Sorry about the spelling, Achernar.
By the way, that’s a very cool site you have. Including the handwriting sample (the content, not the writing).
Okay, fair enough. I admit I haven’t looked at any of them either. But I’ll try to. And I’ll come back when I have something definite to point to.
The main issue is that none of them were online. So accessing them is tough.
I think my University has a library or two. 
Maybe sometime this weekend. I need to return this book I checked out anyway. I think it’s about two months overdue… Well, if you don’t hear from me again, the Library Cops got me.
Archernar: << Just curious, but does this include things like the NEO PI-R? >>
That’s a different discussion, Arch. The terms are (reasonably) well-defined and the results are repeatable, both of which are different from party-trick pop-psych stuff.
Bill H, sorry if I was unclear, and I apologize for the sarcasm. I was trying to draw a line between a serious study and a parlor trick. My sarcasm and derision were devoted to the latter, not the former.
If we’re working on the question of whether graphology can yield measurable, quantifiable results, I’m all for it and I’ll help out and be as objective as I can. If we’re working on the question of whether graphology can provide amusement at a party, like tarot readings or horoscopes or psalmistry or the silly “Take this Personality Quiz” from TEEN MAGAZINE every few weeks… well, I personally don’t have much interest.
And I do understand the huge time commitment you’re offering… which is why I said initially I thought it was impractical. If you’re being nice enough to offer that much time, we should be sure that we’re doing a test that will come up with some interesting results (win, lose, or draw)… and not a test that will be dismissed later as merely clever use of ambiguous personality trait adjectives.
The problem with those horoscope-like party tricks is that a believer comes away saying, “Wow! look how accurate that was!”
and a sceptic comes away saying, “What generalized, ambiguous, flummery!”
So, I guess I’m asking for clarification, Bill. You yourself have used terms like “amusing party trick” and “You can’t tell the magician what your card is. It wrecks the trick.” If that’s what this is about, then I’m not really interested. Let’s clarify what you’re going to do before you get there?
I’m not sure which of us you were trying to insult, but it doesn’t really matter. I certainly am not insulted by such a comparison, because Dex and I have been saying essentially the same thing from the beginning of this snake dance: namely, that claims of ability to discern “personality traits” via handwriting are horse puckey on a level with astrology or phrenology or any other such mumbo-jumbo. The primary reason it’s “bovine excrement” as Dex says, and as he has said clearly, is because saying someone has some certain “personality traits” is just blather, so vague as to be completely meaningless. Saying “I see you have a tendency to be tenacious” or “I sense that you’re a creative person” is simply hot air, whether you sit in a dimly lit parlor and read Tarot cards or fondle the bumps on someone’s skull or stare intently at their handwriting with a look of deep concentration. Such assertions are unprovable, ill-defined and wholly subjective, which anyone who claims to be “a reasonable scientifically-minded person” should know and be willing to acknowledge.
You may truly believe in your snake oil, but your persistent complaints about being picked on (“Accusing me of stereotyping and racism. . .” when no one had “accused” you of any such thing, and so forth) are pretty transparent attempts to ingratiate yourself with the marks and get their sympathy, thus keeping them from noticing how you palm your metaphorical card with phrases like “I need the ability to discard subjects and writings” (thus putting everything under your control, and removing any possible objectivity from the whole business). You’re a pretty good snake oil salesman, with a true carnie’s subtle touch, and I really don’t care who swallows your pitch, but if you’re going to try to attach words like “science” and “objectivity” to your show, you have to expect some people to call you on it. Acting all offended-like doesn’t give you any extra cachet of legitimacy.
Yo, Cobweb. Are you going to contribute anything constructive to this discussion, or are you just going to sit on the sidelines taking potshots? Obviously you have “issues” with graphology. May I suggest that you take them to the Pit? Start a “Flame the Handwriting Expert Here” thread or something.
Bill has put himself way out on a limb here, and I’m sure he knows it perfectly well, having spent nearly three years at the Dope. He knows what goes on, and what can happen. I applaud his spirit of scientific inquiry, as it’s extremely likely that if Irishman and Chronos and Dex manage to get this show on the road, we may ultimately end up proving to him that his chosen specialty really is nothing more than hot air.
