate tables, pivot or otherwise?
> 3. Will there be a means of producing output -- ascii or otherwise -- which
> not cluttered up with special characters which have to be edited out?
> 4. Regardless of the answers to the above, will someone write an accessible
> introduction to how the new SPSS output system works? The discussion in
> the SPSS manuals that I have read is terse to say the least. Its not
> to just describe what the various icons do -- you have discuss how one
> might actually use the stuff. Yes, I have looked at the tutorials. They
> don't cut it.
> I raise this last question because my major complaint about the new SPSS
> system is that it adds a new layer of difficulty to teaching students how to
> use the system. I now have to teach students how to handle output,
> how to get it into a word processor if they want to embed tables of various
> kinds in papers, when what I want to teach is statistics.
> Having raised these points, let me just say that I don't see the point of
> releasing a version of a product which requires all previous users to climb
> a steep learning curve just to get the damned output on the printed page in
> a format they can use. The old methods ought to be available for those of
> us who are too stupid, lazy or stubborn to learn the new stuff. If I have to
> learn new stuff, then I am seriously inclined to move to a stat package that
> also offers me an array of modern statistics and graphics such as S-Plus.
> Then at least there is some payoff for me after I have invested the time and
> effort to learn the new product.
> As long as I am on the soapbox, why is the new SPSS missing data package
> not part of the base? Here is a chance for SPSS to allow students to learn
> something about missing data besides listwise deletion. It does no good
> at all to market the package as an add-on if SPSS is interested in
> to improvements in statistical practice and the teaching of statistics.
> The fact is that there are now a couple of stand alone missing data packages
> that are better than the SPSS product so that if someone is going to buy
> a missing data machine, it is unlikely to be SPSS's anyway. So why not make
> it available to those of us who teach stats so at least we can educate our
> students about modern methods?
> Richard T. Campbell | E-Mail : DCAMP@UIC.EDU
> Department of Sociology M/C 312 | Phone (W): 312/413-3759
> University of Illinois at Chicago | Phone (H): 708/386-2263
> 1007 W. Harrison St. | FAX (SOC): 312/996-5104
> Chicago, IL USA 60607-7140 | FAX (PRC): 312/996-2703