Undiagnosed aspie--RDOS aspie results attached

Post Reply New Topic

Of course I never expected you to personally mark every test



Questions don't just fall from the sky. They must be formulated and selected. It's obvious you are doing the selecting by the inclusion of the oh-so-bizarro "traps" question.



Please explain why anyone other than yourself thinks the quiz is legitimate or valid in any way.

The "about you" page has no effect on the scores. It is only used in the data analysis.

About 1,800 questions have been tested. Of these, 150 qualified for the final version. Selection has only been done based on what I posted before (correlation with score + low correlation to other questions). There is no question in Aspie Quiz because "I like it". This only exists in your fantasy.



I won't, and since you are not a professional, I won't send you the article about Aspie Quiz either. You simply wouldn't understand it, but would rather post more nonsense on the web.

That is what I meant. So do you rely on the answers on that page to analyze the accuracy of your questions? I see a number of problems with that. First of all you are relying on the accuracy of those answers. For instance someone who has AS but only suspects that may be the case may well answer 'self diagnosed' or 'NT'. The other problem is that the majority of people taking your test are probably either on the spectrum or suspect they are. I would have thought only a few NTs are likely to take the test unless they have a very good reason to do so.

I am not criticizing you for doing this. I can't see any way you can get a scientifically validated set of test subjects without access to research facilities and a fair amount of money.

Exactly. It appears the questions are somehow validated by the test, the results of which come from the questions. It's entirely circular. With no way to control or verify the status of those taking the test.

So I'm not a professional. As far as I can see your interest in this area is no more than a hobby. What professional body supports your work? Where has it been published?

I leave it up to you to check both.

No



No



As I wrote, I don't use this info for anything related to scoring. Not when somebody does the test, neither when I assign weights for questions, and certainly not for selecting questions. I already described how the question selection process works (correlation to score + as low correlation as possible with other questions). The weights for scoring are averages of factor loadings from factor analysis of previous answers to questions.



Not so. Today, most score NT. Only a little over 30% of participants score "very likely Aspie".



Not so.



Wrong again. If you don't have random samples and lots of money, you simply invent alternative approaches of doing things. That has nothing to do with science.

What I can see from your blog is that you've had a couple of failures with your peer review, so it doesn't look like your ideas are getting much traction.

But getting back to the questions in your quiz, I'm curious about the origins of the questions, specifically the "Have you been fascinated about making traps?" as you seem somewhat evasive. Like I said, the questions don't just spring forth from the void. Did you write the question or not?

Again, on another question "Do you have unusual sexual preferences?". What exactly is unusual? How is anybody supposed to answer that one? Without clarification I don't see how you can expect any meaningful answer.

To the OP

for "how do you read this thing"

I actually read through the entire report. It's pretty interesting to review your answers rearranged to see the way the data flows in the report. For me, I went through and did all the tests that are in the following thread. Each of them explains different things that are interesting to learn about yourself. Aspie or not!

http://www.wrongplanet.net/postt113459.html

for recovery-like scenarios:

If you were not too severe in the first place with symptoms and were able to build coping methods in your life in order to appear NT to the outside world then that proves on thing: that you can seem NT to the outside world. However if it takes significant extra effort to do all of that, then you just might be more ASD than you are realizing. I pretend to be NT all the time, its like playing a role. I know that I seem somewhat absurd to myself when I do it, and it requires a lot of important small decisions and intentional actions. Generally I cant pay as close attention to the subjects at hand while I do it but that's why I sort of switch it on and off.

-----------
To the aspie quiz discussion

Isn't rdos saying that the test is statistically related to other tests? (mainly mentioned is the AQ test) if that is the case, and this was done in a way to remove as much bias as possible then I would say that it is scientific. However it may not scientifically correlated to diagnosis, but of course it and all the other online tests pretty much say that they aren't there to diagnose you but instead to compare your traits with that of people on the spectrum... actually to quote directly:

While this one doesn't give as many disclaimers as the others (which perhaps is slightly irresponsible) anyone who is approaching an online quiz or test of any kind should have caution and realize that the only difference between an online form and a magazine is that an online form can do extra calculations. So it relies on you reporting correctly.

However, one thing I will say is that it does use more technology than quizzes like the AQ test. For example, I had a pretty high score right out the gate Aspie: 164, NT: 55 but I figured maybe there was bias since I had been reading a lot about asperger's at the time so all the traits were in my head. So I put it away for a few days and when I came back I answered conservatively. If I thought maybe it was too much to say yes on something I would say a little/maybe... and there were some where if they were rare enough or it seemed the question didn't fit exactly that I would say no...

