Like it? Share it!

Sign up for news and updates!






Enter word seen below
Visually impaired? Click here to have an audio challenge played.  You will then need to enter the code that is spelled out.
Change image

CAPTCHA image
Please leave this field empty

Login Form



Forget Parties.. Vote by Sign! PDF Print E-mail
Swift
Written by Jeff Wagg   

Reader Steve has pointed us to an interesting thing in the Washington Post. Unsurprisingly, they report on congressional votes. Their online report has some nice sort options:

By party | By state/territory | By region | By boomer status | By gender | By astrological sign

Yes, that last item is noteworthy.

Yes No Not Voting
Aquarius 26 7
Aries 22 7 1
Cancer 40 12 3
Capricorn 20 9
Gemini 33 15 2
Leo 25 6
Libra 32 9 4
Pisces 23 3 1
Sagittarius 23 6
Scorpio 22 7 3
Taurus 24 5
Virgo 33 9
Yes No Not Voting
Female 52 5 3
Male 271 90 11
Yes No Not Voting
Midwest 75 21 3
Northeast 79 2 1
South 96 53 5
West 73 19 5
Yes No Not Voting
None 48 14 4
2008 275 81 10
Yes No Not Voting
Baby boomer 196 67 10
Post-boomer 35 10 2
Pre-boomer 92 18 2
Yes No Not Voting
Democratic 247 2 5
Republican 76 93 9

If you supported this bill, you should have voted Pisces in the last election. If you were against it, Capricorn was the way to go. Scorpio seems to be the sign of choice for those of you who like to abstain.

You see, now that we have this tool, we no longer need to worry about campaigning or party lines. All we need is someone's birthdate, and we can make informed decisions on how our representatives will vote. In fact, maybe we can do without the vote all together and just let a computer calculate the results based on astrological trends.

The scary thing is that some readers will take the above paragraph seriously. Come on Washington Post, publishing horoscopes is one thing (not that I approve), but suggesting that there is some link between astrological sign and congressional voting is dishonest. Why not list tie color or shoe size? How about height or weight? Or maybe what their favorite sandwich is?

Strangely, I wouldn't be surprised if there was a slight correlation between sign and voting, simply because there is a correlation between birthdate and year of graduation from high school. But to suggest that astrological sign is just as important as region, party, or age is ridiculous. You can chart anything and find correlation, but to find causality is a completely different matter. Listing votes by sign is useless noise, and the Washington Post should know better.

Trackback(0)
Comments (11)Add Comment
Bad election advice ;-)
written by BMN, February 27, 2009
Your calculations seem to be a bit off. If you're in favour of the bill, then Pisces is better than Taurus. If you're against, then Capricorn is a better bet than Gemini. And if you like to abstain, Scorpio is better than Libra.
So, err, 0 out of 3 smilies/sad.gif

EDITED BY JEFF WAGG: I've reconsidered, and you're correct. I'll change it. Thanks.
report abuse
vote down
vote up
Votes: +2
...
written by Bruno, February 28, 2009
I can imagine one's outlook on life correlating with the time of year one is born. However, with on average 36 congressmembers per sign, sampling error is 16%. Too early to tell if there is a correlation.
report abuse
vote down
vote up
Votes: +0
...
written by BillyJoe, February 28, 2009
with on average 36 congressmembers per sign, sampling error is 16%.

How did you calculate the 16%?

report abuse
vote down
vote up
Votes: -1
...
written by Bruno, February 28, 2009
1/sqrt(36)
report abuse
vote down
vote up
Votes: +0
Thanks, but..
written by BillyJoe, February 28, 2009
In the mean time I had looked up the formula:

1.96 * sqrt{(1-f)p(1-p)(n-1)}

It is different from yours but gives the same answer:
Because...

Taking p = 0.5 for the largest value for p(1-p) which = 0.25
Approximating 1.96 = 2 = sqrt 4
Approximating 1-f = 1 for small sample sizes compared to the whole population where f = sample/population
Approximating n-1 = n for reasonably large sample sizes

...this formula aproximates to your version:

1/sqrt n

How neat is that??
But...what does this mean?
report abuse
vote down
vote up
Votes: -1
Zodiac Voting
written by StarTrekLivz, March 01, 2009
But how do their voting patterns correlate to their Chinese astrological signs?

And are we taking into account the fact the sun signs have shifted since the days of the Babylonians?

And isn't it amazing that in a country that many proudly proclaim is "Christian", people follow the horoscope, a behavior forbidden by Leviticus (it's as abominable as homosexual relations or eating shrimp).
report abuse
vote down
vote up
Votes: +0
...
written by BillyJoe, March 01, 2009
Oops, I meant:

1.96 * sqrt{(1-f)p(1-p)/(n-1)}
report abuse
vote down
vote up
Votes: -1
But...
written by BillyJoe, March 01, 2009
What I am actually asking is how do you determine whether the result obtained here is statistically significant.
report abuse
vote down
vote up
Votes: +0
...
written by Careyp74, March 02, 2009
What Bruno is saying, is that with a 16% sampling error, if the actual population voting was equal with regards to sign and choice, you could still have what would appear to be a heavy favoring by any sign to a particular choice in this sampling. You need a much bigger sampling for the results to reflect the true population.
report abuse
vote down
vote up
Votes: +1
And the Zodiacal constellations don't match sun signs anyway
written by garyg, March 02, 2009
Then there's the "extra" sun signs and the fact that the sun isn't where
astrologers say they are:
http://secure5.ozhosting.com/skeptics/skeptic/1999/3.pdf
report abuse
vote down
vote up
Votes: +0
...
written by BillyJoe, March 02, 2009
What Bruno is saying, is that with a 16% sampling error, if the actual population voting was equal with regards to sign and choice, you could still have what would appear to be a heavy favoring by any sign to a particular choice in this sampling. You need a much bigger sampling for the results to reflect the true population.

Thanks, but I still don't have an asnwer to my question: "how do you determine whether the result obtained here is statistically significant". You're saying 16% is too big, but what is the cut off point?
report abuse
vote down
vote up
Votes: -1

Write comment
This content has been locked. You can no longer post any comment.
You must be logged in to post a comment. Please register if you do not have an account yet.

busy