The political compass questions were very ADBOC and generally meaningless. Apparently I’m left libertarian, whatever that means.
The Political Compass isn’t that great for handling nuanced opinions. It mostly seems to test agreement with partisan slogans, some of them quite far out of date, and the wording subtly favors the libertarian interpretation in a lot of cases; as such it’s a decent metric of ideological affiliation, especially if you’re somewhere on the libertarian spectrum, but it gets a lot noisier as soon as you start forming your own opinions. The tendency to use absolute phrasing (“must”, “every”, etc.) is especially problematic for a consequentialist.
The autism test was utterly without interpretation. What does 18 mean?
If this is normed like the last autism-spectrum test I saw, the cutoff point for autistic tendencies (bearing in mind the usual caveats about Internet self-testing) is somewhere in the low to mid-20s. Lower means you’re more likely to be considered neurotypical along that spectrum, higher means an autism (or, historically, AS) diagnosis is more likely.
How to read the autism score was explained on the test page itself.
quote: “Psychologist Simon Baron-Cohen and his colleagues at Cambridge’s Autism Research Centre have created the Autism-Spectrum Quotient, or AQ, as a measure of the extent of autistic traits in adults. In the first major trial using the test, the average score in the control group was 16.4. Eighty percent of those diagnosed with autism or a related disorder scored 32 or higher. The test is not a means for making a diagnosis, however, and many who score above 32 and even meet the diagnostic criteria for mild autism or Asperger’s report no difficulty functioning in their everyday lives.”
So a score above 32 means you are highly likely of being autistic.
The Big 5 test was suspect on some things. Am I really lowest quartile conscientiousness and agreeableness? I defy the data.
I noticed that my conscientiousness and agreeableness score were significantly lower than the last time I took the big five test (slightly before I really got into LW). I don’t have solid evidence, and it does seem a bit convenient, but what I think is happening (please take a second to recall your subjective experience of taking the test before reading further). Despite what the test says, most of the questions do have “right” (or at least socially desirable) answers and LWers are better than the population used to norm the test at avoiding self-serving biases in their answers. Do you remember conscious thinking of biases while taking the test and/or have you developed habits meant to make you less prone to having self severing biases to begin with?
Do you remember conscious thinking of biases while taking the test and/or have you developed habits meant to make you less prone to having self severing biases to begin with?
I wasn’t consciously keeping biases in mind and I don’t think I’ve consciously developed habits to ameliorate them, although I’ve probably gotten better at recognizing and avoiding them anyway. I scored higher on Agreeableness and Conscientiousness than the last time I took a Big 5 test, by almost a quartile each. If I had to guess, I’d probably ascribe this to improving my social life over the last year or so, although my participation in LW would be hard to disentangle from that.
The autism test was utterly without interpretation. What does 18 mean?
It says at the top:
In the first major trial using the test, the average score in the control group was 16.4. Eighty percent of those diagnosed with autism or a related disorder scored 32 or higher. The test is not a means for making a diagnosis, however, and many who score above 32 and even meet the diagnostic criteria for mild autism or Asperger’s report no difficulty functioning in their everyday lives.
but that isn’t enough information to calculate P(autism|score=N). There’s more information about the test here.
The Big 5 test was suspect on some things. Am I really lowest quartile conscientiousness and agreeableness? I defy the data.
I think the test is noisy on agreeableness. Six months ago I got 76%, and a couple of days ago I got 17%. The same person answering the same questions with the same friend being used as a comparison.
The other four (Open 95% Conscientious 41% Extravert 91% and Neurotic 4%) were as stable as you’d expect, and reflect my self image.
I have no idea whether I’m actually agreeable or not. I certainly can be, and I can also be spectacularly nasty. And I guess what you think of me would probably depend on which face I’ve mostly chosen to show around you. (it usually is a conscious choice).
I wonder if the test score is being influenced by which memories are most salient when I take the test.
Did it.
The political question was dumb. why can’t I pick “FAI” or “rational consequentialist”.
