DON'T Use spreadsheets like Excel for much stat. A soapbox post.
Locked
DON'T Use spreadsheets like Excel for much stat. A soapbox post.
 
|
This is one of my soapbox posts, apologies to those of you who are
tired of my making this point.
Using spreadsheets for stat can be like using a hammer to drive a screw. IF I RECALL CORRECTLY any accounting system that uses computer spreadsheets cannot pass ISO certification. I do use Excel or Quattro Pro or tables in WordPerfect several times a week. Sometimes it is practical to enter data via a spreadsheet and then read it into SPSS for checking double entry or quality checks. [This is because spread sheets are usually available, but it not cost effective to provide SPSS to every person who will do data entry.] For an overview of the issues A video can be found at http://www.spss.com/events/event.cfm?E_ID=2921&Country=US A pdf can be found by clicking <The Risks of Using Spreadsheets in Data Analysis> For more technical review see. http://www.pages.drexel.edu/~bdm25/excel-intro.pdf http://www.pages.drexel.edu/~bdm25/excel2007.pdf http://www.pages.drexel.edu/~bdm25/excel-rng.pdf Art Kendall Social Research Consultants ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD
Art Kendall
Social Research Consultants |
Re: DON'T Use spreadsheets like Excel for much stat. A soapbox post.
Administrator
|
Thanks for the post Art!
Sounds rather incompetent/irresponsible on the part of M$! Now I wonder if Open Source projects such as Open Office and NeoOffice have similar issues or whether they actually have competent statistical programmers involved with their implementation?
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me. --- "Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis." Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?" |
Re: DON'T Use spreadsheets like Excel for much stat. A soapbox post.
|
In reply to this post by Art Kendall
Hi Art, We talked about this before but I agree with you that your point *cannot* be emphasized enough. Here's a very interesting link: http://www.eusprig.org/ . It contains one paper about average cell error rates (CERs) in excel sheets. Albert-Jan ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
All right, but apart from the sanitation, the medicine, education, wine, public order, irrigation, roads, a fresh water system, and public health, what have the Romans ever done for us? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From: Art Kendall <[hidden email]> To: [hidden email] Sent: Sat, December 4, 2010 3:10:57 PM Subject: [SPSSX-L] DON'T Use spreadsheets like Excel for much stat. A soapbox post. This is one of my soapbox posts, apologies to those of you who are tired of my making this point. Using spreadsheets for stat can be like using a hammer to drive a screw. IF I RECALL CORRECTLY any accounting system that uses computer spreadsheets cannot pass ISO certification. I do use Excel or Quattro Pro or tables in WordPerfect several times a week. Sometimes it is practical to enter data via a spreadsheet and then read it into SPSS for checking double entry or quality checks. [This is because spread sheets are usually available, but it not cost effective to provide SPSS to every person who will do data entry.] For an overview of the issues A video can be found at http://www.spss.com/events/event.cfm?E_ID=2921&Country=US A pdf can be found by clicking <The Risks of Using Spreadsheets in Data Analysis> For more technical review see. http://www.pages.drexel.edu/~bdm25/excel-intro.pdf http://www.pages.drexel.edu/~bdm25/excel2007.pdf http://www.pages.drexel.edu/~bdm25/excel-rng.pdf Art Kendall Social Research Consultants ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
Re: DON'T Use spreadsheets like Excel for much stat. A soapbox post.
|
|
In reply to this post by Art Kendall
Dear Art, This issue looks very interesting and important, and well worthy of getting up onto your soapbox. However, when I tried clicking on the link "The Risks of Using..." I had to log in with address, phone number, etc. to an IBM site, which then took me to Cognos, which seemed completely ignorant of the topic, and brought back many apparently-irrelevant hits when I searched on "risks of using spreadsheets." Do you have a more direct link? Thanks very much! Allan At 09:10 AM 12/4/2010, you wrote: This is one of my soapbox posts, apologies to those of you who are tired of my making this point. Research Consulting [hidden email] Business & Cell (any time): 215-820-8100 Home (8am-10pm, 7 days/week): 215-885-5313 Address: 108 Cliff Terrace, Wyncote, PA 19095 Visit my Web site at www.dissertationconsulting.net ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
The issues involved in using Excel or any statistical software
is an old one and
goes back to questions of whether mainframe computers
accurately calculated
statistics. An important paper in this regard is Leland
Wilkinson (creator of the
software package Systat) and Gerald Dallal's 1977 American
Statistician's article
that showed that several of the packages
(including SPSS) failed to produce
accurate statistics because of how they were programmed (e.g.,
using single
precision instead of double precision representation; using
the "desk calculator"
algorthim, i.e., calculating sums and sums of sqaured values,
instead of the
deviation from the mean formula). This article can be
accessed on www.jstor.org:
see:
PDF copies may be floating around on the internet.
Responsbile statistical
software companies heed these kinds of problems and fix them
(Microsoft
appears not to fall into this category).
Wilkinson built upon this work, creating a battery of tests
for statistical software
known as the "Statistics Quiz". A paper introducing this
battey is available here:
Google and other types of database search will turn up related
publications and
application of the statistics quiz.
Bruce McCullough of Drexel U has done a fair amount of work in
this area with
respect to accuracy of Excel and Art
Kendall provides several links to his papers
on his webpage. A search for name will turn up more
papers. McCullough's
webpage does contain a list of his publications, some of which
can be directly
accessed (such as Art's links below); see:
For fans of econometrics (I'm not much of one myself) and TSP
package,
here is a page that provides various benchmarks for different
types of
data; see:
-Mike Palij
New York University
|
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |