# Stats Experts: Use Gnumeric, Donâ€™t Use Excel

Discussion in 'NZ Computing' started by Lawrence D'Oliveiro, Feb 21, 2010.

1. ### Lawrence D'OliveiroGuest

Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
the same calculation bugs. However, those have now been fixed, while Excel
remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.

Lawrence D'Oliveiro, Feb 21, 2010

2. ### SquiggleGuest

Re: Non-stats expert: use a stats tool for doing stats!

Lawrence D'Oliveiro threw some characters down the intarwebs:
> Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
> the same calculation bugs. However, those have now been fixed, while Excel
> remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.
>

Why would you use a spreadsheet to do statistical anaylysis? There are
tools designed for doing statistical analysis, SPSS, R and Matlab with
the stats toolkit all being good examples, and far better choices than
gnumeric or excel.

Using excel,gnumeric or any other general purpose spreadsheet for
statistical analysis sounds a bit like using a baseball bat to drive nails.

Squiggle, Feb 22, 2010

3. ### peterwnGuest

Re: Non-stats expert: use a stats tool for doing stats!

On Feb 22, 2:03 pm, Squiggle <> wrote:
> Lawrence D'Oliveiro threw some characters down the intarwebs:> Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
> > the same calculation bugs. However, those have now been fixed, while Excel
> > remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.

>
> Why would you use a spreadsheet to do statistical anaylysis?  There are
> tools designed for doing statistical analysis, SPSS, R and Matlab with
> the stats toolkit all being good examples, and far better choices than
> gnumeric or excel.
>
> Using excel,gnumeric or any other general purpose spreadsheet for
> statistical analysis sounds a bit like using a baseball bat to drive nails.

Hence the logical conclusion of your argument is that the Excel bugs
do not need fixing as people should be using specialist software for
the task in hand. What a cop-out!

Enough people seem to use statistical functions in spreadsheets as to
warrant their retention and in the case of Gnumeric to warrant
volunterers to get stuck in and fix the bugs.

peterwn, Feb 22, 2010
4. ### Gib BogleGuest

Re: Non-stats expert: use a stats tool for doing stats!

peterwn wrote:
> On Feb 22, 2:03 pm, Squiggle <> wrote:
>> Lawrence D'Oliveiro threw some characters down the intarwebs:> Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
>>> the same calculation bugs. However, those have now been fixed, while Excel
>>> remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.

>> Why would you use a spreadsheet to do statistical anaylysis? There are
>> tools designed for doing statistical analysis, SPSS, R and Matlab with
>> the stats toolkit all being good examples, and far better choices than
>> gnumeric or excel.
>>
>> Using excel,gnumeric or any other general purpose spreadsheet for
>> statistical analysis sounds a bit like using a baseball bat to drive nails.

>
> Hence the logical conclusion of your argument is that the Excel bugs
> do not need fixing as people should be using specialist software for

That is a non sequitur.

Gib Bogle, Feb 22, 2010
5. ### peterwnGuest

Re: Non-stats expert: use a stats tool for doing stats!

On Feb 22, 6:00 pm, Gib Bogle <> wrote:
> peterwn wrote:
> > On Feb 22, 2:03 pm, Squiggle <> wrote:
> >> Lawrence D'Oliveiro threw some characters down the intarwebs:> Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
> >>> the same calculation bugs. However, those have now been fixed, while Excel
> >>> remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.
> >> Why would you use a spreadsheet to do statistical anaylysis?  There are
> >> tools designed for doing statistical analysis, SPSS, R and Matlab with
> >> the stats toolkit all being good examples, and far better choices than
> >> gnumeric or excel.

>
> >> Using excel,gnumeric or any other general purpose spreadsheet for
> >> statistical analysis sounds a bit like using a baseball bat to drive nails.

>
> > Hence the logical conclusion of your argument is that the Excel bugs
> > do not need fixing as people should be using specialist software for
> > the task in hand.

>
> That is a non sequitur.

Of course. If spreadsheet programs have statistical functions for
whaever reason, or lack of, they should operate correctly.

peterwn, Feb 22, 2010
6. ### Gib BogleGuest

Re: Non-stats expert: use a stats tool for doing stats!

peterwn wrote:
> On Feb 22, 6:00 pm, Gib Bogle <> wrote:
>> peterwn wrote:
>>> On Feb 22, 2:03 pm, Squiggle <> wrote:
>>>> Lawrence D'Oliveiro threw some characters down the intarwebs:> Interesting. Gnumeric initially cloned Excel so exactly that it reproduced
>>>>> the same calculation bugs. However, those have now been fixed, while Excel
>>>>> remains as buggy as ever <http://www.csdassn.org/reportdetail.cfm?ID=508>.
>>>> Why would you use a spreadsheet to do statistical anaylysis? There are
>>>> tools designed for doing statistical analysis, SPSS, R and Matlab with
>>>> the stats toolkit all being good examples, and far better choices than
>>>> gnumeric or excel.
>>>> Using excel,gnumeric or any other general purpose spreadsheet for
>>>> statistical analysis sounds a bit like using a baseball bat to drive nails.
>>> Hence the logical conclusion of your argument is that the Excel bugs
>>> do not need fixing as people should be using specialist software for

>> That is a non sequitur.