It told me I was too inconsistent lol... I sort of liked that It knew that I was trying to botch my answers in the opposite directions and stopped me.

So maybe a better clarifying question to help people feel better about the test rdos...

What all happens when people sign up to help with calibration? Even if you don't want to get too deep into the statistical tests you use or data validation or whatever you might use I would be curious to learn more about it from a high level overview.

For the record, while I am not a psych professional, I am a professional Analyst with specific background in behavior and quality analysis so I am sure I'll understand whatever you throw out at me if it relates to data engineering and that sort of thing.

-------------------

As for diagnosis from online sources: the only source I would trust is the DSM IV and for forward thinkers the DSM V. If you think that both or either of those have too ambiguous of wording then you probably aren't suited to self diagnose based on their criteria alone and might need additional assistance from doctors, further reading, comparisons to professional dx people, or just hanging out here on WP a bunch. I read it as straight forward and consider it the only online resource that I could use for myself to define my self diagnosis.

I can understand that but how do you confirm that your results are correct? Tweaking the individual scores for maximum correlation to the overall score makes sense but how do you know the overall results aren't biased in one direction or the other?

30% is much higer than the population average so it appears you do have a bias in the people taking the test. This is to be expected as the vast majority of the general population don't know anything about ASD. In most cases people will need to have had some experience of ASD (personally or someone they know) to want to take the test.

Just out of interest, have you analyzed the results based on age? Most Aspies learn to compensate to a greater or lesser extent over time. This is a problem that is common to most of the tests out there. For instance questions like 'do you tend to go on about a particular subject' could get the answer 'no' because the person taking the test has learned that doing so causes problems. Although they want to go on about their subject they consciously hold themselves back.


There is nothing wrong with using alternative approaches but you do need to be able to confirm your results agains some sort of benchmark. A fundamental part of scientific research is being able to PROVE your results.

Who wrote the questions is not important. What is important is if the questions are meaningful for neurodiversity, and this question is meaningful for neurodiversity.



Another tool used in the selection of question is number of "?" responses. If those were above 10%, a question would not qualify. Since this question did qualify, it has below 10% of "I don't know" answers, and thus 90% or more answer to it. This should answer your question.

There are 5 control questions that you should answer opposite to. If you fail more than one, you will get this text.



If you like the manuscript sent to PLoS, which provides details for the construction and properties of Aspie Quiz, send me a PM with your email address, and I'll send it to you.

This is described in detail in the paper sent to PLoS, but in short, I've shown that the factors underlying scoring in Aspie Quiz are highly stable, and you can recreate them with any question selection as long as your test is diverse enough, and contains both Aspie and neurotypical questions in sufficient amount.



True. There naturally is a bias in participation. What I claimed is that there is not a minority NTs. NTs are actually a majority.



Aspie Quiz has no age bias, in the sense that factors and explained variance is similar regardless of age. Scores are not age related either, but young males on average score considerably lower, which is related to many more young males participating than expected. If older people score lower is not possible to show as participation is not controlled.



The benchmark is the two primary factors. The main objective was to show that these are natural properties of human variation (shown with factor congruence coefficients), and then provide a set of items that covered all aspects related to these factors. The usefulness of Aspie Quiz as a diagnostic tool for ASD actually was just a side-effect. Nothing in the development process cared about DSM criteria for ASD.

OK, I take it from your reluctance that you did write that question. Why is it meaningful for neurodiversity? You haven't explained or even provided any evidence that this is so. When I read the question the first time I was in hysterics - it is so far out. Can't you see that?



Not really. Everyone has their own standard of sexual preferences, and cultural and religious factors also come into play. Whether or not some people don't know is neither here nor there. Unusual is such an ambiguous word, leaving the question open to personal interpretation. No matter how much I may laugh at your "traps" question at least that is specific.

The question "Do you have unusual sexual preferences?" was selected because I didn't want to be too specific about which sexual deviations I think are related to neurodiversity. All I want to say about this is that it is not HBT (Homo, Bi, Trans). I actually don't want ASDs to be thought to be related to specific sexual deviations right now as these are even more stigmatized than DSM diagnosis. Additionally, the link is not that strong, but it was strong enough to include the more general "unusual sexual preferences"
.
And I can assure you that the sexual deviation(s) that are linked to neurodiversity are not cultural. Some strong candidates are not legal and have no known cure, so cannot be cultural.