...from Jim Alcock
Hi John and Kramer and Randi:
I am back in town now, and I have been giving some thought to this test. We have to be very careful with regard to the choice of participants. For example, people who maintain an interest in horoscopes may well come to view themselves in terms of what they have read about the meaning of their sun signs. If a person is a "Virgo" and knows what that is supposed to mean about his/her personality astrology-wise, then he/she may respond to te questionnaire according to that belief (whether or not he/she has any inkling about the nature of this test and its focus on astrological signs). Indeed, there is some reason to believe that horoscope devotees may even have to some extent shaped their personalities and behaviours according to that belief. "I'm a Vïrgo, we are all ..."
So, to be sure that we are examining the correspondence between astrological sign and personalities, we need to be sure that neither the personality itself, nor the individual's responses to the test, have been distorted to some extent by knowledge of the person's sun sign. This is NOT a minor problem.
How do we eliminate this problem? Keeping the participants ignorant of the astrological aspect of the study is a good first step, but certainly on its own does NOT eliminate the problem. Ideally, we need to choose people who have no knowledge of their sun sign and what it is supposed to mean.
However, it is hard to recruit people by asking,"Do you know your sun sign?" for that obviously tells them that the study is about astrology.
Further, people who do not know their sun sign are probably less likely to have knowledge of their time of birth, for they will not have bothered to find out. (I am not sure off-hand whether most, or all, or any birth certificates also record time of birth. If they do, this would not be a problem).
John, I don't know how this would sit with you, or with Randi and Kramer, but I would be inclined towards the following: Since we do not know in advance the possible confounding effect that knowledge of sun sign might cause, and since it would take some considerable effort to eliminate its possible effects, a two-step procedure would be in order:
1. First, run the study without regard for possible knowledge of sun sign.
If the results are negative, there is no point in going any further. The test fails.
2. If the results reach the criterion for success (10/12 correct), we cannot know whether the results support astrology or simply reflect the effects of the participants' knowledge of what they should be like if astrology is correct. We then run a second, definitive, study where we take all the pains necessary to eliminate that possible influence.
Please let me know your thoughts on this. We also have to discuss and agree upon the procedure for selection of participants.
Sincerely,
Jim Alcock
Hi John and Kramer and Randi:
I am back in town now, and I have been giving some thought to this test. We have to be very careful with regard to the choice of participants. For example, people who maintain an interest in horoscopes may well come to view themselves in terms of what they have read about the meaning of their sun signs. If a person is a "Virgo" and knows what that is supposed to mean about his/her personality astrology-wise, then he/she may respond to te questionnaire according to that belief (whether or not he/she has any inkling about the nature of this test and its focus on astrological signs). Indeed, there is some reason to believe that horoscope devotees may even have to some extent shaped their personalities and behaviours according to that belief. "I'm a Vïrgo, we are all ..."
So, to be sure that we are examining the correspondence between astrological sign and personalities, we need to be sure that neither the personality itself, nor the individual's responses to the test, have been distorted to some extent by knowledge of the person's sun sign. This is NOT a minor problem.
How do we eliminate this problem? Keeping the participants ignorant of the astrological aspect of the study is a good first step, but certainly on its own does NOT eliminate the problem. Ideally, we need to choose people who have no knowledge of their sun sign and what it is supposed to mean.
However, it is hard to recruit people by asking,"Do you know your sun sign?" for that obviously tells them that the study is about astrology.
Further, people who do not know their sun sign are probably less likely to have knowledge of their time of birth, for they will not have bothered to find out. (I am not sure off-hand whether most, or all, or any birth certificates also record time of birth. If they do, this would not be a problem).
John, I don't know how this would sit with you, or with Randi and Kramer, but I would be inclined towards the following: Since we do not know in advance the possible confounding effect that knowledge of sun sign might cause, and since it would take some considerable effort to eliminate its possible effects, a two-step procedure would be in order:
1. First, run the study without regard for possible knowledge of sun sign.
If the results are negative, there is no point in going any further. The test fails.
2. If the results reach the criterion for success (10/12 correct), we cannot know whether the results support astrology or simply reflect the effects of the participants' knowledge of what they should be like if astrology is correct. We then run a second, definitive, study where we take all the pains necessary to eliminate that possible influence.
Please let me know your thoughts on this. We also have to discuss and agree upon the procedure for selection of participants.
Sincerely,
Jim Alcock