You’re making a big deal out of his expressed need to discard entries, but here’s what he actually said:
My understanding of his points here is that first, what he wants to do is use the handwriting samples to illustrate how handwriting reflects personality traits, but he’s worried that if he gets 20 samples from 20 GQ Google-Fanatics, and if one makes the not unreasonable assumption that all GQ Google-Fanatics are detail-oriented, anal-retentive trivia addicts, then all 20 GQ samples that he receives in the mail might reveal to his analysis 20 virtually identical people, which won’t enable him to demonstrate anything about graphology, and it will be boring and pointless for him to do the analysis. “Number 17…Yep, here’s another one who shits a brick at typos in ad copy…”
And second, he wants the ability to discard or discount certain subjects because their real life persona might turn out to be totally different from their SDMB persona, which would be something that wouldn’t come out until afterwards, in the e-mail and posts to the thread, after he had published his analyses.
I don’t see that either of these stipulations would be a problem. For the first, we could just get the talent pool from sources other than “Just GQ” or “Just Comments”. There are a number of long-time posters who spend all their time in MPSIMS, Cafe Society, IMHO, and the Pit, and of course there’s a large population of “Great Debates Is My Life” Dopers in permanent residence down there, and I’d defy anyone to characterize a “typical Great Debater”. They’re all over the board.
For the second, as long as he was honest about it and told us which subjects he was discarding and why (“Well, I said this one was Elucidator, and it actually turned out to be Minty Green…”), I don’t have a problem with it. Because that would actually be part of the proof that graphology is just hot air.
It may seem like a different discussion at this point, but I’ll bet that it’s more or less what Cecil had in mind when he used the term “personality”.
C K Dexter Haven wrote
No problem.
Absolutely. That’s my goal too.
Absolutely. Let me clarify what I meant earlier.
There is a possiblity (and I don’t know how big, perhaps .1%, perhaps 80%. The possibility that it could be 80% makes me very nervous…) that I’ll receive a set of samples that are too similar for me to differentiate.
If that happens, I may be forced to say “I can’t do this.” Many people, myself included will be disappointed, and many people will consider me a fraud if this happens. But in the chance it does, I will show exactly what I mean by similar. i.e. what parts of the handwriting were too close to differentiate. And I’ll show examples of other non-test handwriting that does differentiate. That should at least give others (hopefully) an understanding of why things failed.
And… if that happens, I’ll explain what handwriting analysis claims is the meaning of those similar traits in all of the samples. Which is not pure science, I’ll agree. But at this very moment, I have zero doubt personally of what those traits will actually be. And if it turns out (as I suspect) that those traits happen to match with traits a reasonable person would attribute to a CCC reader, I don’t think that’s irrelevant information.
Now…
All that said, I hope that we do get a good varied group, and this is a reasonable test. I hope I’m worried with no reason.
You know, as I typed that line about magic tricks, I thought someone would say this, and I wavered. I wondered if it were best just to say “It’s not scientific for me to accept a sample with a pre-known, unique trait that I’m 100% certain matches to one of the subjects”, and leave it at that.
But I figured, “well, I am explaining this on the objective scientific side, and the objective scientific people will get my meaning, and perhaps those inclined to humor may like the magic reference.” And I still hope they do.
BTW, for those who didn’t get my earlier attempts at humor, here are some Cliffs Notes:
[ul]
[li]The reference to tea leaves predicting the sharpness of razor blades was a combination of two silly notions, a) that looking at the tea leaves that are left after drinking tea can predict one’s future, and b) that putting a razor blade under a pyramid will keep it sharp.[/li][li]The query to Cecil for his cites, saying Where’s the beef? is a reference to a humerous television commercial produced by a hamburger-selling establishment named “Wendys”, wherein a little old lady wondered why the burger was so small.[/li][li]The reference to Cecil’s claimed cites as studies in a briefcase was a reference to Senator Joe McCarthy, who claimed to have a list of communists in a briefcase.[/li][li]When I wondered how I ended up drafted into some serious effort when the people who are defining the rules are already discarding the results, I used the phrase How did I get here? This is also a line from a song by a group known as “The Talking Heads.”[/li][/ul]
Hope this helps.
Oh, one more:
[ul]
[li]That last list of explanations was another attempt at humor; you can add it to the list.[/li][/ul]
And that last explanation as well.
Uh, and so on.
Thanks for the explanations. Were you also being humorous when you said "it’s just an amusing party trick " in your opening post? Because, for me, that’s the crux of the issue.
I certainly understand your need to reject samples. No problem there. But I’d still like clarification: are you going to try to identify the writers, or are you only going to assign personality traits (“ambitious”, “creative”, “arrogant”, “intelligent”) to them?
