How capitalize "eBay" or "iPhone" at the start of a sentence?

Jennifer Murphy

unread,

Jan 14, 2014, 8:35:44 AM1/14/14

to

Sentences like "eBay had a good quarter." or "iPhone sales are down."
look odd to me, but I am reluctant to change the spelling of a proper
name. I agree with this website that writing "Ebay (or EBay) had a good
quarter." is like changing Sallee's name to "Sally".

http://english.stackexchange.com/questions/9063/how-do-you-capitalize-a-proper-noun-such-as-iphone

Don't eBay and Sallee have a right to determine how their names are
spelled, including capitalization?

Bertel Lund Hansen

unread,

Jan 14, 2014, 9:01:11 AM1/14/14

to

Jennifer Murphy skrev:

> Sentences like "eBay had a good quarter." or "iPhone sales are down."
> look odd to me, but I am reluctant to change the spelling of a proper
> name. I agree with this website that writing "Ebay (or EBay) had a good
> quarter." is like changing Sallee's name to "Sally".

The companies have created the problem themselves by breaking a
very old tradition. They have to suffer the consequences.

Changing "eBay" to "EBay" is no worse than changing "well" to
"Well".

--
Bertel, Denmark

Peter T. Daniels

unread,

Jan 14, 2014, 9:09:13 AM1/14/14

to

That depends on exactly what the registered trademark is. Presumably
the initial lowercase letter is a definitive part of the sequence.

CDB

unread,

Jan 14, 2014, 9:20:56 AM1/14/14

to

Sure. Maybe that means you won't be able to use them in certain
contexts. Sales of iPhones are down.

I don't think Sallee's case is comparable; I can't think of any context
that would force you to change final "ee" to "-y" (a 141-character tweet
can be dealt with in some other way).

Message has been deleted

Jerry Friedman

unread,

Jan 14, 2014, 9:42:13 AM1/14/14

to

On 1/14/14 7:01 AM, Bertel Lund Hansen wrote:
> Jennifer Murphy skrev:
>
>> Sentences like "eBay had a good quarter." or "iPhone sales are down."
>> look odd to me, but I am reluctant to change the spelling of a proper
>> name. I agree with this website that writing "Ebay (or EBay) had a good
>> quarter." is like changing Sallee's name to "Sally".

I don't agree.

> The companies have created the problem themselves by breaking a
> very old tradition. They have to suffer the consequences.

I agree, when you can't use the better method of keeping the name out of
the beginnings of sentences. "The quarter went well for eBay." "The
sales of iPhones are down."

I hope people who don't capitalize their on-line monikers don't mind if
they get capitalized at the beginnings of sentences. Micky may have an
opinion on this.

> Changing "eBay" to "EBay" is no worse than changing "well" to
> "Well".

Maybe a little worse.

--
Jerry Friedman

Paul Wolff

unread,

Jan 14, 2014, 9:50:04 AM1/14/14

to

On Tue, 14 Jan 2014, Peter T. Daniels <gram...@verizon.net> posted:

Trademarks don't govern ordinary discourse.

Apart from that, companies aren't consistent in handling their own
names. The British ebay web site displays the ebay name in consecutive
lower case letters of red, blue, orange and green. The trading company
in Europe is advertised by them as eBay Europe S.� r.l; company
registration offices often have their own rules for capitalisation, or
ignoring it. Companies House, the company registration authority for
England and Wales, doesn't recognise lower case as far as I recall, but
makes all letters upper case, however they are submitted by the
registrant. Whether or not Ebay Limited is connected to the shopping
site, the name is shown all in caps on the CH namecheck web site.

--
Paul

Jennifer Murphy

unread,

Jan 14, 2014, 10:24:22 AM1/14/14

to

On Tue, 14 Jan 2014 15:01:11 +0100, Bertel Lund Hansen
<kanon...@lundhansen.dk> wrote:

>Jennifer Murphy skrev:
>
>> Sentences like "eBay had a good quarter." or "iPhone sales are down."
>> look odd to me, but I am reluctant to change the spelling of a proper
>> name. I agree with this website that writing "Ebay (or EBay) had a good
>> quarter." is like changing Sallee's name to "Sally".
>
>The companies have created the problem themselves by breaking a
>very old tradition. They have to suffer the consequences.

What are those consequences? (Sounds punitive.)

>Changing "eBay" to "EBay" is no worse than changing "well" to
>"Well".

That seems like a weaker analogy than Sallee/Sally.

Bertel Lund Hansen

unread,

Jan 14, 2014, 11:05:07 AM1/14/14

to

Jennifer Murphy skrev:

>>The companies have created the problem themselves by breaking a
>>very old tradition. They have to suffer the consequences.

> What are those consequences? (Sounds punitive.)

That I write "Ebay" if the word is the first in a sentence.

>>Changing "eBay" to "EBay" is no worse than changing "well" to
>>"Well".

> That seems like a weaker analogy than Sallee/Sally.

It seems to me more precise since my example has to do with the
first letter. The Sally-comparison seems far-fetched.

--
Bertel, Denmark

Nathan Sanders

unread,

Jan 14, 2014, 3:50:47 PM1/14/14

to

A better comparison would be with the plural of a brand name that ends
in -y, like "BlackBerry". Neither (1) nor (2) really look right to me,
though I style guides generally favor (2):

(1) He owns a lot of BlackBerries.
(2) He owns a lot of BlackBerrys.

Nathan

--
Department of Linguistics
Swarthmore College
http://sanders.phonologist.org

R H Draney

unread,

Jan 14, 2014, 4:15:01 PM1/14/14

to

Jennifer Murphy filted:

How do you feel about:

"Mg" is the abbreviation for "milligram".

?...r

--
Me? Sarcastic?
Yeah, right.

Message has been deleted

Garrett Wollman

unread,

Jan 14, 2014, 4:57:36 PM1/14/14

to

In article <lb49c...@drn.newsguy.com>,

R H Draney <dado...@spamcop.net> wrote:

>How do you feel about:
>
> "Mg" is the abbreviation for "milligram".
>
>?...r

Wrongity wrong wrong wrong.

"mg" is the unit symbol for "milligram". It is not an abbreviation.

-GAWollman

--
Garrett A. Wollman | What intellectual phenomenon can be older, or more oft
wol...@bimajority.org| repeated, than the story of a large research program
Opinions not shared by| that impaled itself upon a false central assumption
my employers. | accepted by all practitioners? - S.J. Gould, 1993

Peter Duncanson [BrE]

unread,

Jan 14, 2014, 6:34:07 PM1/14/14

to

On 14 Jan 2014 21:21:56 GMT, r...@zedat.fu-berlin.de (Stefan Ram) wrote:

>R H Draney <dado...@spamcop.net> writes:
>>"Mg" is the abbreviation for "milligram".
>

> In such a case, I use a restrictive apposition:
>
> »The shortening "mg" is the abbreviation for "milligram".«
> »The keyword "int" was derived from the english word "integer".«
> »The brand "xAmple" was just made up by me.«

There is a British musician who has the professional name "Example".
This derives from the initials of his first and last names: E.G.

