Graphology
- Thread starter The Mighty Thor
- Start date
No no... believe OVERALL that the results would come out to around ... say... 70%
Unless I am missing some crucial information here, your statement that you believe the results would come out to around ... say ... 70% implies that you in fact have no real idea how accurate your results are; you are guessing. At that, you are guessing that your level of accuracy is still pretty low. Bear in mind, that since you start from a position of tending to believe that there is something to graphology then the natural presumption must be that you are going to over-estimate the your degree of success rather than underestimate it. This is not in any way to suggest that you are dishonest or intellectually under-endowed, we all have this tendency which is one of the reasons for all the safeguards in scientific investigation (double-blind trials, the drive to falsify rather than prove one's theory, etc.)
Even if it is the case that your overall results are 70% accurate and this were to be confirmed by a reasonable sized controlled study of your results over time, this would only be enough to indicate that there is something to graphology it would not indicate that anything useful was going on. A 70% accuracy rate would be terribly useful if they were odds for gambling and you had sufficiently large pockets to play for a while but it is well below the level of predictability that any scientific study would have to show in order for it's results to be deemed valid. On a more "common-sense" level, should any important decision be based on a method which is wrong 30% of the time (and I include in important decisions, "should this person get the job?", "should I vote for him at the next election" or even more importantly, "is this a nice person?", "are they nice enough to be my friend?").
Unless I am missing some crucial information here, your statement that you believe the results would come out to around ... say ... 70% implies that you in fact have no real idea how accurate your results are; you are guessing. At that, you are guessing that your level of accuracy is still pretty low. Bear in mind, that since you start from a position of tending to believe that there is something to graphology then the natural presumption must be that you are going to over-estimate the your degree of success rather than underestimate it. This is not in any way to suggest that you are dishonest or intellectually under-endowed, we all have this tendency which is one of the reasons for all the safeguards in scientific investigation (double-blind trials, the drive to falsify rather than prove one's theory, etc.)
Even if it is the case that your overall results are 70% accurate and this were to be confirmed by a reasonable sized controlled study of your results over time, this would only be enough to indicate that there is something to graphology it would not indicate that anything useful was going on. A 70% accuracy rate would be terribly useful if they were odds for gambling and you had sufficiently large pockets to play for a while but it is well below the level of predictability that any scientific study would have to show in order for it's results to be deemed valid. On a more "common-sense" level, should any important decision be based on a method which is wrong 30% of the time (and I include in important decisions, "should this person get the job?", "should I vote for him at the next election" or even more importantly, "is this a nice person?", "are they nice enough to be my friend?").
Well, my 70% is me trying to be realistic
As already mentioned, it's a wild guess based on subjective feedback.
Now about using it for eg. job interviews and girlfriend assessment - I've already stated that I don't agree with it being used in that way.
Even if the results are, say, 85% accurate on average ... (1) there's a chance that this particular girlfriend has a situation/condition causing confusion in the handwriting. Thus judgement upon the writing is unfair. and (2) I wouldn't like to be rejected from a job or girlfriend because of a 15% chance. That's unfair in itself.
Thus I agree already that it has little practical value outside of a party trick. IMO and thus far.
As already mentioned, it's a wild guess based on subjective feedback.
Now about using it for eg. job interviews and girlfriend assessment - I've already stated that I don't agree with it being used in that way.
Even if the results are, say, 85% accurate on average ... (1) there's a chance that this particular girlfriend has a situation/condition causing confusion in the handwriting. Thus judgement upon the writing is unfair. and (2) I wouldn't like to be rejected from a job or girlfriend because of a 15% chance. That's unfair in itself.
Thus I agree already that it has little practical value outside of a party trick. IMO and thus far.
The figures mentioned don't sem to have much meaning as the exact same feedback could have been considered 70% accurate to all the people you gave it to.
This wouldn't mean that graphology is 70% accuracy, merely that you had described some character traits that most people believe applies to them.
A good way to actually test graphology (I think this has been mentioned previously) is to get, say, ten people, analyse their writing then get them to pick which description they feel applies best to themselves.
If actually measured in testing I don't think you'd find it has any relation to personality at all.
This wouldn't mean that graphology is 70% accuracy, merely that you had described some character traits that most people believe applies to them.
A good way to actually test graphology (I think this has been mentioned previously) is to get, say, ten people, analyse their writing then get them to pick which description they feel applies best to themselves.
If actually measured in testing I don't think you'd find it has any relation to personality at all.