>
> Of course. If spreadsheet programs have statistical functions for
> whaever reason, or lack of, they should operate correctly.

Plus willful misconstruing. As you well know, he didn't say anything about the
need to fix bugs in Excel, and to claim such a "logical conclusion" is illogical.

Gib Bogle, Feb 22, 2010
7. ### peterwnGuest

Re: Non-stats expert: use a stats tool for doing stats!

On Feb 23, 9:12 am, Gib Bogle <> wrote:

>
> > Of course.  If spreadsheet programs have statistical functions for
> > whaever reason, or lack of,  they should operate correctly.

>
> Plus willful misconstruing.  As you well know, he didn't say anything about the
> need to fix bugs in Excel, and to claim such a "logical conclusion" is illogical.

Which begs the question why he made an off topic comment to start
with. I construed his comment to mean "you might as well stick eith
Excel, as no serious statistics person uses spreadsheets for
statistics" and drew my conclusions from this.

peterwn, Feb 22, 2010
8. ### Gib BogleGuest

Re: Non-stats expert: use a stats tool for doing stats!

peterwn wrote:
> On Feb 23, 9:12 am, Gib Bogle <> wrote:
>
>>> Of course. If spreadsheet programs have statistical functions for
>>> whaever reason, or lack of, they should operate correctly.

>> Plus willful misconstruing. As you well know, he didn't say anything about the
>> need to fix bugs in Excel, and to claim such a "logical conclusion" is illogical.

>
> Which begs the question why he made an off topic comment to start
> with. I construed his comment to mean "you might as well stick eith
> Excel, as no serious statistics person uses spreadsheets for
> statistics" and drew my conclusions from this.
>

Eh? He said this "Using excel,gnumeric or any other general purpose spreadsheet
for statistical analysis sounds a bit like using a baseball bat to drive nails."
Does this sound like advocating sticking with Excel? Not to me.

Gib Bogle, Feb 22, 2010
9. ### Lawrence D'OliveiroGuest

Re: Non-stats expert: use a stats tool for doing stats!

In message <hluvm4\$qrc\$>, Gib Bogle wrote:

> He said this "Using excel,gnumeric or any other general purpose
> spreadsheet for statistical analysis sounds a bit like using a baseball
> bat to drive nails."
> Does this sound like advocating sticking with Excel? Not to me.

But there was a clear implication that the bugs somehow didnâ€™t matter.

Lawrence D'Oliveiro, Feb 22, 2010
10. ### Gib BogleGuest

Re: Non-stats expert: use a stats tool for doing stats!

Lawrence D'Oliveiro wrote:
> In message <hluvm4\$qrc\$>, Gib Bogle wrote:
>
>> He said this "Using excel,gnumeric or any other general purpose
>> spreadsheet for statistical analysis sounds a bit like using a baseball
>> bat to drive nails."
>> Does this sound like advocating sticking with Excel? Not to me.

>
> But there was a clear implication that the bugs somehow didnâ€™t matter.

Not to me. The implication to me was that only a fool would use Excel for stats.

Gib Bogle, Feb 23, 2010
11. ### Lawrence D'OliveiroGuest

Re: Non-stats expert: use a stats tool for doing stats!

In message <hlvm31\$vfk\$>, Gib Bogle wrote:

> Lawrence D'Oliveiro wrote:
>
>> But there was a clear implication that the bugs somehow didnâ€™t matter.

>
> Not to me. The implication to me was that only a fool would use Excel for
> stats.

But what would the relevance be of that, then?

Lawrence D'Oliveiro, Feb 23, 2010
12. ### peterwnGuest

Re: Non-stats expert: use a stats tool for doing stats!

On Feb 23, 5:37 pm, Gib Bogle <> wrote:
> Lawrence D'Oliveiro wrote:
> > In message <hluvm4\$>, Gib Bogle wrote:

>
> >> He said this "Using excel,gnumeric or any other general purpose
> >> spreadsheet for statistical analysis sounds a bit like using a baseball
> >> bat to drive nails."
> >>   Does this sound like advocating sticking with Excel?  Not to me.

>
> > But there was a clear implication that the bugs somehow didn’t matter..

>
> Not to me.  The implication to me was that only a fool would use Excel for stats.

Since allegedly one is born every minute, there would be millions of
people out there who would potentially use spreadsheet software for
stats.

peterwn, Feb 23, 2010