--
Peter Duncanson, UK
(in alt.usage.english)

Bertel Lund Hansen

unread,

Jan 14, 2014, 7:04:48 PM1/14/14

to

Nathan Sanders skrev:

> (2) He owns a lot of BlackBerrys.

Shouldn't that be "BlackBerry's"?

--
Bertel, Denmark

Message has been deleted

Bertel Lund Hansen

unread,

Jan 14, 2014, 7:08:08 PM1/14/14

to

R H Draney skrev:

> How do you feel about:

> "Mg" is the abbreviation for "milligram".

I don't say that I like writing "EBay", but that's what I would
do. Likewise with your example if I ever needed such a sentence.

--
Bertel, Denmark

Message has been deleted

Robert Bannister

unread,

Jan 14, 2014, 7:23:30 PM1/14/14

to

It would have been aesthetically more pleasing had such spellings never
been adopted in the first place, but they were and so we must learn to
live with the fact.

--
Robert Bannister - 1940-71 SE England
1972-now W Australia

Robert Bannister

unread,

Jan 14, 2014, 7:31:40 PM1/14/14

to

On 14/01/2014 10:37 pm, Stefan Ram wrote:

> I'd say often the style is best that is most unobtrusive
> and natural. The readers should not be distracted by an
> eye-catching spelling. So usually, one should go with the
> most common spelling. This can be found using fulltext
> web search engines.

So are you saying that if a retail grocery calls itself "Finefayre" or
"FineFair", we should ignore their trade mark name and call them
"Finefare"? I don't think that will work.

Message has been deleted

Dr Nick

unread,

Jan 14, 2014, 8:07:45 PM1/14/14

to

You try being part of a family with something that works that way as
your surname, sunshine.

Nick, one of the Attys/Atties - I'm for the former BTW.

Garrett Wollman

unread,

Jan 14, 2014, 8:41:35 PM1/14/14

to

In article <mg-symbol-20...@ram.dialup.fu-berlin.de>,
Stefan Ram <r...@zedat.fu-berlin.de> wrote:

>wol...@bimajority.org (Garrett Wollman) writes:
>>"mg" is the unit symbol for "milligram". It is not an abbreviation.
>

> The unit symbol is given on the line marked U below.
> The character sequence »mg« is given on the line marked A.
>
>U ㎎
>A mg

The SI unit symbols are defined in ISO 31 (or its successor, ISO/IEC
80000), not ISO 10646.

Jennifer Murphy

unread,

Jan 14, 2014, 9:21:32 PM1/14/14

to

I'm pretty sure I've seem both "Grammys" and "Grammies" and similarly
for "Tony" and "Emmy". And despite the advice of some websites, I
believe the correct form is "Grammys".

Peter T. Daniels

unread,

Jan 14, 2014, 11:44:42 PM1/14/14

to

On Tuesday, January 14, 2014 7:04:48 PM UTC-5, Bertel Lund Hansen wrote:
> Nathan Sanders skrev:

> > (2) He owns a lot of BlackBerrys.

>
> Shouldn't that be "BlackBerry's"?

That's German.

R H Draney

unread,

Jan 15, 2014, 12:00:58 AM1/15/14

to

Peter Duncanson [BrE]" <ma...@peterduncanson.net> filted:

I used to work with a guy named Xavier Escamilla, nicknamed "Ecksy"...it's much
to late to ask him if it was a diminutive of his given-name initial or a
phonetic rendering of first and last initial together....r

Message has been deleted

Reinhold {Rey} Aman

unread,

Jan 15, 2014, 12:07:57 AM1/15/14

to

PeteY Daniels wrote:
>>>
>>> Shouldn't that be "BlackBerry's"?
>
> That's German.
>

Wrong.