My alleged “potshots” are constructive. It’s always constructive to point out flummery and foolishness, and it’s always appropriate to ridicule the ridiculous. My posts say the same things that others have been saying throughout this thread, such as:
The point being made is that claims of graphology being of any use in determining anything about the personality of the writer are nonsense, on equal footing with astrology, palmistry or tossing chicken bones. Cecil stated it quite succinctly in his original column, and numerous legitimate studies have been cited. Graphology is twaddle.
Bill H. claims otherwise, but admits:
When there is an abundance of evidence of bunk and not one jot of evidence that the claim is anything but bunk, the burden of proof is on the claimant. Bill H. says there is no evidence that he knows of in his favor, and refuses to admit any evidence against him as legitimate, claiming it’s “biased”. That’s not “scientific”; it’s standard paralogia for pseudo-science aficionados: deny all the evidence with a wave of your hand, and then press on.
Then when he’s called on it, the claimant resorts to complaints of being picked on, such as:
Note, nobody ever accused the poor downtrodden fellow of stereotyping or racism; he’s merely playing the “poor little picked-on me” card. And observe how he’s twisted the entire matter onto its head: suddenly he’s the poor beleaguered victim who, though he is the one making claims contradictory to all evidence, is demanding cites for more studies while admitting that he has none on his side.
This is scientific method? Deny the validity of all legitimate studies (they’re “biased” of course, by definition, because they don’t confirm the claimant’s assertions), use the fact that absolutely no evidence exists to back your claims as a selling point, demand more citations of more studies (which can conveniently be dismissed as “biased” or via some other twaddle), and, when anybody has the temerity to call you on it, whine about being picked on and turn yourself into a victim.
Sorry, but paralogia, nonsense and folderol don’t impress me, and my calling nonsense and folderol nonsense and folderol isn’t “taking potshots”, it’s simply pointing out in plain language that this emperor has no clothes.
This is a heckuva interesting discussion, but there are underlying questions that I feel the need to bring up. The big two are at the end of the post, above the cites.
Is there a real relationship between the proposed “testing” of Bill H. and anyone’s (dis)belief in graphology? Would anyone’s position on graphology be altered (even partly) by the results of this test?
With regards to the original question (a legitimate interest in seeing some citations for the 200-odd studies disproving the validity of graphology): Is there such a thing as a set of studies with results for/against graphology
that would be accepted as valid by the persons on this thread?
My own experience has been that the number of studies found and cited has little impact on the final beliefs of the interested people. I can back that up with cites, actually: there are multiple studies that have demonstrated
that belief perseverance is still about 75% when (1) the belief has no known past emotional investment for the people tested and (2) the belief is shown to be completely false. I’ll put the cites for the studies involved at the end of this post, so I can get to my main questions:
My question for Bill H in particular and the group in general, then, is this: exactly what level of demonstration or number of studies would it take for you to change your mind on the topic of graphology?
Also, for Bill H. specifically: if Cecil showed up and listed 200 studies, how many of them would it actually take to change your mind? 10? 50? All 200? Would 200 be enough?
Cites for perservance of beliefs (more cites available on request):
Anderson, C. A., Lepper, M. R., and Ross, L. (1980). Perserverance of social theories: the role of explanation in the persistence of discredited information. Journal of Personality and Social Psychology, 39, 1037-1049.
Davies, M. F. (1997). Belief persistence after evidential discrediting: The impact of generated versus provided explanations on likekihood of discredited outcomes. Journal of Experimental Social Psychology, 33, 561-578.
See? You can be constructive, if you try.
On the one hand, I am genuinely interested in any replies people have time to make to my questions re the perserverance of beliefs, and so wanted to be all detached and (gawd help me) reasonable. Also, I’m new, so I don’t want to be too lamely newbie in my posts.
Nonetheless, with regards to Dr. Cobweb’s most recent post, I can’t help saying, “Dude! Awesome!!! Preach it, brother!!!”
[Oh, the shame of it! To be “postin ‘Me too!’ like some brain-dead AOL-er!”]
Excellent Weird Al quote!!!
Not necessarily. Handwriting starts with a norm, which makes it a learned performance. Its basic forms are imposed upon the writer in grade-school penmanship classes, although less these days than, say, 50 years ago–but still, the forms have to be learned. Once people have acquired a degree of manipulative skill they refashion the forms in individual ways as an expression of personal style. It’s the deviations from the norm that handwriting experts look for–both to discern (not “predict”) personality traits, for whatever reason, and to compare handwriting so they can see, for instance, if the person who claims not to have signed the check really did sign it.