As a party trick, it's fine. Many years ago, I used to read palms as a party trick and so far as my, very likely faulty, memory recalls every person whose palm I read thought I was extremely accurate. It was, nevertheless a trick just like graphology and both seem likely to work on the basis of the Forer effect.
I suppose I am a little surprised that, while from your posts you are obviously intelligent and have the good taste to quote George Bernard Shaw (who would, I suspect, have despised graphology) you seem willing to promote graphology from party trick to science on the basis of no real evidence: the Forer effect clearly can account for all your successes (as well as mine as a palmist) and you have not countered the objection to your assumption that a few distinct explanations could account for all the possible reasons one might scribe a fancy A or draw a matchstick cat (did you ever look at Lowry, a man who could most definitely paint?)
I suppose I am a little surprised that, while from your posts you are obviously intelligent and have the good taste to quote George Bernard Shaw (who would, I suspect, have despised graphology) you seem willing to promote graphology from party trick to science on the basis of no real evidence: the Forer effect clearly can account for all your successes (as well as mine as a palmist) and you have not countered the objection to your assumption that a few distinct explanations could account for all the possible reasons one might scribe a fancy A or draw a matchstick cat (did you ever look at Lowry, a man who could most definitely paint?)
Ashles said:
Well put.
That's a basic problem. It's very easy for me to write statements that 95+% of the population would agree accurately describe themselves. For example, who on this forum does not like warm, just out of the oven chocolate-chip cookies? Is there anyone who does not "at times [...]have serious doubts as to whether you have made the right decision or done the right thing" (to quote Forer)?
Placebo,
I can’t go through the entire list of references on the one link that provides references, but I started at the top of the list and went down as far as time allowed.
I skipped those available only in French.
Here’s what I found (it may help you to learn not to accept people at their word and to check out references even if presented in an apparently professional manner):
---
REFERENCE: APA Guidelines for Educational and Psychological Testing
American Psychological Association
APA: 1974
MY COMMENTS: These are guidelines for testing. They are not research. And they do not address the efficacy of graphology.
---
REFERENCE: Barrow, N K and Scott, R H
Validation of a personnel Selection System to meet EEOC Guidelines
Journal of Handwriting Psychology
1984, 1 (1), 15 – 17
MY COMMENTS: Not available on line
---
REFERENCE: Baruch Nevo
Graphology Validation Studies in Israel
1986
MY COMMENTS: Not available on line and not listed on Baruch’s online C.V. A similar paper is, though, but is listed as being published in a non-refereed journal.
---
REFERENCE: Beumont, P.
British Journal of Psychiatry
Small handwriting in some patients with anorexia nervosa
1971. 119, 349-350
MY COMMENTS: Listed on line without abstracts or comment. Other articles give abstracts. The title indicates that this is not a validation of graphology but a study of behavior associated with a medical condition. And it’s two pages long. Not indicative of research; more indicative of recorded observation.
---
REFERENCE: Broom, M E, Thompson, B & Bouton, M T
Sex Differences in Handwriting
Journal of Applied Psychology
1929, 13, 159-166
MY COMMENTS: Not available online. No references to it at all except in your link.
Without the article it is impossible to know if the author suggests anything beyond determining gender through graphology; even that much is in doubt as the author may very well be exploring how, at the time, boys and girls were taught differently and so learned different styles of penmanship (conjecture on my part). And look at the year. 1929?
---
REFERENCE: Campbell, D T & Fiske, D W:
Convergent and Discriminant Validation by the Multitrait-MultiMethod Matrix.
Psychological Bulletin
1959, 56, 81 – 1105.
MY COMMENTS: Mentioned on PubMed but without an abstract or text.
Discussions of the matrix in other articles, though, reveal that it does not discuss graphology.
---
REFERENCE: Castelnuovo-Tedesco, P
Ratings of Intelligence and Personality from Handwriting
American Psychologist
MY COMMENTS: Can’t find it online nor a reference to it except on your link.
---
REFERENCE: Cattell, Raymond B
A Culture Free Intelligence Test I
Journal of Educational Psychology
March 1940: 31(30, 161 – 170
MY COMMENTS: This test is an intelligence test, not a personality test. And it uses simple geometric drawings with simplified instructions provided, not handwriting.
---
REFERENCE: Crumbaugh, J C & Stockholm, E
Validation of Graphoanalysis by ‘Global’ or ‘Holistic’ Method.
Perceptual and Motor Skills
1977, 44, 03 – 410.
