No definitely never,

Objective criteria is one of the" complimentary", integration for connecting the observer to the subject. It is important for the observer to realise his/her relevance with the case. It is assessment for the observer's "SELF".

Multidimensional tools/ combined methods & Multidiciplinary, in addition to full engagement of "first person Psychology.

Interesting paper. Where is it published?

It is in the bioRxvi and submitted to the Journal of Neural Computation undergoing peer review. Thanks!

OK. Let us know when it is accepted and date of publication.

sure thing!

https://www.biorxiv.org/content/10.1101/717827v1

here is the site; please feel free to leave or send me comments to improve it

I agree with a lot of your comments re the use of the ADOS but was under the impression from Cathy that it was a diagnostic test (i.e., criterion referenced) meant to serve as a formal way of assessing cases and as an adjunct to inform clinical impressions and it does a pretty good job of that. This hasn't stopped folks from using it inappropriately however and the same can be said for the ADI with some using the total score as a measure of severity. I created the PDDBI (and it's subsequent variants) to get at some of the same issues you bring up (standardizing on age, for example). Not sure about the male-female issue though because the instrument wasn't designed to assess these more subtle issues. Adding quantitative biometric measures may be helpful at the research level as we have published using Noldus tracking data so I applaud you on that. Keep up the good work from my alma mater!

Thanks! Yes, it appears to be a test that is not designed to track change. Yet, I have seen a paper by Cathy and colleagues already doing so as a measure of neurodevelopmental trajectories. For clinical needs, it may be adequate to face the crisis we have in our hands, but the pickle is to have to use it in research, without having a proper metric to be able to measure change. It is too costly and too arduous for the children to have to complete. At the clinic, OK, you have to, but for research, not so sure anymore. We have digitized it and with 2-3 tasks top one gets what one needs below 15min. But we need to massively test it in neurotypicals to build the normative data bank. Until then, we are not going to have a proper metric scale to quantify trajectories in an age-dependent manner. I wished we could do a large study with her team and design the proper distance metric for this test in digitized form. Then I can get the self-emerging stratification that we need for research studies -no heuristics. It is totally doable

OK, that's an issue. There are also problems with the severity metric as you note, which also includes the lack of confidence intervals (unless she addressed that in her paper) and the issue of comparability of severity across modules. BTW, it is also important to have controls for ID, not just typicals. Good luck with that.

absolutely, but for a metric space we need the normative data first and it has never been collected; when we did it in a hand full of people, we saw the need to turn this into a norm-referenced test; and of course we need the ID controls. At ~30% of the reported prevalence, ID makes this a very worrisome and real issue, beyond possible false positives or probable inflated rates to give out much needed service. ID is the real thing that I see; with numbers way to high. It makes one wonder why the increase in ID?

The numbers have switched. It used to be 70% id.

Interesting...in NJ the rate of ASD is exponentially growing since 2002; taking off at about 2007-2008 (conservative estimate from open access special ed records publicly available up to 2016); with ~220K receiving special ed each year since then and the ASD seemingly absorbing other generic diagnoses; but it is capped at about 220K individuals from 3-21 years of age each year. Maybe that is the special ed budget? Regardless, it is a large number of adults by now if you think that children who were 4 and above in 2002 are now transitioning to adulthood (no id in there, just various disorders). Not sure what NJ will do; but I was curious to know how many adults are receiving services/benefits and no one could tell me. So it occurred to me to go into the public schools records. This is not the number but it gives me at least a clue.

It's a bunch of things - primarily changes in DSM criteria - broadening them considerably partially explaining the IQ effects as well as access to early intervention ABA services which also increase IQ; greater demand for ABA at all ages; greater awareness of ASD; parental involvement leading to increased funding of research and more specialized behavioral intervention programs (especially in NJ); the de-institutionalization movement in the 70s; increased survival rates of NICU infants and greater rates of multiple births; and pre-natal environmental exposure to toxins, etc. When I started, ASD was considered a psychosis and quite rare - no longer. (see my "What's New...") paper.

yes, very complex indeed; thank God for data science today, so we can retrospectively learn what didn't work and learn from it -thanks for your replies! they are very informative

Great. Best of luck going forward

thanks! I will need it! this is a difficult field -I have computational neuroscience as back up :-)

That's super.

and Parkinson's Disease :-)

https://www.jove.com/video/59827/dynamic-digital-biomarkers-motor-cognitive-function-parkinson-s

Does Jamie Mcpartland know of your work re objective measures for asd? Or Roz Picard?

Roz Picard knows a little bit about it, because we coincided at the Boston Club meetings that the Nancy Lurie Marks Family Foundation organizes ( a few years back). I received their Career Development Award and met her there. I do not know Jamie Mcpartland, but if there is interest in these, I wrote a book with a companion site that has the code, sample data to use it and so forth. https://www.elsevier.com/books/objective-biometric-methods-for-the-diagnosis-and-treatment-of-nervous-system-disorders/torres/978-0-12-804082-9

And the lab can provide services to analyze data in general, including transcriptome data now (working on that at Salk now on my sabbatical at the computational neurobiology Sejnowski lab). It is a very much needed area in ASD (I think) and I am hoping to help as much as needed/wanted. It is up to people really; at our end, we are making the methods openly available to the community.

Ok. Sounds great.