Hensachi

239 views
Skip to first unread message

imagina...@gmail.com

unread,
Apr 30, 2014, 5:25:13 AM4/30/14
to hon...@googlegroups.com
In a paper on automatic problem solving, a score is given as 受験者中偏差値約60. I hear people talking about 偏差値, and I believe it means something like standard deviation, so this may well mean "1 SD above the mean". But I have no idea if this sort of terminology is used in the English speaking world (it certainly wasn't when I took exams), or quite how to say it. Any suggestions?

Thanks
Brian Chandler

timl...@aol.com

unread,
Apr 30, 2014, 5:37:36 AM4/30/14
to hon...@googlegroups.com
Brian,

"one standard deviation above the mean" gets a huge number of googits.

FWIW
Tim Leeney
--
You received this message because you are subscribed to the Google
Groups "Honyaku E<>J translation list" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to honyaku+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Oroszlany Balazs

unread,
Apr 30, 2014, 6:19:02 AM4/30/14
to hon...@googlegroups.com

imagina...@gmail.com

unread,
Apr 30, 2014, 6:21:11 AM4/30/14
to hon...@googlegroups.com
** Sorry, I can make no sense whatsoever of this Google form. This is my reply to Tim Sweeney, who said:

"one standard deviation above the mean" gets a huge number of googits.

Thanks for quick response.
Yes, of course these are standard terms from statistics (which is not my subject). But the problem is how to talk about the exam/test results. By about the 10th result in this google results, I found this (I reproduce the url after a bit more googlepain):
www.wrightslaw.com/advoc/articles/bellcurve.pdf

At the bottom it comments that some test scores use "T scores" where the mean is 50 and the SD is 10. I guess this means the system universally used in Japan, but the fact that this is a footnote suggests it is not the universal or even a standard way of describing exam results in the English-speaking world.

Does anyone else have experience of translating such scores?
Brian Chandler


Jens Wilkinson

unread,
Apr 30, 2014, 6:45:02 AM4/30/14
to hon...@googlegroups.com
The problem with the term is that it has a statistical meaning, but in practice people use it to mean a standardized score. So a school with a hensachi of 60 is moderately difficult to get into; one with a hensachi of 80 is very hard. So it's sort of like SAT scores in the sense that getting 600 isn't that hard but 700 is much tougher. But in Japan for some reason it's called a standard deviation even though I guess that 95% of parents (p=0.002) have no idea what a standard deviation is. All we know is that if our kid's hensachi is lower than the average for the school they are aiming for, it's time for juku or yobiko.

So I would tend to say, had an average hensachi of 60.

Jens Wilkinson

imagina...@gmail.com

unread,
Apr 30, 2014, 8:43:43 AM4/30/14
to hon...@googlegroups.com
Oroszlany Balazs said:
 > Isn't it standard score?

Well, it's a sort of "standard score", and the US Navy page seems indeed to have the same scaling as the Japanese standard. But note that the Japanese WP page for 偏差値 says "standardised on mean = 50, SD = 10", the corresponding English page does not. It says the "standard score" is the number of standard deviations; so a 偏差値 of 60 would be a "standard score" of 1 on that definition. But the PDF I linked above says that (USA?) "standard scores" usually have a mean of 100 and an SD of 15, which very definitely does not match.

Jens said -- much good sense -- except that I don't think it's appropriate just to cop out with 'hensachi' in an academic paper, where readers will assume it really does mean "standard deviation", and thus makes no sense.

I think we'll have to put a footnote or something. I conclude that there is no really standard way to quote these.

Thanks
Brian Chandler

Joel Dechant

unread,
Apr 30, 2014, 10:22:13 PM4/30/14
to hon...@googlegroups.com

Ah, the dreaded hensachi.

In statistics, it is indeed standard score or standard deviation, but in the realm of university examinations, what it really boils down to is selectivity. 
Todai and its ilk require higher scores, making them harder to get into, hence they are more selective. (http://www.ireg-observatory.org/ireg-6/presentation/akiyoshi.pdf)

A footnote is probably your best bet.

Joel Dechant



Steven P. Venti

unread,
Apr 30, 2014, 11:11:15 PM4/30/14
to hon...@googlegroups.com
Joel Dechant <jhde...@gmail.com> wrote:

> in the realm of university examinations, what it really boils down to is
> selectivity.

In the US, at least, isn't this type of selectivity usually expressed in terms
of percentile rank? Perhaps a reference to percentile rank is one way of
making this easily intelligible.

FWIW

-----------------------------------------------------------------
Steve Venti
spv...@bhk-limited.com

Arima 1-11-5-203
Miyamae-ku, Kawasaki, Kanagawa 216-0003
Tel: 090-8045-5128

-----------------------------------------------------------------

Herman

unread,
Apr 30, 2014, 11:19:42 PM4/30/14
to hon...@googlegroups.com
On 4/30/2014 03:21, imagina...@gmail.com wrote:
> At the bottom it comments that some test scores use "T scores" where the
> mean is 50 and the SD is 10. I guess this means the system universally
> used in Japan, but the fact that this is a footnote suggests it is not
> the universal or even a standard way of describing exam results in the
> English-speaking world.
>

I believe SAT scores are basically computed in the same, only using 500
and 100 instead of 50 and 10. In this case, the term normally used is
simply "score". But in any case, T-score is the correct translation for
"hensachi".

Herman Kahn


Reply all
Reply to author
Forward
0 new messages