MCI Nettest

Ron Resnick (
Fri, 18 Jul 1997 15:32:21 +0300 (EET DST)

Rohit, do you know anything about how this is set up, under the covers?

Here's a note posted by someone here at IBM, which is how I learned
of the test:

----- SCNEWS NEWSFILE appended at 13:53:18 on 97/07/15 GMT (by YOELLE at HAIFASC3)
. . 8 97/07/15 15:48 MCI: Women Surf Better Than Men (YOELLE)
MCI: Women Surf Better Than Men
Results from the first Internet aptitude test, which required surfers to
use a variety of Web search sites to find answers to five questions
covering pop culture, science, history and the arts, found that of the
650,000 initial respondents, only 17,000 (2.6 percent) finished the test.
The average score was 78.8, the median, 84. Overall, two-thirds scored less
than the average and 13 percent scored less than 50, meaning they got two
or fewer questions correct. More men participated, 68 percent, but women
proved to be slightly better at surfing, with an average score of 79.91 vs.
78.29 for men. The highest scoring group was women ages 24-39, with an
average of 80.85. Men ages 24-39 scored slightly lower, 80.52.

PS: If you are interested in such surveys,
(which are often more serious than this one), you can subscribe to
the NUA mailing list by sending mail to
with the word suscribe in the body of the message

Note the 2.6% statistic above, for 'finished the test'! Yet more
evidence for 97%/3% :-)

I took the test twice. Once at work, where I finished in about 18 minutes,
but then it (presumably the cgi server) screwed up and couldn't
process my 'win'. Perhaps because I
took the test from a browser behind the IBM firewall? The second time,
I did it from home, and finished in about 24 minutes, based on my own
watch. It took that long because of long network delays - about 3
minutes at one point between typing a keyword search into Yahoo and getting
a response. I think there's also a bug in MCI's timing - It reported
the following:

You finished in 0:65:34 using a ISDN connection which gives
you a score of 96.

When, as I noted, it took me about 24 minutes. I have no idea what 0:65:34
means anyway. If it was meant to be 1hour+, why not say so? Looks buggy.
Also, I was trying to figure out what the '96' meant. What's the calibration
of their scale? What's the maximum? 100? Based on the averages apparently being
in the 80 range, I would assume that 96 is a pretty good score, but I'd
like to know *how* good.

I told it that I have an ISDN because I set it up at work, where I have
a dedicated connection, which isn't ISDN, but then again isn't T1 either
(I actually don't know what our line is through Tel Aviv and out). But
at home I have a 28.8 modem, and I couldn't change my MCI settings to reflect
that. That's ok - presumably if it 'thinks' I have ISDN when I don't, that
would only make my score that much better, right?

Another complaint: they limit you to only 5 or so search engines that
they pick - Altavista, Lycos, Yahoo, .. 2 others. I think Infoseek.
They should set it up to let you use arbitrary search devices. Lately,
I've grown to like Hotbot, thanks to Robert Thau's recent tip.

So how 'bout a FoRK challenge ;-)? FoRKers take the test and post their
scores. On your honour not to cheat, of course! I wonder what a FoRK
average would be? Probably well above the 'national average'. Unless
we humble ourselves with our collective 97edness.

Ron Resnick IBM:
IBM Haifa Research Lab HRL world:
MATAM 31905 Haifa, Israel
Tel: 972-4-829-6491

"A shadow is the ..
... mobile, persistent, distributed, self-descriptive,
self-contained, pro-active, cooperative, trustworthy,
object-oriented, software analog of some real world object"
-- Sandor Spruit