(Crumbaugh 1977)
MY COMMENTS: Text is not available, but the abstract is. Actually, the abstract sounds promising. It says it did five trials each involving five subjects who had a personality profile developed by a graphologist based on handwriting samples. The profiles were then randomly distributed to five other people who each knew the five subjects and were asked to match them up. Results were reportedly significantly above chance. I’d like to see this actual paper. There is, off the top of my head, at least one potential fatal flaw: Did the graphologist know/meet the subjects or develop the profiles based on blinded analyses? If not blinded, then the profiles could have been based on observation more so than on graphology. Just a thought.
---
REFERENCE: Crumbaugh, J C
A Reply to ‘Validity and Student Acceptance of A Graphoanalytic Approach to Personality.’
Journal of Personality Assessment.
1977, 41, 351 – 352
MY COMMENTS: I couldn’t find it, but the title indicates it’s not a study.
---
REFERENCES: Dennis, Wayne
Handwriting conventions as determinants of human figure drawings
Journal of Consulting Psychology
1958: 22, 293 – 295
&
Dennis, Wayne & Raskin, Evelyn
Further evidence concerning the effect of handwriting habits upon the location of drawing.
Journal of Consulting Psychology
1960, 24, 548 - 549
MY COMMENTS: Neither is available online. Look, however, at the dates and the lengths. But most importantly, look at the titles. It seems clear that that these are not about personality determination based upon handwriting but upon the correlation of handwriting styles with the placement of drawings upon the paper.
---
REFERENCE: Eliasberg, Wladimir:
Political Graphology
Journal of Psychology
1943, 16, 177 -201
MY COMMENTS: It only shows up on the link; nowhere else. Also, look at the date (1943, the height of World War II and xenophobia) and compare with the title. Tell me this is not an ideologically driven study, if indeed it is a study at all.
---
REFERENCE: Eliasberg, Wladimir:
Methods in graphological diagnostics.
Psychiatric Quarterly
1952, 26,l 399 – 413
MY COMMENTS: Only found on your link
---
REFERENCE: Epstein, L., Hartford, H & Tumarkin, I
The relationship of certain letter form variants in the handwriting of female subjects to their education.
Journal of Experimental Education
1961, 29(4), 385 – 392
MY COMMENTS: Only found on your link. But read the title carefully. Nothing to do with personality determination. Everything to do with the correlation between handwriting and education level. I’ll buy that possibility but pose myself as a counter-example.
---
REFERENCES: Epstein, S
The Stability of Behavior: I
Journal of Personality and Social Psychology
1979, 37, 1097 – 1126
&
The Stability of Behavior: 2
American Psychologist
1980, 35, 790 – 806
MY COMMENTS: These have nothing to do with graphology.
---
REFERENCE: Eysenck, H. J. & Eyseneck, S. B. G.
Manual of the Eysenck Personality Inventory
London: University of London Press: 1964
MY COMMENTS: There’s no graphology in this
---
REFERENCE: Eysenck, H.J.
Graphological analysis and psychiatry: An Experimental Study
British Journal of Psychology
1945, 35, 70 – 81
MY COMMENTS: Not available on line, but other sources indicate poor protocols, i.e., Eysenck asked a graphologist, based on handwriting samples, to predict how subjects would answer 27 questions (presumably yes/no questions). The graphologist got 62% correct (far above the 50% expected). But the graphologist had been previously told that all the subjects were diagnosed neurotics. In other words, the graphologist already had outside information that would allow him to predict their answers.
---
---
To summarize:
I looked at the first 19 English language references on your list.
Eleven do not refer to graphology or are not a study of it.
One is a study published in a non-refereed journal and is not available on line.
Five are possibly studies published in journals but not available on line. This includes one which is almost certainly ideologically driven and another by the same author.
One is a study of graphology with poor protocols but which is not available on line.
One is a promising study of graphology that may or may not have poor protocols which is not available on line.
The two actual studies are dated 1945 (for the poor protocol one) and 1977.
The five studies we can’t tell anything about date back to 1943.
The non-refereed journal study is from 1986.
---
Not sure about you, but this does not lead me to lend much credence to the list or the discipline.
I can’t go through the entire list of references on the one link that provides references, but I started at the top of the list and went down as far as time allowed.
I skipped those available only in French.
Here’s what I found (it may help you to learn not to accept people at their word and to check out references even if presented in an apparently professional manner):
---
REFERENCE: APA Guidelines for Educational and Psychological Testing
American Psychological Association
APA: 1974
MY COMMENTS: These are guidelines for testing. They are not research. And they do not address the efficacy of graphology.
