John,
Way back in time, around 1977-78, at the first SPSS users group meeting in
Chicago, I remember at how surprised Norman Nie and others were that most
users were using SPSS for data management needs, and not for statistical
procedures. I believe the most heavily used statistical procedures at that
time, based on a quick survey they had done, were Frequencies, Crosstabs and
Breakdown. Certainly my use of SPSS over the years involved mostly taking
a data set of some structure, doing Recodes/Computes, and then Exporting the
data into a new format that would be used later for more in-depth
statistical analysis, say a Regression procedure or Correlation analysis.
For example, one of the biggest things I did with SPSS was to take a raw
data file, and use Computes and Recodes to clean the data set up, use
Aggregate to create summary totals, and Export the data to be used by
spreadsheet software. SPSS, at least to me, had a syntax that made it
easy for an analyst to take a raw data file and transform it into what they
ultimately wanted to do. When SPSS added the Mult Response command set,
and then followed that up with the Complex Files procedures, a user really
had a very powerful data management tool set available that could really
begin to tease out data relationships that prior to this had required
expensive custom computer programming to get similar results.
Obviously, the statistical procedures are important. But for those who
are analyzing large amounts of data, the restructuring or reorganization of
the data to get to the point of using a stat procedure are where analysts
spend a large amount of time. Anything that can make that task easier is
viewed as a big plus by most analysts.
Marshall
-----Original Message-----
From: address@hidden
[mailto:address@hidden On Behalf Of John
Darrington
Sent: Wednesday, May 11, 2011 3:59 AM
To: Roman Seidl
Cc: address@hidden
Subject: Re: Import / Export Functions and other features missing for
academic and office use in urban planning
On Tue, May 10, 2011 at 06:36:05PM +0200, Roman Seidl wrote:
I was confused when mentioning ODT - which seem to work for the output
(which is certainly very nice to have).
I was actually thinking about ODS, XLS or DBF as a means to import and
export the current dataset.
I could not find out how to import ODS in 0.7.7 though Wikipedia (well
Wikipedia...) claims it should be possible (s.
http://en.wikipedia.org/wiki/PSPP, last modified on 30 January 2011 at
04:03).
The only way to get Data out of PSPP for now seems to be exporting to a
text file which is a quite tedious job as the docs seem to be rather
bad
and there is no such thing as a wizard.
I'm rather interested to know why you place so much importance on exporting
the data. I've always considered PSPP (and spss) as a tool to analyse data,
not
(primarily) to process it. Therefore, there are a number of ways in which
you can import data, and there are a number of formats to which you can
export the results of your analysis (ie, the tables showing T-test, Anova,
etc). However, like you say, exporting data hasn't been a feature upon
which the developers have placed great emphasis.
But the fact that you are wanting to export data from PSPP seems to suggest
that you're using it primarily for its data manipulation capabilities (eg,
SORT, FLIP, AGGREGATE etc) rather than for performing statistical analysis.
Can you give a typical workflow example of your usage? And what features do
you actually use within PSPP (between having got the data in, and getting it
back out again)?
J'
--
PGP Public key ID: 1024D/2DE827B3
fingerprint = 8797 A26D 0854 2EAB 0285 A290 8A67 719C 2DE8 27B3 See
http://pgp.mit.edu or any PGP keyserver for public key.
_______________________________________________
Pspp-users mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/pspp-users