--
~~~ Reinhold {Rey} Aman ~~~

Peter T. Daniels

unread,

Jan 15, 2014, 12:14:42 AM1/15/14

to

On Wednesday, January 15, 2014 12:07:57 AM UTC-5, Reinhold {Rey} Aman wrote:
> PeteY Daniels wrote:

> >>> Shouldn't that be "BlackBerry's"?
> > That's German.
>
> Wrong.

Liar.

Apteryx

unread,

Jan 15, 2014, 12:35:16 AM1/15/14

to

Sallee/Sally is a weak analogy in itself - it is just flouting Sallee's
preference for no reason.

Changing eBay to EBay (or Ebay) when at the start of a sentence would be
like changing sAlly to SAlly (or Sally), but only at the start of a
sentence, if Sallee had chosen to change her name to sAlly. Actually
there is a stronger reason to do it in the case of eBay - the chances of
their feelings being hurt are negligible.

aPteryx

Reinhold {Rey} Aman

unread,

Jan 15, 2014, 12:51:33 AM1/15/14

to

Ignorant PeteY Daniels wrote:
>
> Reinhold {Rey} Aman wrote:
>> PeteY Daniels wrote:
>
>>>>> Shouldn't that be "BlackBerry's"?
>>>
>>> That's German.
>>>
>> Wrong.
>>
> Liar.
>

You ignorant asshole. Say "Hi" to Bertel "sch" Hansen, who also doesn't
know the difference between German and Dutch.

Garrett Wollman

unread,

Jan 15, 2014, 1:31:27 AM1/15/14

to

In article <lb56mj$9vh$1...@speranza.aioe.org>,

Apteryx <apt...@xtra.co.nz> wrote:

>Changing eBay to EBay (or Ebay) when at the start of a sentence would be
>like changing sAlly to SAlly (or Sally), but only at the start of a
>sentence, if Sallee had chosen to change her name to sAlly. Actually
>there is a stronger reason to do it in the case of eBay - the chances of
>their feelings being hurt are negligible.

At least in the U.S., trademark registrations for "words" are folded
to uppercase. A registrant can claim a specific typographic
presentation (font, capitalization, colors, etc.) by providing a
specimen, but most filings simply include a "typed drawing", which is
nothing more than the word mark being claimed written in capital
letters. Such a registration covers the word in all typefaces,
capitalizations, etc., although some companies -- particularly drug
companies for some reason -- seem to believe that they are bound from
then on to write their brand name in all-caps-and-a-registered-sign
on every single use.

The word mark "EBAY" was registered by "eBay Inc. CORPORATION
CALIFORNIA 2005 Hamilton Avenue, Suite 350 San Jose CALIFORNIA 95125"
and is now owned by "EBAY INC. CORPORATION DELAWARE 2145 HAMILTON
AVENUE SAN JOSE CALIFORNIA 95125". The precise use claimed at that
time was "IC 035. US 100 101 102. G & S: on-line trading services in
which seller posts items to be auctioned and bidding is done
electronically, and providing evaluative feedback and ratings of
sellers' goods and services, the value and prices of sellers' goods,
buyers' and sellers' performance, delivery, and overall trading
experience in connection therewith"; they have since filed numerous
registrations for other fields of commerce and for other typographic
presentations, most of which show the name in all lower case, and
several of which claim the specific colors.

As to the basic issue, I feel the same way about "eBay" as I do about
"bell hooks": I don't care what the owner's preference is, when I'm
writing in English, I will capitalize them in all of the situations
where the normal rules of English call for that. They don't get a
choice about the rules of the language in which they are written
about.

Nathan Sanders

unread,

Jan 15, 2014, 2:44:16 AM1/15/14

to

On 2014-01-15 06:31:27 +0000, Garrett Wollman said:

> As to the basic issue, I feel the same way about "eBay" as I do about
> "bell hooks": I don't care what the owner's preference is, when I'm
> writing in English, I will capitalize them in all of the situations
> where the normal rules of English call for that. They don't get a
> choice about the rules of the language in which they are written
> about.

unread,

Jan 15, 2014, 6:33:38 AM1/15/14

to

On 15 Jan 2014 00:13:39 GMT, Stefan Ram wrote:
> Oops! It's possible that I might have messed up on the
> proper encoding declaration. But maybe one could still
> get the idea (there is a single Unicode mg symbol).
>

That seems wrong to me. The milligram is not an atomic unit like a
scruple; it is a composite unit. The "m" of mg is the same "m" as
the "m" of mV, ml, and the rest.

Surely there are not Unicode symbols for every single milli-unit? If
there are, what about kilo-, not to mention yocto-?

Stan Brown

unread,

Jan 15, 2014, 6:37:18 AM1/15/14

to

On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:
> How do you feel about sentences like:
>
> i is the only glyph in the English alphabet that is not fully connected.

I don't agree. What about j?

> p-values greater than 0.5 are highly insignificant.

I don't agree. Aside from the silliness of "highly insignificant", a
p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
and the same is true of any other measurement.

pensive hamster

unread,

Jan 15, 2014, 6:43:28 AM1/15/14

to

On Wednesday, 15 January 2014 07:44:16 UTC, Nathan Sanders wrote:

> How do you feel about sentences like:
>
> i is the only glyph in the English alphabet that is not fully connected.
> g, j, p, q, and y are the only English letters with descenders.
> "m" is the SI unit symbol for meter.
> "password" is a terrible password, in part because it is all lowercase.
> mRNA carries information from the DNA to the cytoplasm.
> iPod is a silly name because it begins with a lowercase letter.
> pH test strips are available upon request.
> p-values greater than 0.5 are highly insignificant.
> grep, sed, and awk are useful commands in Unix.
>
> Of course, you can almost always avoid the issue by rewriting the
> sentence (that's my solution), but for the sake of argument, assume
> that no such rearrangement is possible/allowed (maybe the title of an
> article, for example, where words like "mRNA" are known to appear in
> intial position).

You could just put the awkward word with the initial lower-case letter
inside quote marks, as you have done with your third and fourth example
sentences above.

Personally I wouldn't bother, if the word or name is well-known and
well-established, as in the case of iPod, e. e. cummings, or pH. Of course,
it is to some extent subjective whether a name is well-known or not.

Grammar should serve clarity of meaning. Starting a sentence with a word
which itself starts with lower-case letter should not be a problem, unless
it introduces some ambiguity or lack of clarity in the information you are
trying to convey.

Bertel Lund Hansen

unread,

Jan 15, 2014, 6:47:49 AM1/15/14

to

Stan Brown skrev:

>> Oops! It's possible that I might have messed up on the
>> proper encoding declaration. But maybe one could still
>> get the idea (there is a single Unicode mg symbol).

> That seems wrong to me.

I agree.

--
Bertel, Denmark

Leslie Danks

unread,

Jan 15, 2014, 7:28:16 AM1/15/14

to

Stan Brown wrote:

> On 14 Jan 2014 21:21:56 GMT, Stefan Ram wrote:
>>
>> R H Draney <dado...@spamcop.net> writes:
>> >"Mg" is the abbreviation for "milligram".
>>
>> In such a case, I use a restrictive apposition:
>>
>> »The shortening "mg" is the abbreviation for "milligram".«
>> »The keyword "int" was derived from the english word "integer".«
>> »The brand "xAmple" was just made up by me.«
>
> "Milligram" is abbreviated mg.

Some smart-aleck will then ask you whether it's "mg." or "mg".

--
Les (BrE)
I frog mi wos i do dua - STS

Peter T. Daniels

unread,

Jan 15, 2014, 9:00:06 AM1/15/14

to

On Wednesday, January 15, 2014 12:51:33 AM UTC-5, Reinhold {Rey} Aman lied:
> Peter T. Daniels wrote:
> > Reinhold {Rey} Aman lied:
> >> Peter T. Daniels wrote:

[no, he didn't]

> >>>>> Shouldn't that be "BlackBerry's"?
> >>> That's German.
> >> Wrong.
> > Liar.
>
> You ignorant asshole. Say "Hi" to Bertel "sch" Hansen, who also doesn't
> know the difference between German and Dutch.

Maybe I've read more German than you over the past 40 years. Less
familiar proper names are pluralized in German with apostrophe-s.

James Silverton

unread,

Jan 15, 2014, 9:26:28 AM1/15/14

to

On 1/15/2014 6:21 AM, Stan Brown wrote:
> On Tue, 14 Jan 2014 05:35:44 -0800, Jennifer Murphy wrote:
>>
>> Sentences like "eBay had a good quarter." or "iPhone sales are down."
>> look odd to me, but I am reluctant to change the spelling of a proper
>> name. I agree with this website that writing "Ebay (or EBay) had a good
>> quarter." is like changing Sallee's name to "Sally".
>>
>> http://english.stackexchange.com/questions/9063/how-do-you-capitalize-a-proper-noun-such-as-iphone
>>
>> Don't eBay and Sallee have a right to determine how their names are
>> spelled, including capitalization?
>
> This is not a new problem. What do you do with e. e. cummings at the
> beginning of a sentence?
>
> The answer I give, in such cases, is to rewrite the sentence so that
> the offending name is not the first word. Yes, I said "offending" --
> I find this interior-capital stuff annoying.
>

The various companies have chosen to be "cute" about spelling and I
don't feel any great need to follow. I think I would capitalize if the
word came first in a sentence. "IPads can be a nuisance" does look
strange but there is no doubt about the meaning.

--
Jim Silverton (Potomac, MD)

Extraneous "not." in Reply To.

Message has been deleted

Athel Cornish-Bowden

unread,

Jan 15, 2014, 11:20:04 AM1/15/14

to

Be thankful for small mercies, Rey. At least we've been spared a
statement of how BlackBerry is made plural in Danish.

--
athel

Athel Cornish-Bowden

unread,

Jan 15, 2014, 11:25:52 AM1/15/14

to

On 2014-01-14 22:57:36 +0100, wol...@bimajority.org (Garrett Wollman) said:

> In article <lb49c...@drn.newsguy.com>,
> R H Draney <dado...@spamcop.net> wrote:
>
>> How do you feel about:

>>
>> "Mg" is the abbreviation for "milligram".
>>

>> ?...r
>
> Wrongity wrong wrong wrong.
>
> "mg" is the unit symbol for "milligram". It is not an abbreviation.

I'm glad you said that. I was working through to the end of the thread
before saying it myself.

--
athel

Athel Cornish-Bowden

unread,

Jan 15, 2014, 11:27:15 AM1/15/14

to

On 2014-01-15 12:33:38 +0100, Stan Brown <the_sta...@fastmail.fm> said:

> On 15 Jan 2014 00:13:39 GMT, Stefan Ram wrote:
>> Oops! It's possible that I might have messed up on the
>> proper encoding declaration. But maybe one could still
>> get the idea (there is a single Unicode mg symbol).
>>
>
> That seems wrong to me. The milligram is not an atomic unit like a
> scruple; it is a composite unit. The "m" of mg is the same "m" as
> the "m" of mV, ml, and the rest.

Right.

>
> Surely there are not Unicode symbols for every single milli-unit? If
> there are, what about kilo-, not to mention yocto-?

--

athel

Athel Cornish-Bowden

unread,

Jan 15, 2014, 11:37:52 AM1/15/14

to

On 2014-01-15 12:37:18 +0100, Stan Brown <the_sta...@fastmail.fm> said:

> On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:
>> How do you feel about sentences like:
>>
>> i is the only glyph in the English alphabet that is not fully connected.
>
> I don't agree. What about j?

That's what I wondered.

>
>> p-values greater than 0.5 are highly insignificant.
>
> I don't agree. Aside from the silliness of "highly insignificant", a
> p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
> and the same is true of any other measurement.

Likewise. I think the sentence is badly expressed. I'd say something
like "p-values greater than 0.5 are conventionally consider to be not
significant." (Incidentally, in this context I don't regard "not
significant" as synonymous with "insignificant": the former is
declining to make a statement; the latter is making a definite
statement.)

Of all Nathan's examples the one that comes up most often in my writing
is pH: there I don't hesitate to put the lower-case p at the beginning
of a sentence. I wouldn't hesitate with mRNA either if I had much
occasion to write it.

--
athel

Nick Spalding

unread,

Jan 15, 2014, 12:04:51 PM1/15/14

to

Garrett Wollman wrote, in <lb4bsg$1621$1...@grapevine.csail.mit.edu>
on Tue, 14 Jan 2014 21:57:36 +0000 (UTC):

> In article <lb49c...@drn.newsguy.com>,
> R H Draney <dado...@spamcop.net> wrote:
>
> >How do you feel about:
> >
> > "Mg" is the abbreviation for "milligram".
> >
> >?...r
>
> Wrongity wrong wrong wrong.
>
> "mg" is the unit symbol for "milligram". It is not an abbreviation.

Surely it must have been started out as one, or did the compilers of the symbol
list just say – "mg", that's a combination we haven't used before, lets use it
for "milligram".
--
Nick Spalding
BrE/IrE

unread,

Jan 15, 2014, 4:55:04 PM1/15/14

to

Athel Cornish-Bowden filted:

ObHomerSimpson: Mmmm, blackberry Danish!...r

--
Me? Sarcastic?
Yeah, right.

R H Draney

unread,

Jan 15, 2014, 5:18:02 PM1/15/14

to

pensive hamster filted:

>
>Grammar should serve clarity of meaning. Starting a sentence with a word
>which itself starts with lower-case letter should not be a problem, unless
>it introduces some ambiguity or lack of clarity in the information you are
>trying to convey.

Mind you, it'd be a lot clearer if the full stop at the end of the preceding
sentence were followed by two spaces, as custom demands....r

R H Draney

unread,

Jan 15, 2014, 5:20:50 PM1/15/14

to

Peter Duncanson [BrE]" <ma...@peterduncanson.net> filted:

Didn't our Daniel have a lesser obsession with the proper internal
capitalization of the the name "Edward FitzGerald"?...r

Peter T. Daniels

unread,

Jan 15, 2014, 5:47:14 PM1/15/14

to

As custom demanded back in typewriter days.

Nathan Sanders

unread,

Jan 15, 2014, 7:03:20 PM1/15/14

to

On 2014-01-15 11:37:18 +0000, Stan Brown said:

> On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:
>> How do you feel about sentences like:
>>
>> i is the only glyph in the English alphabet that is not fully connected.
>
> I don't agree. What about j?

