This paper explores to what extent the zodiac is reasonable to be accepted as a determinant of compatibility between two people. People have always been curious about their destiny and usually approached astrologers to find it out. It is generally believed that certain zodiac signs work better with each other, rather than with other signs. Nowadays any individual with simply an access to the internet is aware of this to be true and verifiable. However, it is not clear how justifiable it is to believe in astrological predictions and this confusion needs to be addressed.
Statistical analysis in one of the ways to check the relevance of astrological statements about the success of love relationships. For this purpose, a number of divorced couples were investigated for compliance with respect to horoscope.
Astrology, perhaps one of the most famous personality theories, has created little empirical research, because it lacks sanction as scientific inquiry, so actually it is recognized as pseudoscience (Silverman, 1971).
The principal concept of astrology is that the individual’s personality is greatly influenced at the moment of birth by the position of the Sun, moon and planets within the zodiac, an area of the sky 16° wide that extends around the Earth roughly parallel to the equator. Astrologers have divided the zodiac into 12 equal portions, each corresponding to a particular sign. The position of the Sun alone, at the instant of birth, determines the individual’s sign.
There are a few studies that tested how valid astrology is and, generally, the results did not manage to support predictions of the astrologers. However, there are still quite a lot of people who trust it.
The purpose of this study is to undermine astrology’s popularity and to persuade people not to pay attention to the zodiac sign of your soulmate. When it comes to the true love, even the stars are powerless.
Procedures and Data Collection
This section provides a brief overview of the data collection process. It includes an explanation of the methods of data gathering, parameters of the sample, techniques employed to create the sample and the size of the sample.
To determine whether the astrological incompatibility truly affects the divorce rates we have used three different astrological tables. For each of the three tables, if the signs were incompatible and got divorced, we put “1” point to astrologers; if the signs were partially compatible and got divorced, we put “0.5” points to astrologers, if the signs were perfectly compatible but still got divorced, we put “0” points to astrologers. Subsequently, we calculated astrologers’ points and divided it by the number of divorced couples, obtaining our sample proportion of the true astrological predictions. Afterwards, we performed a one-sample z-test for
proportion on the two appropriate significance levels.
The aim of the research is to determine whether the astrological forecast tables can really predict an outcome of the marriage, so the target population in this study is divorced couples. Firstly, we decided to conduct a survey on www.surveymonkey.com. Using this site we obtained roughly about 70 % of our final sample, while the other 30 % we acquired using the random sample of divorced celebrities from all over the world. Of the vast range of possibilities, we decided to use this technique due to the fact that our group wanted to cover different layers of society, starting from simple Kazakhstani couples and ending with the famous American politicians. To ensure confidentiality, names and surnames of of the subjects were removed from the conclusive paper. The final data set yielded a sample size of 85 divorced couples.
Statistical Analysis
H0: p=0.5 and Ha: p<0.5In the context of the research:
H0: proportion of true astrological predictions about zodiac incompatibility=0.5
Ha: proportion of true astrological predictions about zodiac incompatibility<0.5
Conducted hypothesis tests are both one sample z-tests. The conditions are met:
1) n * p>10, as 0.5*85=42.5 > 10
n * (1-p)>10, as 0.5*85=42.5>10 → The population is normal.
2) n<0.1N, as the sample, that is 85, more than 10 times less than the amount of all divorced people in the world. 3) We assume that the sample is an SRS.
The first table gives the following results: p̂1=0.382, means that based on the first table, astrologists did a correct prediction 38.2 % of times. We test hypothesis at α=5 % and α=10 %
The corresponding p-value, according to table 5 of the New Cambridge Statistical tables, is equal to 0.015. 0.015<0.05<0.1, that is p-value=0.015< α=5 %< α=10 %
The p-value is less than α at both significance levels, so we reject the H0. There is strong evidence that the astrologists are right in less than 50 % of the time.
The second table gives the following results: p̂2=0.471, means that based on the second table, astrologists did correct prediction 47.1 % of times, in our sample. We construct a 95 % C.I.:
We are 95 % confident that the probability that astrologists make accurate predictions about incompatibility of two people is between 0.3648 and 0.5771
The third table gives the following results:
p̂3=0.371, means that based on the third table, astrologists did a correct prediction 37.1 % of times. We test hypothesis at 2 significance levels: α=5 % and α=10 %.
The corresponding p-value, according to table 5 of the New Cambridge Statistical tables, is equal to 0.0089. 0.0089<0.05<0.1, that is p-value=0.0089<α=5 %<α=10 %
The p-value is less than α at both significance levels, so we reject the H0. There is strong evidence that the astrologists are right in less than 50 % of the time.
Identification and elimination of biases.
Among ethical issues in contemporary society the most serious and disturbing might be divorce and remarriage. Therefore, people are not always ready to answer the questions about it. The fact that we did not use a random sampling technique helps us to avoid non-response, response biases and save time. However, it also leads to the following consequences:
- Inability to accurately determine the sampling error.
- Selection bias. We recognize that not all of the segments of population respond to surveys in social networks. Hence, due to the assumption that famous people do not usually participate in surveys, we used data about the divorces of celebrities (politicians, musicians, actors, models and etc). The limitations of this research include:
- Project raises an ethical issue, so the majority of people do not want to share the required information.
- There are a lot astrological tables and it takes a lot of time to check each table.
Conclusion
The results of the hypothesis tests, according to the data of the 1st and 3rd tables, show that the astrologists’ predictions become filled in less than 50 % of the time. Moreover, the confidence interval for the population of all divorced couples, derived from the second table, indicated that the proportion of true astrological predictions about zodiac incompatibility is between 0,3648 and 0,5771. Therefore, even if this proportion is 0,5771, i.e. at its maximum possible value, astrology is still not a good predictor, because the probability is still too low to be valid. Despite the fact that we could not obtain a simple random sample, we still tried to minimize biases as much as possible and create the sample that would represent the whole population of our interest.
References:
- Silverman, B. I. (1971). Studies of Astrology. The Journal of Psychology, 141–149.
