Problems with the AQ test.

Page 3 of 3 [ 46 posts ] Go to page Previous 1, 2, 3

Joined: 14 Jan 2011
Age: 38
Gender: Female
Posts: 942
Location: SoCal

06 May 2011, 8:07 pm

My problem with most of the online tests is that I'm far too familiar with the symptoms (partially because AS itself has turned into something of a special interest!) so I have trouble answering the questions as objectively as I might once have before I knew. I know what answers the test is looking for, so I would be able to "skew" the results if I wanted to, either towards the Aspie side or towards the NT side. Sometimes it's unintentional, even, as there are times when I lack the outside perspective to tell me if I truly have xyz problem or if I'm just seeing things where there are none to be seen. Or, for example, I am seeing normal variation that would be present in the general population. Ugh!

_________________
I don't suffer from insanity. I enjoy every minute of it.

swbluto
Veteran
Veteran

Joined: 26 Feb 2011
Age: 37
Gender: Male
Posts: 2,899
Location: In the Andes, counting the stars and wondering if one of them is home to another civilization

06 May 2011, 8:12 pm

syrella wrote:

That's true, but the test has high test-retest reliability (Around .9, I believe) and I personally scored a 32, 35 and 36 upon my re-assessments and those were the intentionally honest times I tested, so there's a fairly high likelihood there won't be a large variation in the scores for an "honest effort" (The large variation that could be due to "unintentionally swaying the results"). That's assuming your understanding of the questions doesn't significantly change between retests.

My particular problem is comprehension bias. When it says "Do you have troubles imagining what the character in a book looks like?"(Paraphrased), I'm thinking I can usually imagine *something*, but I always thought that I never quite imagined it the way the author was imagining it, and so I didn't think I imagined it *that well*. So there's those interpretative and perceptual biases to contend with.

Last edited by swbluto on 06 May 2011, 8:16 pm, edited 1 time in total.

draelynn
Veteran
Veteran

Joined: 24 Jan 2011
Age: 56
Gender: Female
Posts: 2,304
Location: SE Pennsylvania

06 May 2011, 8:16 pm

Verdandi wrote:

Yeah, the scores don't correlate to severity.

I'm so glad it's not a measure of the severity of my hubby's NT status...

Verdandi
Veteran
Veteran

Joined: 7 Dec 2010
Age: 55
Gender: Female
Posts: 12,275
Location: University of California Sunnydale (fictional location - Real location Olympia, WA)

06 May 2011, 8:19 pm

draelynn wrote:

Verdandi wrote:

Yeah, the scores don't correlate to severity.

I'm so glad it's not a measure of the severity of my hubby's NT status...

I know one guy who scored a 6.

puddingmouse
Veteran
Veteran

Joined: 24 Apr 2010
Age: 38
Gender: Female
Posts: 8,777
Location: Cottonopolis

06 May 2011, 8:27 pm

Verdandi wrote:

Yeah, the scores don't correlate to severity.

What are they supposed to correlate to? The stereotypes in the questions?

_________________
Zombies, zombies will tear us apart...again.

puddingmouse
Veteran
Veteran

Joined: 24 Apr 2010
Age: 38
Gender: Female
Posts: 8,777
Location: Cottonopolis

06 May 2011, 8:27 pm

double post

_________________
Zombies, zombies will tear us apart...again.

Last edited by puddingmouse on 06 May 2011, 8:28 pm, edited 1 time in total.

draelynn
Veteran
Veteran

Joined: 24 Jan 2011
Age: 56
Gender: Female
Posts: 2,304
Location: SE Pennsylvania

06 May 2011, 8:28 pm

Verdandi wrote:

draelynn wrote:

Verdandi wrote:

Yeah, the scores don't correlate to severity.

I'm so glad it's not a measure of the severity of my hubby's NT status...

I know one guy who scored a 6.

Woah... that guy is so totally the posterboy for normal...