I really liked the CFAR questions. MORE OF THOSE.
The political compass questions were very ADBOC and generally meaningless. Apparently I’m left libertarian, whatever that means.
The Big 5 test was suspect on some things. Am I really lowest quartile conscientiousness and agreeableness? I defy the data.
The iq test was fun and challenging. Got 133, which is also what I’ve gotten on previous iq tests.
The autism test was utterly without interpretation. What does 18 mean?
EDIT: sorry I’m being so negative, a good survey overall. Maybe this is where the low agreeableness comes from.
The Political Compass isn’t that great for handling nuanced opinions. It mostly seems to test agreement with partisan slogans, some of them quite far out of date, and the wording subtly favors the libertarian interpretation in a lot of cases; as such it’s a decent metric of ideological affiliation, especially if you’re somewhere on the libertarian spectrum, but it gets a lot noisier as soon as you start forming your own opinions. The tendency to use absolute phrasing (“must”, “every”, etc.) is especially problematic for a consequentialist.
If this is normed like the last autism-spectrum test I saw, the cutoff point for autistic tendencies (bearing in mind the usual caveats about Internet self-testing) is somewhere in the low to mid-20s. Lower means you’re more likely to be considered neurotypical along that spectrum, higher means an autism (or, historically, AS) diagnosis is more likely.
How to read the autism score was explained on the test page itself.
quote: “Psychologist Simon Baron-Cohen and his colleagues at Cambridge’s Autism Research Centre have created the Autism-Spectrum Quotient, or AQ, as a measure of the extent of autistic traits in adults. In the first major trial using the test, the average score in the control group was 16.4. Eighty percent of those diagnosed with autism or a related disorder scored 32 or higher. The test is not a means for making a diagnosis, however, and many who score above 32 and even meet the diagnostic criteria for mild autism or Asperger’s report no difficulty functioning in their everyday lives.”
So a score above 32 means you are highly likely of being autistic.
I noticed that my conscientiousness and agreeableness score were significantly lower than the last time I took the big five test (slightly before I really got into LW). I don’t have solid evidence, and it does seem a bit convenient, but what I think is happening (please take a second to recall your subjective experience of taking the test before reading further). Despite what the test says, most of the questions do have “right” (or at least socially desirable) answers and LWers are better than the population used to norm the test at avoiding self-serving biases in their answers. Do you remember conscious thinking of biases while taking the test and/or have you developed habits meant to make you less prone to having self severing biases to begin with?
I noticed self-serving a bit on the first few questions and adjusted. I definitely remember thinking of biases.
Most tests (I think including this one) are normed on people who (mostly) don’t do that. Thus I’d expect it to under rate your positive qualities.
Gotta keep that in mind next time...
(Relevant comment)
I wasn’t consciously keeping biases in mind and I don’t think I’ve consciously developed habits to ameliorate them, although I’ve probably gotten better at recognizing and avoiding them anyway. I scored higher on Agreeableness and Conscientiousness than the last time I took a Big 5 test, by almost a quartile each. If I had to guess, I’d probably ascribe this to improving my social life over the last year or so, although my participation in LW would be hard to disentangle from that.
http://lesswrong.com/lw/7s4/poll_results_lw_probably_doesnt_cause_akrasia/ is relevant.
We’ll find out what it means when we have an average of all the other LW results.
It says at the top:
but that isn’t enough information to calculate P(autism|score=N). There’s more information about the test here.
They address some of those issues in their FAQs.
I think the test is noisy on agreeableness. Six months ago I got 76%, and a couple of days ago I got 17%. The same person answering the same questions with the same friend being used as a comparison.
The other four (Open 95% Conscientious 41% Extravert 91% and Neurotic 4%) were as stable as you’d expect, and reflect my self image.
I have no idea whether I’m actually agreeable or not. I certainly can be, and I can also be spectacularly nasty. And I guess what you think of me would probably depend on which face I’ve mostly chosen to show around you. (it usually is a conscious choice).
I wonder if the test score is being influenced by which memories are most salient when I take the test.