---
REFERENCE: Barrow, N K and Scott, R H
Validation of a personnel Selection System to meet EEOC Guidelines
Journal of Handwriting Psychology
1984, 1 (1), 15 – 17
MY COMMENTS: Not available on line
---
REFERENCE: Baruch Nevo
Graphology Validation Studies in Israel
1986
MY COMMENTS: Not available on line and not listed on Baruch’s online C.V. A similar paper is, though, but is listed as being published in a non-refereed journal.
---
REFERENCE: Beumont, P.
British Journal of Psychiatry
Small handwriting in some patients with anorexia nervosa
1971. 119, 349-350
MY COMMENTS: Listed on line without abstracts or comment. Other articles give abstracts. The title indicates that this is not a validation of graphology but a study of behavior associated with a medical condition. And it’s two pages long. Not indicative of research; more indicative of recorded observation.
---
REFERENCE: Broom, M E, Thompson, B & Bouton, M T
Sex Differences in Handwriting
Journal of Applied Psychology
1929, 13, 159-166
MY COMMENTS: Not available online. No references to it at all except in your link.
Without the article it is impossible to know if the author suggests anything beyond determining gender through graphology; even that much is in doubt as the author may very well be exploring how, at the time, boys and girls were taught differently and so learned different styles of penmanship (conjecture on my part). And look at the year. 1929?
---
REFERENCE: Campbell, D T & Fiske, D W:
Convergent and Discriminant Validation by the Multitrait-MultiMethod Matrix.
Psychological Bulletin
1959, 56, 81 – 1105.
MY COMMENTS: Mentioned on PubMed but without an abstract or text.
Discussions of the matrix in other articles, though, reveal that it does not discuss graphology.
---
REFERENCE: Castelnuovo-Tedesco, P
Ratings of Intelligence and Personality from Handwriting
American Psychologist
MY COMMENTS: Can’t find it online nor a reference to it except on your link.
---
REFERENCE: Cattell, Raymond B
A Culture Free Intelligence Test I
Journal of Educational Psychology
March 1940: 31(30, 161 – 170
MY COMMENTS: This test is an intelligence test, not a personality test. And it uses simple geometric drawings with simplified instructions provided, not handwriting.
---
REFERENCE: Crumbaugh, J C & Stockholm, E
Validation of Graphoanalysis by ‘Global’ or ‘Holistic’ Method.
Perceptual and Motor Skills
1977, 44, 03 – 410.
(Crumbaugh 1977)
MY COMMENTS: Text is not available, but the abstract is. Actually, the abstract sounds promising. It says it did five trials each involving five subjects who had a personality profile developed by a graphologist based on handwriting samples. The profiles were then randomly distributed to five other people who each knew the five subjects and were asked to match them up. Results were reportedly significantly above chance. I’d like to see this actual paper. There is, off the top of my head, at least one potential fatal flaw: Did the graphologist know/meet the subjects or develop the profiles based on blinded analyses? If not blinded, then the profiles could have been based on observation more so than on graphology. Just a thought.
---
REFERENCE: Crumbaugh, J C
A Reply to ‘Validity and Student Acceptance of A Graphoanalytic Approach to Personality.’
Journal of Personality Assessment.
1977, 41, 351 – 352
MY COMMENTS: I couldn’t find it, but the title indicates it’s not a study.
---
REFERENCES: Dennis, Wayne
Handwriting conventions as determinants of human figure drawings
Journal of Consulting Psychology
1958: 22, 293 – 295
&
Dennis, Wayne & Raskin, Evelyn
Further evidence concerning the effect of handwriting habits upon the location of drawing.
Journal of Consulting Psychology
1960, 24, 548 - 549
MY COMMENTS: Neither is available online. Look, however, at the dates and the lengths. But most importantly, look at the titles. It seems clear that that these are not about personality determination based upon handwriting but upon the correlation of handwriting styles with the placement of drawings upon the paper.
---
REFERENCE: Eliasberg, Wladimir:
Political Graphology
Journal of Psychology
1943, 16, 177 -201
MY COMMENTS: It only shows up on the link; nowhere else. Also, look at the date (1943, the height of World War II and xenophobia) and compare with the title. Tell me this is not an ideologically driven study, if indeed it is a study at all.
---
REFERENCE: Eliasberg, Wladimir:
Methods in graphological diagnostics.