Verdandi
Veteran
Veteran

Joined: 7 Dec 2010
Age: 55
Gender: Female
Posts: 12,275
Location: University of California Sunnydale (fictional location - Real location Olympia, WA)

06 May 2011, 8:30 pm

draelynn wrote:

Verdandi wrote:

draelynn wrote:

Verdandi wrote:

Yeah, the scores don't correlate to severity.

I'm so glad it's not a measure of the severity of my hubby's NT status...

I know one guy who scored a 6.

Woah... that guy is so totally the posterboy for normal...

Actually, he has bipolar and ADHD. I don't think he comes off as "normal" very often, but he's very much not autistic.

puddingmouse wrote:

What are they supposed to correlate to? The stereotypes in the questions?

I think they're supposed to correlate to chances of having an ASD.

draelynn
Veteran
Veteran

Joined: 24 Jan 2011
Age: 56
Gender: Female
Posts: 2,304
Location: SE Pennsylvania

06 May 2011, 8:38 pm

Verdandi wrote:

Actually, he has bipolar and ADHD. I don't think he comes off as "normal" very often, but he's very much not autistic.

You know, I find that incredibly interesting... my hubby is bipolar too. Is the difference there really brain wiring vs. brain chemistry? Beyond the scope of this thread but made me go... hmmmmmm...

rdos
Veteran
Veteran

Joined: 6 Jul 2005
Age: 63
Gender: Male
Posts: 6,096
Location: Sweden

07 May 2011, 2:28 am

swbluto wrote:

Even though this quiz doesn't really have a high predictability for possessing aspergers for "moderate scores"(in the 30-40 range), the multiple-regression analysis I performed suggested that it was nearly twice as predictive as the Aspie quiz. However, someone might want to look over the formula I derived in computing the relative predictability just to be sure...

What is your material? I have compared with the AQ-test (same participants), twice in Aspie-quiz, with some 1000 participants. The figures Simon-Baron Cohens have calculated for the reliability is nowhere near his in this population.

rdos
Veteran
Veteran

Joined: 6 Jul 2005
Age: 63
Gender: Male
Posts: 6,096
Location: Sweden

07 May 2011, 2:30 am

swbluto wrote:

Aspie Quiz has 3-4% of coefficient of variation in the results. Don't know how this translates to correlation.

AllieKat
Toucan
Toucan

Joined: 22 Apr 2011
Age: 48
Gender: Female
Posts: 294

07 May 2011, 2:39 am

I still predict that the more social impaired Aspies who love to talk will probably score LOWER than mild Aspies who tend to me more aware of their social difficulties. I have a NT friend who's just a really introverted person and he scores higher on the AQ test than I do. Yet, he's never had problems reading other people or saying things that unintentionally offend them. It's just that he prefers to spend time alone even though his social skills, IMO, are actually quite good.

Verdandi
Veteran
Veteran

Joined: 7 Dec 2010
Age: 55
Gender: Female
Posts: 12,275
Location: University of California Sunnydale (fictional location - Real location Olympia, WA)

07 May 2011, 2:42 am

AllieKat wrote:

The more aware of my social impairments I've been, the higher my score got.

Although I am not sure how impaired I actually am, but I think that my experience seems to align with your theory, although it's subjective.

rdos
Veteran
Veteran

Joined: 6 Jul 2005
Age: 63
Gender: Male
Posts: 6,096
Location: Sweden

06 Jul 2014, 10:18 am

TPE2 wrote:

My main problem with the AQ is other - several statistical studies made about the AQ have show that the AQ questions don't really measure any relevant trait:

- The "Cronbach alpha" is usually between 0.6 and 0.7 (if a test is an alpha lower than 0.7, this usually means that the questions don't really have a relation with each other)

- Factor analysis usually show factors different of the sub-scales of AQ, and many questions of AQ are not really associated with any factor

That applied to Aspie Quiz as well in the beginning. However, final version 3 has groups / factors that have been validated with factor analysis and has both CFI and TLI over 0.95.

Page 3 of 3 [ 46 posts ] Go to page Previous 1, 2, 3

Jump to:

Problems with the AQ test. Post Reply New Topic

Problems with the AQ test.