Welcome!!

Here is a study that intermingles science with mysticism.

The art of astrology is ancient, dating back thousands of years. As science advances, astrology has become outmoded and regarded by "real" scientists as fiction. But is there any truth to the assumptions? Is astrology for real, or is it merely a way of arbitrarily classifying people? I undertook the task of discovering who was right and who was the crackpot.

The study involved a survey and 38 participants, who were random high school students. I discovered quite a few flaws in my experiment along the way, so if I were to conduct another similar experiment, I would follow through in a much different way. But hey, I’m a novice at this. This study was a growing experience. Isn’t that what they say about everything that turns out unexpectedly?

At any rate, I chose to set a hypothesis of pi equaling 0.25. That is, I thought that 25 percent of my sample would have two out of three traits that are agreed upon by most astrologers as typical of the birth sign. I had the subjects choose six traits of 36 that best described him or her. I took down the subjects’ birth dates. Later, I decided whether it was a success or failure according to my key. If the subject had two traits that were designated as their birth sign, that was a success. Anything else was a failure. In English, my hypothesis is that 25 percent of the population has character traits that are typical for his or her birth sign.

The result? Well, you’ll see.

Just so you know. . .

To tell the truth, I had some problems locating background information. Seems as though people who believe will do so without question, and those who don’t also do so without proof, which to me defies the notion of scientists needing proof before rejecting a theory. Therefore, those who didn’t believe didn’t care enough to study it and those who did didn’t want proof of nonexistence. Before I get into other studies of this nature, though, it may be useful to know something about the zodiac.

According to astrology, every person who is born during a particular time frame will share some basic characteristics with other people born during the same time frame. for example, people born from September 24 to October 23 are supposedly indecisive, but very fair to both sides of an issue. Members of the scientific community scorn this, arguing that the traits given to each sign are too vague and that the personality is determined half by genetics and half by environment. These are valid arguments, even I admit. But then, that’s why I chose a proportion of 25 percent rather than anything higher.

Now for the other study.

At the University of Indiana, a study much more comprehensive than my own was conducted using real astrologers and eleven test subjects. The subjects were interviewed and that data was analyzed by said psychics. Not only did the astrologers fail to match the subjects to his or her birth sign, but they did not agree with each other. Obviously, this study concluded that there was not sufficient evidence to conclude that character traits and zodiacal signs were associated.

The Big Nasty

Here’s the part of the experiment that we all know and love: the survey. The population is every person in the world, since everyone has a birth date and everyone has a personality. I chose my quarry for this part from my resources at the high school. I randomly went up to people during lunch and sought friends of friends to vary the sample. Therefore, I didn’t study only Honors level senior students.

On the survey was a list of 36 traits, three from each astrological sign. I asked the students to choose six that best described them, in hopes that two would be from their own zodaical sign. Why six? Because I’m not a fool. I know that some personality traits are inherited and others arise due to the person’s environment. I also asked that the students write their birth dates so I could determine success and failure. As I said, if they had two or more traits that were attributable to the sign, the survey was a success. One or none denoted a failure.

The Survey

Here's the survey I handed out to my victims, er, subjects, word for word.

Pretty Pictures

Here's the first graph. The x-axis is actually the birth signs.

1-Aries

2-Taurus

3-Gemini

4-Cancer

5-Leo

6-Virgo

7-Libra

8-Scorpio

9-Sagittarius

10-Capricorn

11-Aquarius

12-Pisces

Here's the pie chart, just because I think it's cool.

As you can see, the histogram for the character traits is everywhere. One notable thing is that stubbornness was a Taurus quality, but half the people chose it.

The Big Test

First, I have to determine if I can use a z-test or if I have to use a t-test. That is, I have to determine if the sample size is large enough.

The formula is:

np >5

and

n(1-p) > 5

38 * .25 > 5                                              38*(1-.25) > 5

9.5 > 5                                                      28.5 > 5

yes                                                              yes

All right, that means that a z-score is appropriate for this data.

level of significance = .05

z = (p-pi)/(p(1-p)/n)^(1/2)

It looks complicated, but it’s not.

(.1063-.25)/(.1063(1-.1063)/38)^(1/2)

z =-2.87

P-value = .0021

P-value < .05

The End is Near

Since the P-value is less than the level of significance, the hypothesis is rejected. There is not sufficient evidence of 25% of the population having two or more character traits that belong to their birth sign at the 5% significance level. In other words, nope. I can’t say that that character traits and birth signs aren’t associated; they aren’t associated at the proportion I believed. Besides, my study was nowhere near extensive enough.

If I had known then . . .

Anyway, I don’t know where to begin as far as weaknesses go. I would do a thousand things differently if I knew then what I know now. For starters, i wouldn’t have chosen this topic. It’s too extensive to be studied by one wimpy survey, which brings me to the next weak link. The survey was hard to make and I see now how badly I botched it. The problem is, a lot of those traits have some duality. Some traits should have been put on and others left off. I’ve come to the conclusion that a survey method was a bad way to go on this particular study. Furthermore, some of the traits were not positive, which adds the confounding factor of response bias. Add to that the problem in choosing a sample. With the population being all people in existence, choosing only teenagers in a middle-class suburb is sampling bias. I didn’t even choose my sample out of hat or via a random number sample. You might as well throw in a research bias as well.  Despite the results, I still believe that there is some kind of connection between astrological signs and personality, despite that it's not logical and I have no proof.  In other words, I wouldn’t trust this study or this conclusion as a reliable source of information.

THE END!!!