Psychiatric Quarterly
1952, 26,l 399 – 413
MY COMMENTS: Only found on your link
---
REFERENCE: Epstein, L., Hartford, H & Tumarkin, I
The relationship of certain letter form variants in the handwriting of female subjects to their education.
Journal of Experimental Education
1961, 29(4), 385 – 392
MY COMMENTS: Only found on your link. But read the title carefully. Nothing to do with personality determination. Everything to do with the correlation between handwriting and education level. I’ll buy that possibility but pose myself as a counter-example.
---
REFERENCES: Epstein, S
The Stability of Behavior: I
Journal of Personality and Social Psychology
1979, 37, 1097 – 1126
&
The Stability of Behavior: 2
American Psychologist
1980, 35, 790 – 806
MY COMMENTS: These have nothing to do with graphology.
---
REFERENCE: Eysenck, H. J. & Eyseneck, S. B. G.
Manual of the Eysenck Personality Inventory
London: University of London Press: 1964
MY COMMENTS: There’s no graphology in this
---
REFERENCE: Eysenck, H.J.
Graphological analysis and psychiatry: An Experimental Study
British Journal of Psychology
1945, 35, 70 – 81
MY COMMENTS: Not available on line, but other sources indicate poor protocols, i.e., Eysenck asked a graphologist, based on handwriting samples, to predict how subjects would answer 27 questions (presumably yes/no questions). The graphologist got 62% correct (far above the 50% expected). But the graphologist had been previously told that all the subjects were diagnosed neurotics. In other words, the graphologist already had outside information that would allow him to predict their answers.
---
---
To summarize:
I looked at the first 19 English language references on your list.
Eleven do not refer to graphology or are not a study of it.
One is a study published in a non-refereed journal and is not available on line.
Five are possibly studies published in journals but not available on line. This includes one which is almost certainly ideologically driven and another by the same author.
One is a study of graphology with poor protocols but which is not available on line.
One is a promising study of graphology that may or may not have poor protocols which is not available on line.
The two actual studies are dated 1945 (for the poor protocol one) and 1977.
The five studies we can’t tell anything about date back to 1943.
The non-refereed journal study is from 1986.
---
Not sure about you, but this does not lead me to lend much credence to the list or the discipline.
Thanks again everyone
Seems like my arguments aren't holding up too well so far
Thanks for looking at the references, Garrette. I'm quite disappointed at the results, if what you state is true.
As I said, I hadn't had time to look at them when I posted the links.
Let's see.. where do I disagree with everyone so far.
Well let's start with the bit about being able to make a broad statement that 95+ of the population would accept as truth.
1. I have NEVER received direct feedback from the subject themselves, but rather the opinion of the acquaintance. As such, the person providing the feedback generally had no desire to mince words. Aside for pleasing me, perhaps (a possible problem)
2. I NEVER do double positive statements of the sort found in reference to Forer.
3. I do NOT keep back the blatant negative statements as Forer demonstrated. 'A very rebellious, opinionated and aggressive person' is not positive by any means.
In regard to me 'promoting it as science' - I am not personally claiming it to be a science based on no real evidence.
I was simply investigating the general defence of graphologists - who state that it is a science. I'd have expected them willing to back up a dangerous statement like that, and found that list of references. As mentioned at the time, I hadn't had the chance to investigate them yet and as such had no idea how valid they were.
As it turns out, it seems to be baloney. As Garrette states, 11 of 19 of those references seem to be unrelated to graphology ... quite a silly attempt by the author to validate it as a science.
Alright, let's summarise:
a. Thus far it seems graphology is NOT a science and has not been sufficiently researched/experimented upon in it's development. (This does not necessarily mean the science is impossible to develop however)
b. My feedback would likely have been slanted by those who wished to provide me with a positive response.
c. Any positive analysis would result in Forer effect. Keep in mind however that not all of my analyses were on a positive note.
It is certainly leaning towards BS thus far
But I'm still not entirely convinced and would like to try a few experiments when my life calms a little.
I'll let everyone know
[EDIT]
On a different note - I'd like to explain how an 'obviously intelligent' (thanks for that
) person is able to believe in something without evidence.
It's quite simple - I wasn't as gifted as Randi to be a skeptic from birth
As a child, my mother had the book lying around. I tried it out, and found it worked. As we've discussed, it does seem to work - but possibly not for the reason I believed.
My childlike mentality simple accepted it without too much question.