Job applicants these days get a lot of personality tests, such as Myers-Briggs and others, and most of them are easy enough to see through as they’re being taken (in other words, if you’re looking for a sales job and you know the sales manager is looking for an extravert, it’s pretty easy to decode the test questions and slant your answers that way). And many other factors go into whether you’ll be successful at sales in general, or that job in particular. You couldn’t safely predict it from handwriting analysis, but you couldn’t safely predict it from a standard personality test, either.
I don’t understand this. You’re saying that people don’t have personality traits?
What’s hard is quantifying personality traits. It may be obvious from two handwriting samples that A is more tenacious than B, or that C has more indications of deception/dishonesty than B. Taken in isolation these mean nothing–although they could mean that A will do anything to succeed and B could be disloyal and C will steal office supplies. Or not.
But that’s the problem with an experiment like the one proposed here. Bill is going to take samples and match them with people, but these are the same people who are going to decide, in some equally unscientific way, whether they are more or less tenacious (let’s say) than some other person they don’t even know? I don’t see how that could possibly do anything but fail.
Not necessarily. Handwriting starts with a norm, which makes it a learned performance. Its basic forms are imposed upon the writer in grade-school penmanship classes, although less these days than, say, 50 years ago–but still, the forms have to be learned. Once people have acquired a degree of manipulative skill they refashion the forms in individual ways as an expression of personal style. It’s the deviations from the norm that handwriting experts look for–both to discern (not “predict”) personality traits, for whatever reason, and to compare handwriting so they can see, for instance, if the person who claims not to have signed the check really did sign it.
Job applicants these days get a lot of personality tests, such as Myers-Briggs and others, and most of them are easy enough to see through as they’re being taken (in other words, if you’re looking for a sales job and you know the sales manager is looking for an extravert, it’s pretty easy to decode the test questions and slant your answers that way). And many other factors go into whether you’ll be successful at sales in general, or that job in particular. You couldn’t safely predict it from handwriting analysis, but you couldn’t safely predict it from a standard personality test, either.
I don’t understand this. You’re saying that people don’t have personality traits?
What’s hard is quantifying personality traits. It may be obvious from two handwriting samples that A is more tenacious than B, or that C has more indications of deception/dishonesty than B. Taken in isolation these mean nothing–although they could mean that A will do anything to succeed and B could be disloyal and C will steal office supplies. Or not.
But that’s the problem with an experiment like the one proposed here. Bill is going to take samples and match them with people, but these are the same people who are going to decide, in some equally unscientific way, whether they are more or less tenacious (let’s say) than some other person they don’t even know? I don’t see how that could possibly do anything but fail.
To reiterate, Cicada2003, the test is not to determine whether Bill H. can tell us who’s tenacious or sensitive or whatever. As you point out, that would be nearly impossible to scientifically control. The test is to determine whether he can match handwriting samples to people. This is, in fact, objectively measureable. If he looks at my handwriting and says that it belongs to Dex, that’s wrong. If he looks at my writing and says that it belongs to me, that’s right. Repeat for some appropriate number of samples, and the test starts becoming meaningful.
As for the “parlor trick” comments, suppose that Bill H. is, in fact, able to do what he claims. Suppose that he matches all ten (or however many) samples correctly. That would certainly be interesting, but of what use is it? I think that it is fair to call the level of ability Bill claims a parlor trick. But the key is, it’s a potentially measureable parlor trick.
And Achernar, what are the problems you see with the test setup, as currently established? And what would you recommend to improve the test?
Check. Check. Is this on? Did anybody bother to read my post at the bottom of the first page? Because I swear I addressed Dex’s concerns about the point of the test in that post.
Second bolding added for emphasis.
I certainly agree with you a list of personality traits alone would read like a horoscope or palm reading or talking with the dead episode. That’s why I specifically stated that he must match the writer with their sample.
I said some other things in the post that I think are important that nobody has commented one way or another on. Especially the last paragraph about ramifications of the results of this test.
Bill H., you originally asked for cites. While we’re still waiting for Cecil to provide his list of 200, I think DDG did a pretty good job with the list she gave. True, there are only two links, but the citations are there. Now your quibble is that they aren’t online? That seems specious. Is it harder for you to look them up and see what they say? Sure. But they’re still citations of studies on graphology, including numerous articles in peer-reviewed psychology journals. So I think it’s a little unfair to say nobody has posted cites. (Note: 2 sites, but 23 cites. Don’t confuse the terminology.)