Oops! Definite oversight on my part.

>> p-values greater than 0.5 are highly insignificant.
>
> I don't agree. Aside from the silliness of "highly insignificant", a

It's no sillier than "highly significant":

"Most authors refer to *statistically significant* as P < 0.05 and
*statistically highly significant* as P < 0.001"
<http://www.statsdirect.co.uk/help/basics/pval.htm>

"When p value <= .01 -> the observed difference is 'highly significant'"
<http://www.sjsu.edu/faculty/gerstman/StatPrimer/hyp-test.pdf>

"In practice, many scientists use phrases like 'Significant', 'Highly
Significant', 'Extremely Significant', or 'Borderline Significant'"
<http://webpages.marshall.edu/~denvir/BMS617/Spring2012/Lecture6_2012-01-30.pdf>

And though "highly significant" is probably more common than "highly
insignificant", the latter is not unheard of:

"Highly insignificant values (P > 0.50) are indicated by red"
<http://www.talkorigins.org/faqs/comdesc/incongruent.html>

"Many are highly insignificant, with p-values above .5."
<http://www.elon.edu/docs/e-web/students/ipe/volumes/Jackson%202012.pdf>

"The coefficient for 3 x is now highly insignificant (p-value = 0.856)
as it should be"
<http://scholar.lib.vt.edu/theses/available/etd-07302003-183651/unrestricted/Chapter7.pdf>

Differences in degree of (in)significance can certainly be worth noting
when, for example, you are using a one particular multiple comparison
correction (say Bonferroni correction). If your p-value is very close
to the corrected alpha (that is, it is not "highly" (in)significant), a
different correction method might alter the significance of your
p-value, whereas a different correction method is unlikely to alter the
significance of a "highly" (in)significant p-value.

> p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
> and the same is true of any other measurement.

The given sentence makes no claims about p-values below 0.5.

Nathan

--
Department of Linguistics
Swarthmore College
http://sanders.phonologist.org

unread,

Jan 15, 2014, 7:48:54 PM1/15/14

to

Garrett Wollman:

>> "mg" is the unit symbol for "milligram". It is not an abbreviation.

Stefan Ram:
> The unit symbol is [Unicode character 338E].

No, as usual, Garrett is right. The symbol for the unit is the two
characters "mg"; this is defined in the SI standard. The character
338E is as a representation of this character sequence provided for
use in Chinese, Japanese, or Korean writing. See:

http://www.unicode.org/charts/PDF/U3300.pdf
--
Mark Brader | "Should array indices start at 0 or 1? My ecumenical
Toronto | compromise of 0.5 was rejected without, I thought,
m...@vex.net | proper consideration." -- Stan Kelly-Bootle

My text in this article is in the public domain.

Robert Bannister

unread,

Jan 15, 2014, 7:52:19 PM1/15/14

to

On 15/01/2014 7:27 pm, Stan Brown wrote:
> On Tue, 14 Jan 2014 18:21:32 -0800, Jennifer Murphy wrote:
>> I'm pretty sure I've seem both "Grammys" and "Grammies" and similarly
>> for "Tony" and "Emmy".
>>
>
> I don't think I've ever seen "Grammies", "Tonies", or "Emmies" in the
> context of awards.
>
> For that matter, even in ordinary proper names I think the rules are
> different. If our own Mr. Cooper were to meet Mr. Danza, the actor
> turned English teacher, they would be a pair of Tonys, not a pair of
> Tonies, would they not?
>

"Tonies" looks more natural to me, and I know I have seen it.
The old ad when we were still at school avoided the question by saying
"Which girl has the Toni?".

--
Robert Bannister - 1940-71 SE England
1972-now W Australia

Robert Bannister

unread,

Jan 15, 2014, 7:53:21 PM1/15/14

to

On 15/01/2014 1:07 pm, Stefan Ram wrote:
> R H Draney <dado...@spamcop.net> writes:

>> I used to work with a guy named Xavier Escamilla, nicknamed "Ecksy"...it's much
>> to late to ask him if it was a diminutive of his given-name initial or a
>> phonetic rendering of first and last initial together....r
>
> I could use �camera�, my initials being SLR.
>
Not "rifle"?

unread,

Jan 15, 2014, 8:14:52 PM1/15/14

to

On 15/01/14 22:21, Stan Brown wrote:

> The answer I give, in such cases, is to rewrite the sentence so that
> the offending name is not the first word. Yes, I said "offending" --
> I find this interior-capital stuff annoying.

Many MacDonalds have listened to you and given up camel case, but the
McDonalds have not followed suit.

Nathan Sanders

unread,

Jan 15, 2014, 8:16:32 PM1/15/14

to

On 2014-01-16 00:43:01 +0000, Stan Brown said:

> On Wed, 15 Jan 2014 17:37:52 +0100, Athel Cornish-Bowden wrote:
>>
>> On 2014-01-15 12:37:18 +0100, Stan Brown <the_sta...@fastmail.fm> said:
>>
>>> On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:
>>>> p-values greater than 0.5 are highly insignificant.
>>>
>>> I don't agree. Aside from the silliness of "highly insignificant", a
>>> p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
>>> and the same is true of any other measurement.
>>
>> Likewise. I think the sentence is badly expressed. I'd say something
>> like "p-values greater than 0.5 are conventionally consider to be not
>> significant." (Incidentally, in this context I don't regard "not
>> significant" as synonymous with "insignificant": the former is
>> declining to make a statement; the latter is making a definite
>> statement.)
>
> You made the same mistake that I did, initially. I saw "0.5" (50%)
> and read it as "0.05" (5%).
>
> But you're dead right about the difference between "not significant"
> and "insignificant". Statistics can never prove that there is no
> effect; it can only fail to determine whether there is one or not.
> It's frightening how many textbooks get that wrong.

My dictionaries give "not significant" as the definition of "insignificant":

<http://www.merriam-webster.com/dictionary/insignificant>

<http://ahdictionary.com/word/search.html?q=insignificant>

And in my experience, this is indeed how these expressions are used in
the wild by real statisticians, so it's not just lay usage.
"Statistically insignificant" is just a shorter way of saying "not
statistically significant". It's analogous to using "p is greater
0.05" instead of "p is not less than or equal to 0.05":

p not <= 0.05 : not statistically significant
::
p > 0.05 : statistically insignificant

Since "p not <= 0.05" and "p > 0.05" are equivalent, so are "not
statistically significant" and "statistically insignificant".

What do you think "statistically insignificant" (with respect to
p-values) could mean instead? And have you ever seen a statistician
use it that way?

Message has been deleted

Garrett Wollman

unread,

Jan 15, 2014, 8:59:08 PM1/15/14

to

In article <MPG.2d40f6257...@news.individual.net>,

Stan Brown <the_sta...@fastmail.fm> wrote:
>In the olden days when "word processing" did not exist, Typists were
>taught to put a double space between sentences. But even then,
>typographers did not. And now, when everything but Usenet is word
>processed, there's no reason to do it.