Also keep in mind that I didn't have such an invaluable resource as the internet to delve into for answers
Assuming we don't find any other interesting information, so far it seems to me that I was fooled.
Seems like my arguments aren't holding up too well so far
Thanks for looking at the references, Garrette. I'm quite disappointed at the results, if what you state is true.
As I said, I hadn't had time to look at them when I posted the links.
Let's see.. where do I disagree with everyone so far.
Well let's start with the bit about being able to make a broad statement that 95+ of the population would accept as truth.
1. I have NEVER received direct feedback from the subject themselves, but rather the opinion of the acquaintance. As such, the person providing the feedback generally had no desire to mince words. Aside for pleasing me, perhaps (a possible problem)
2. I NEVER do double positive statements of the sort found in reference to Forer.
3. I do NOT keep back the blatant negative statements as Forer demonstrated. 'A very rebellious, opinionated and aggressive person' is not positive by any means.
In regard to me 'promoting it as science' - I am not personally claiming it to be a science based on no real evidence.
I was simply investigating the general defence of graphologists - who state that it is a science. I'd have expected them willing to back up a dangerous statement like that, and found that list of references. As mentioned at the time, I hadn't had the chance to investigate them yet and as such had no idea how valid they were.
As it turns out, it seems to be baloney. As Garrette states, 11 of 19 of those references seem to be unrelated to graphology ... quite a silly attempt by the author to validate it as a science.
Alright, let's summarise:
a. Thus far it seems graphology is NOT a science and has not been sufficiently researched/experimented upon in it's development. (This does not necessarily mean the science is impossible to develop however)
b. My feedback would likely have been slanted by those who wished to provide me with a positive response.
c. Any positive analysis would result in Forer effect. Keep in mind however that not all of my analyses were on a positive note.
It is certainly leaning towards BS thus far
But I'm still not entirely convinced and would like to try a few experiments when my life calms a little.
I'll let everyone know
[EDIT]
On a different note - I'd like to explain how an 'obviously intelligent' (thanks for that
) person is able to believe in something without evidence.It's quite simple - I wasn't as gifted as Randi to be a skeptic from birth
As a child, my mother had the book lying around. I tried it out, and found it worked. As we've discussed, it does seem to work - but possibly not for the reason I believed.
My childlike mentality simple accepted it without too much question.
Also keep in mind that I didn't have such an invaluable resource as the internet to delve into for answers
Assuming we don't find any other interesting information, so far it seems to me that I was fooled.
Don't worry about it.
You were not the first and won't be the last intelligent person to fall for something that isn't true.
I was a big believer in things such as this (though, strangely, never graphology).
I think many others on this board are/were in the same boat.
There is no shame whatsoever in being fooled. Actually, I think fooled is far too strong a word. You merely believed what was in a book that you came across as a child and have not, until now been confronted with any reason to suppose that the information contained within it was inaccurate. That is absolutely normal. What is out of the ordinary is that you ever came to question such a belief at all. That is, indeed, intelligent. Even more remarkable is that you are able to consider views opposed to your own and to modify your position according to those views which you find reasonable. In my experience, that is a rare and admirable thing.
By the way, I think you might want to look again at the Forer effect. It is not the case that we are only likely to identify with descriptions which we consider positive, there are many apparently negative characteristics with which we all tend to identify. Neither is it particularly relevant, in terms of the effect, whether you delivered your analysis in person or through a third party, though had you done the former there would have been a whole range of additional conflating factors including, as you identified, the tendency of subjects to want to please the experimenter.
Oh, and I was puzzled by you promoting a part trick to a science not as a science. It appeared to me that you were treating it as a science (albeit it an imperfect one) not that you were in any way proselytising.
Anyway, congratulations on having the intelligence to question your own beliefs and the courage to do so in such an open manner which can only benefit all of us.
By the way, I think you might want to look again at the Forer effect. It is not the case that we are only likely to identify with descriptions which we consider positive, there are many apparently negative characteristics with which we all tend to identify. Neither is it particularly relevant, in terms of the effect, whether you delivered your analysis in person or through a third party, though had you done the former there would have been a whole range of additional conflating factors including, as you identified, the tendency of subjects to want to please the experimenter.
Oh, and I was puzzled by you promoting a part trick to a science not as a science. It appeared to me that you were treating it as a science (albeit it an imperfect one) not that you were in any way proselytising.
Anyway, congratulations on having the intelligence to question your own beliefs and the courage to do so in such an open manner which can only benefit all of us.