Bill H. said:
Please please take a look at my post at the bottom of the first page and give me some feedback. I’m particularly interested in if you feel what I’ve said accurately sums up what you think the scope and ramifications for this test are. I would like to see this go forward, but I think it important we agree about what we’re doing and why, and if that is enough to make it worth your particular effort.
DDG makes some good points about how to try to get a diverse enough sample pool so Bill H. feels the test is fair. I don’t think it would be that difficult. The trick to me seems to be limiting the number of respondents. If Dex got us an announcement at the top of every forum, he’d be swamped with replies, and then have the logistical nightmare of sorting who read the announcement in which forum. I’m trying to think of a better way to get a diverse sample. Off the top of my head, I say that several of us can think of a poster or two in other forums that have their own distinct style. Maybe a personal request via email to them and invite them to participate?
Does anyone else have any good ideas for making the candidate list diverse while managable to sort the applicants? Let’s hear them.
Dr. Cobweb said:
I think you should re-read the thread.
C K Dexter Haven said:
And the specific example Dex used related to blacks and crime. So while Dex didn’t quite accuse him of racism, he very much did accuse him of stereotyping, and I can see how Bill H. could feel accused of racism from that post.
So now I am asking you, lay off it. Several of us are trying to make this a useful learning experience, and all you’re doing is getting in the way.
Regarding Bill H. being able to eliminate writing samples, here’s my thought. After we cull the list of participants to, say 25, and they each submit two copies of the writing sample to Archenar, then when Bill H. receives the copies he is allowed 5 preremptory challenges for whatever reason he wants - too similar, to dishevelled, too sticky - whatever. That leaves 20 samples upon which to perform the actual test.
Regarding the non-graphologist control evaluator, my thoughts on that were to not include emails precisely to eliminate the use of personality information. However, I see your point about making the correspondence public afterwards in order for everyone to be able to verify nobody disclosed some particular trait about their writing, such as how they dot their i’s with hearts, or the aforementioned use of custom ink. Thus I can see how sharing those communications with the control would help eliminate those type of factors. To me this is still something to be decided, do we use this control or not. Let’s have clear discussion on this.
zut said:
I think it does. I think it adds a level of control that the pure statistical approach does not. Just like the placebo effect is a real effect that must be taken into account in medical trials, to show that the treatment not only beats statistics but placebo, in our study their does exist the small chance that non-personality data is being used to evaluate the posts. This control helps correct for that. I would propose that both people, Bill H. and the control (still Dex?) list their reasons for matching the samples the way they do. Again, the pass/fail criteria for the test is actually identifying the writer with the sample, not the list of personality traits. But the list of reasons for matching the way you do shows what is being looked at and why. This would be useful from the control. If all twenty end up being eeny meeny miney moe, then that’s one thing, but if he can say “pink ink with hearts over the i’s have to be a woman, and the only one of those was DDG” then that tells us something else.
I’m also trying to think about how to reassure Bill H. about the writing samples being on the up and up. I think it important that the participants not only write their samples like normal (and not use their left foot or whatever), but also that they don’t mislead Bill in any correspondence he uses to base his personality judgment. After all, if I write normally and then act differently than normal that wouldn’t be a fair test, either. I suggest the participants sign an affidavit that they will provide an accurate sample of their normal writing and that the information provided during correspondence is true and not intended to deceive. These affidavits could be a form written up here and signed, then mailed with the samples to Archenar. He would retain them until the conclusion of the test, and then they could be made available with the rest of the correspondence and writing samples.
Okay, who’s going to archive all this stuff? 
Should have previewed.
Jennifer McIlwee Myers, I wanted to respond to your post separately.
I think a well-constructed test with conclusive and duplicable results published in a peer-reviewed professional journal would go a long way to influencing my opinion. I would expect there to be criticism over how one study shows positive results when so many others have not.
I’m curious how our test will come out. I’m fairly certain given my going in position of what the outcome will be, but I could be surprised. That is why I’m wanting to make sure our test is as rigorous as we can make it, and I’m looking for any help I can find that will control the test for validity and eliminate post hoc justifications for why results turn out different than expectations (either way).
I would also ask you to read my post on the bottom of page one and answer your thoughts about the ramifications of the outcome of this test.