Unless you use a text editor to edit text, rather than one of these
bastardized "word processor" things. (I guess a word processor does
to words what a food processor does to food....)

-GAWollman

--
Garrett A. Wollman | What intellectual phenomenon can be older, or more oft
wol...@bimajority.org| repeated, than the story of a large research program
Opinions not shared by| that impaled itself upon a false central assumption
my employers. | accepted by all practitioners? - S.J. Gould, 1993

Peter Moylan

unread,

Jan 15, 2014, 11:11:11 PM1/15/14

to

On 16/01/14 12:06, Stefan Ram wrote:
> Robert Bannister <rob...@clubtelco.com> writes:

>> On 15/01/2014 1:07 pm, Stefan Ram wrote:
>>> R H Draney <dado...@spamcop.net> writes:
>>>> I used to work with a guy named Xavier Escamilla, nicknamed "Ecksy"...it's much
>>>> to late to ask him if it was a diminutive of his given-name initial or a
>>>> phonetic rendering of first and last initial together....r
>>> I could use �camera�, my initials being SLR.
>> Not "rifle"?
>

> Now, that I just spotted Unicode symbol U+33DB �sr�,
> I will omit the �L� and then use U+33DB as �my symbol�.
>
That might make your symbol invisible to those of us who haven't
installed Japanese support. At least, I think it's Japanese; Firefox
can't show me most of the symbols in that section.

Peter Moylan

unread,

Jan 15, 2014, 11:17:01 PM1/15/14

to

On 16/01/14 11:47, Stan Brown wrote:
> On 15 Jan 2014 14:18:02 -0800, R H Draney wrote:
>> Mind you, it'd be a lot clearer if the full stop at the end of the preceding
>> sentence were followed by two spaces, as custom demands....r
>
> I can't agree that "custom demands" that. Robert Bringhurst exploded
> that myth in /The Elements of Typographic Style/, 30+ years ago, and
> I'm sure he wasn't the first.
>
> In the olden days when "word processing" did not exist, Typists were
> taught to put a double space between sentences. But even then,
> typographers did not. And now, when everything but Usenet is word
> processed, there's no reason to do it.
>
> (Yet old habits die hard, as you'll discern from the preceding
> paragraph.)
>

For me the rule is clear enough. In proportionally-spaced text you type
a single space, and then the software converts that to approximately one
and a half spaces. (The exact amount depending on things like
justification.) In fixed-space fonts that option is not available, so
you should type two space characters.

Of course with Usenet you don't know what font the person reading your
article will be using, so you just have to toss a coin.

In practice, I leave the decision to my fingers. I might have to change
that, because my fingers now make more errors than they used to.

Bertel Lund Hansen

unread,

Jan 15, 2014, 11:36:26 PM1/15/14

to

Peter Moylan skrev:

> In practice, I leave the decision to my fingers. I might have to change
> that, because my fingers now make more errors than they used to.

Don't change anything.

When I moved into this apartment, I started placing my kitchen
utensils in the drawers in the kitchen. That worked well, and I
could remember where they all were. At some time later I decided
that the order was a bit illogical, so I rearranged the tools.
The result is that I now consistently open the wrong drawer. I
can't even trick myself, because if I stop midways and then
change my choice - I still get the wrong drawer.

I am kind of proud, because it proves that my memory works to
perfection ... sort of. But it's not very useful.

--
Bertel, Denmark

R H Draney

unread,

Jan 16, 2014, 1:50:29 AM1/16/14

to

Robert Bannister filted:

>
>On 16/01/2014 6:20 am, R H Draney wrote:
>> Peter Duncanson [BrE]" <ma...@peterduncanson.net> filted:
>>>

>>> I take the approach that the spellings of proper names, including
>>> registered business names, etc, are invariant.
>>
>> Didn't our Daniel have a lesser obsession with the proper internal
>> capitalization of the the name "Edward FitzGerald"?...r
>>
>>
>There're also all those Mac/Mc names.

And the "van der" names, and the "de la" names, and the "'t" names....r

Dr Nick

unread,

Jan 16, 2014, 2:26:13 AM1/16/14

to

There are some Dutchmen - at least - who would disagree with you.

It's fairly old-fashioned to refer to someone by their surname, but it
does still happen.

"Smith was accused of posting without a license".
"van Gogh never did it".

James Hogg

unread,

Jan 16, 2014, 3:08:35 AM1/16/14

to

I see that the Dutch Wiki article about van Gogh has no qualms about
starting a sentence with Van Gogh.

--
James

Peter Duncanson [BrE]

unread,

Jan 16, 2014, 6:06:07 AM1/16/14

to

On Thu, 16 Jan 2014 09:01:52 +0800, Robert Bannister

<rob...@clubtelco.com> wrote:

>On 16/01/2014 6:20 am, R H Draney wrote:
>> Peter Duncanson [BrE]" <ma...@peterduncanson.net> filted:
>>>
>>> On Wed, 15 Jan 2014 08:23:30 +0800, Robert Bannister
>>> <rob...@clubtelco.com> wrote:
>>>
>>>> On 14/01/2014 9:35 pm, Jennifer Murphy wrote:
>>>>> Sentences like "eBay had a good quarter." or "iPhone sales are down."
>>>>> look odd to me, but I am reluctant to change the spelling of a proper
>>>>> name. I agree with this website that writing "Ebay (or EBay) had a good
>>>>> quarter." is like changing Sallee's name to "Sally".
>>>>>
>>>>> http://english.stackexchange.com/questions/9063/how-do-you-capitalize-a-proper-noun-such-as-iphone
>>>>>
>>>>> Don't eBay and Sallee have a right to determine how their names are
>>>>> spelled, including capitalization?
>>>>>
>>>>
>>>> It would have been aesthetically more pleasing had such spellings never
>>>> been adopted in the first place, but they were and so we must learn to
>>>> live with the fact.
>>>
>>> Quite.
>>>
>>> I take the approach that the spellings of proper names, including
>>> registered business names, etc, are invariant.
>>
>> Didn't our Daniel have a lesser obsession with the proper internal
>> capitalization of the the name "Edward FitzGerald"?...r
>>
>>
>There're also all those Mac/Mc names.

And the O' names.

--
Peter Duncanson, UK
(in alt.usage.english)

J. J. Lodder

unread,

Jan 16, 2014, 7:40:34 AM1/16/14

to

Entirely correct.
'Van' should be capitalised at the beginning of a sentence.

Standard convention is to write 'Vincent van Gogh'
with a lowercase van, and Mr. Van Gogh with an upprcase V.

't Hooft is a genuine exception,

Jan

Jan

pensive hamster

unread,

Jan 16, 2014, 8:26:04 AM1/16/14

to

On Thursday, 16 January 2014 04:17:01 UTC, Peter Moylan wrote:
> On 16/01/14 11:47, Stan Brown wrote:
> > On 15 Jan 2014 14:18:02 -0800, R H Draney wrote:
>
> >> Mind you, it'd be a lot clearer if the full stop at the end of the preceding
> >> sentence were followed by two spaces, as custom demands....r
> >
> > I can't agree that "custom demands" that. Robert Bringhurst exploded
> > that myth in /The Elements of Typographic Style/, 30+ years ago, and
> > I'm sure he wasn't the first.
> >
> > In the olden days when "word processing" did not exist, Typists were
> > taught to put a double space between sentences. But even then,
> > typographers did not. And now, when everything but Usenet is word
> > processed, there's no reason to do it.
> >
> > (Yet old habits die hard, as you'll discern from the preceding
> > paragraph.)
> >
> For me the rule is clear enough. In proportionally-spaced text you type
> a single space, and then the software converts that to approximately one
> and a half spaces.

Interesting, I never knew that. I have done a bit of typography and graphic
design in the past (using FreeHand, and Fontographer a little bit), so I am
not completely ignorant of the subject.

I have sometimes manually adjusted kerning, or spacing between letters,
but I have done it by eye. I used to squint up my eyes so that everything
is a bit of a blur, and then sometimes tweak the letter spacing slightly so
that a line of text would appear an even shade of grey, with no lumpy bits.
Though with a slightly paler (wider spaced) gap between sentences.

But a double space between sentences seems visually excessive to me,
especially considering it comes after a full stop, which is quite a narrow
and short 'letter', more space than letter really. So effectively a double
space between sentences plus the miniscule full stop would amount to
about 2.75 spaces, visually speaking.

> (The exact amount depending on things like
> justification.) In fixed-space fonts that option is not available, so
> you should type two space characters.

I have rarely used a typewriter, and never use fixed-space fonts, I think
they look horrible.

I was a bit worried when RHD said earlier that 'custom demands'
double spacing between a full stop and the start of a new sentence, I
had never heard that. Trying it out, it just looks w r o n g to me.

But I use (proportionally-spaced) Arial, in Google Groups, in Firefox;
don't know what RHD or others use.

CDB

unread,

Jan 16, 2014, 8:34:56 AM1/16/14

to

Ah, so is the rule capitalisation if it begins the name as presented?
Beginning the sentence would be a subset of that.

> 't Hooft is a genuine exception,

"Vincent 't Hooft", "Mr 't Hooft", "'t Hooft has arrived."?

Nick Spalding

unread,

Jan 16, 2014, 9:15:09 AM1/16/14

to

pensive hamster wrote, in
<4e17c2ba-fd76-48eb...@googlegroups.com>
on Thu, 16 Jan 2014 05:26:04 -0800 (PST):

As does your single spacing to me.

> But I use (proportionally-spaced) Arial, in Google Groups, in Firefox;
> don't know what RHD or others use.
>
> > Of course with Usenet you don't know what font the person reading your
> > article will be using, so you just have to toss a coin.
> >
> > In practice, I leave the decision to my fingers. I might have to change
> > that, because my fingers now make more errors than they used to.

--
Nick Spalding
BrE/IrE

Anton Shepelev

unread,

Jan 16, 2014, 10:01:16 AM1/16/14

to

Nathan Sanders:

> How do you feel about sentences like:
>
> i is the only glyph in the English alphabet that is not fully connected.
> g, j, p, q, and y are the only English letters with descenders.
> "m" is the SI unit symbol for meter.
> "password" is a terrible password, in part because it is all lowercase.
> mRNA carries information from the DNA to the cytoplasm.
> iPod is a silly name because it begins with a lowercase letter.
> pH test strips are available upon request.

> p-values greater than 0.5 are highly insignificant.

> grep, sed, and awk are useful commands in Unix.

I fully agree with Nathan in this. The preference
should be made in the order of locality. Words with
a fixed captalization should retain it regardless of
their position in the senence, i.e word level over-
rides sentence-level. Words without a fixed captal-
ization should obey sentence-level rules. We al-
ready have this with personal names, which are al-
ways capitalized.

--
() ascii ribbon campaign - against html e-mail
/\ www.asciiribbon.org - against proprietary attachments

Anton Shepelev

unread,

Jan 16, 2014, 10:17:06 AM1/16/14

to

Stan Brown:

> [...]a p-value of 0.5001 is essentially the same

> as a p-value of 0.4999 -- and the same is true of
> any other measurement.

How can this be? Is a p-value not the probablity of
the hypothesis under test being wrong and the null
hypothesis being correct? In this case, the lower
the p-value the higher more significance of a posi-
tive result.

Peter Duncanson [BrE]

unread,

Jan 16, 2014, 10:23:47 AM1/16/14

to

In

't Hooft has arrived

the ' is an upper-case ' which just happens to look the same as a
lower-case '.

Anton Shepelev

unread,

Jan 16, 2014, 10:41:40 AM1/16/14

to

pensive hamster:

> Interesting, I never knew that. I have done a bit

> of typography and graphic design in the past (us-

> ing FreeHand, and Fontographer a little bit), so I
> am not completely ignorant of the subject.
>
> I have sometimes manually adjusted kerning, or
> spacing between letters, but I have done it by

> eye. I used to squint up my eyes so that every-

> thing is a bit of a blur, and then sometimes tweak
> the letter spacing slightly so that a line of text
> would appear an even shade of grey, with no lumpy
> bits. Though with a slightly paler (wider spaced)
> gap between sentences.

Sounds rather like tracking to me, not kerning. I
have done both:

http://en.wikipedia.org/wiki/Kerning
http://en.wikipedia.org/wiki/Letter-spacing

Your technique for integral tracking correction is
interesting.

In modern typography, tracking is abused all too of-
ten, and when the reader notices changes in letter
density, it is abuse.

> I have rarely used 7 a typewriter, and never use

> fixed-space fonts, I think they look horrible.

And yet Usenet posts never look correct in propor-
tional fonts, because word-wrapping is based on the
character count in a line, which only in a
monospaced font is directly connected with the actu-
al line length.

Tony Cooper

unread,

Jan 16, 2014, 12:13:00 PM1/16/14

to

On Thu, 16 Jan 2014 05:26:04 -0800 (PST), pensive hamster
<pensive...@hotmail.co.uk> wrote:

>But a double space between sentences seems visually excessive to me,
>especially considering it comes after a full stop,

Yeah, but if you learned to touch-type decades ago on a manual
typewriter, that double tap on the space bar is as unstoppable as
wincing when the dentist's needle pierces the gum.
--
Tony Cooper - Orlando FL

Katy Jennison

unread,

Jan 16, 2014, 12:31:16 PM1/16/14

to

+1.

(I don't say it would be totally and utterly impossible to un-learn, but
it's quite low down on my list of things to expend the necessary time
and effort upon.)

--
Katy Jennison

pensive hamster

unread,

Jan 16, 2014, 12:58:01 PM1/16/14

to

On Thursday, 16 January 2014 15:41:40 UTC, Anton Shepelev wrote:
> pensive hamster:

> > I have sometimes manually adjusted kerning, or
> > spacing between letters, but I have done it by
> > eye. I used to squint up my eyes so that every-
> > thing is a bit of a blur, and then sometimes tweak
> > the letter spacing slightly so that a line of text
> > would appear an even shade of grey, with no lumpy
> > bits. Though with a slightly paler (wider spaced)
> > gap between sentences.
>
> Sounds rather like tracking to me, not kerning. I
> have done both:
> http://en.wikipedia.org/wiki/Kerning
> http://en.wikipedia.org/wiki/Letter-spacing

No, I am pretty sure I mean kerning. I would generally only do it when
using letterforms in a logo, or when laying out a strapline or similar
shortish line of text.

I found a little game where people can test their kerning skills. I scored
100 on the first two, and then got over-confident ...

http://type.method.ac/#

> Your technique for integral tracking correction is
> interesting.
>
> In modern typography, tracking is abused all too of-
> ten, and when the reader notices changes in letter
> density, it is abuse.

I have used tracking or stretching-out letters as well. Short sentences
or phrases displayed in isolation tend to be more readable if the letter
spacing is widened out slightly.

Or if I was writing a place name or other information in two lines on a
map, I might stretch the shorter word so that the two lines weren't too
different in length, eg.

G r e a t e r
Manchester

(I wouldn't do it to anything like that extent, it is just an example
composed within the limits of Usenet typography, to illustrate the principle).

pensive hamster

unread,

Jan 16, 2014, 1:09:30 PM1/16/14

to

On Thursday, 16 January 2014 17:13:00 UTC, Tony Cooper wrote:
> On Thu, 16 Jan 2014 05:26:04 -0800 (PST), pensive hamster

> >But a double space between sentences seems visually excessive to me,
> >especially considering it comes after a full stop,
>
> Yeah, but if you learned to touch-type decades ago on a manual
> typewriter, that double tap on the space bar is as unstoppable as
> wincing when the dentist's needle pierces the gum.

You don't double tap on the space bar in between each word, do
you, only in between each sentence?

(I never learned to touch type. I am just a two-finger typist, with
a flick of the thumb on the space bar.)

Tony Cooper

unread,

Jan 16, 2014, 1:22:03 PM1/16/14

to

On Thu, 16 Jan 2014 10:09:30 -0800 (PST), pensive hamster
<pensive...@hotmail.co.uk> wrote:

>On Thursday, 16 January 2014 17:13:00 UTC, Tony Cooper wrote:
>> On Thu, 16 Jan 2014 05:26:04 -0800 (PST), pensive hamster
>
>> >But a double space between sentences seems visually excessive to me,
>> >especially considering it comes after a full stop,
>>
>> Yeah, but if you learned to touch-type decades ago on a manual
>> typewriter, that double tap on the space bar is as unstoppable as
>> wincing when the dentist's needle pierces the gum.
>
>You don't double tap on the space bar in between each word, do
>you, only in between each sentence?
>

No, just after a period/full stop. It's done with my right thumb on
the space bar. Pure instinct at this point.

Athel Cornish-Bowden

unread,

Jan 16, 2014, 1:45:32 PM1/16/14

to

On 2014-01-16 00:43:01 +0000, Stan Brown said:

> On Wed, 15 Jan 2014 17:37:52 +0100, Athel Cornish-Bowden wrote:
>>
>> On 2014-01-15 12:37:18 +0100, Stan Brown <the_sta...@fastmail.fm> said:
>>
>>> On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:

>>>> p-values greater than 0.5 are highly insignificant.
>>>

>>> I don't agree. Aside from the silliness of "highly insignificant", a

>>> p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
>>> and the same is true of any other measurement.
>>

>> Likewise. I think the sentence is badly expressed. I'd say something
>> like "p-values greater than 0.5 are conventionally consider to be not
>> significant." (Incidentally, in this context I don't regard "not
>> significant" as synonymous with "insignificant": the former is
>> declining to make a statement; the latter is making a definite
>> statement.)
>
> You made the same mistake that I did, initially. I saw "0.5" (50%)
> and read it as "0.05" (5%).

Yes, you're right. It's an obvious way to misread it, because no one
who knows what he's talking about says "significant at the 50% level".

>
> But you're dead right about the difference between "not significant"
> and "insignificant". Statistics can never prove that there is no
> effect; it can only fail to determine whether there is one or not.
> It's frightening how many textbooks get that wrong.

--
athel

Athel Cornish-Bowden

unread,

Jan 16, 2014, 1:48:37 PM1/16/14

to

On 2014-01-16 01:16:32 +0000, Nathan Sanders said:

> On 2014-01-16 00:43:01 +0000, Stan Brown said:
>
>> On Wed, 15 Jan 2014 17:37:52 +0100, Athel Cornish-Bowden wrote:
>>>
>>> On 2014-01-15 12:37:18 +0100, Stan Brown <the_sta...@fastmail.fm> said:
>>>
>>>> On Wed, 15 Jan 2014 02:44:16 -0500, Nathan Sanders wrote:
>>>>> p-values greater than 0.5 are highly insignificant.
>>>>
>>>> I don't agree. Aside from the silliness of "highly insignificant", a
>>>> p-value of 0.5001 is essentially the same as a p-value of 0.4999 --
>>>> and the same is true of any other measurement.
>>>
>>> Likewise. I think the sentence is badly expressed. I'd say something
>>> like "p-values greater than 0.5 are conventionally consider to be not
>>> significant." (Incidentally, in this context I don't regard "not
>>> significant" as synonymous with "insignificant": the former is
>>> declining to make a statement; the latter is making a definite
>>> statement.)
>>
>> You made the same mistake that I did, initially. I saw "0.5" (50%)
>> and read it as "0.05" (5%).
>>
>> But you're dead right about the difference between "not significant"
>> and "insignificant". Statistics can never prove that there is no
>> effect; it can only fail to determine whether there is one or not.
>> It's frightening how many textbooks get that wrong.
>
> My dictionaries give "not significant" as the definition of "insignificant":
>
> <http://www.merriam-webster.com/dictionary/insignificant>
>
> <http://ahdictionary.com/word/search.html?q=insignificant>

OK, but I didn't get my training in statistics from dictionaries, and
I'd be surprised if you did.

>
> And in my experience, this is indeed how these expressions are used in
> the wild by real statisticians, so it's not just lay usage.

Maybe, but it's just sloppy wording.

> "Statistically insignificant" is just a shorter way of saying "not
> statistically significant". It's analogous to using "p is greater
> 0.05" instead of "p is not less than or equal to 0.05":
>
> p not <= 0.05 : not statistically significant
> ::
> p > 0.05 : statistically insignificant
>
> Since "p not <= 0.05" and "p > 0.05" are equivalent, so are "not
> statistically significant" and "statistically insignificant".
>
> What do you think "statistically insignificant" (with respect to
> p-values) could mean instead?

I wouldn't (I hope) ever use such a term.

> And have you ever seen a statistician use it that way?

I don't remember (but Stan says that many textbooks get it wrong).

>
> Nathan

--
athel