Researchers Develop 2-Minute Autism Screening Tool

Post Reply New Topic

Disability Scoop

"For the study, parents of 1,959 children ages 18 to 36 months with no known developmental issues were screened using the method during pediatrics visits at clinics in New Jersey."

1. There is no other sampling information given. We don't know if this was all boys, or mainly boys, or girls.

2. They made conclusions based on a very poor sampling technique apparently without any awareness of sampling error. Sampling error results in unreliable and invalid conclusions. Robust sampling technique is needed to produce credible results.

3. There was no control group of diagnosed AS people to compare results with.

4. What instructions did they give to the participants? How did testing affect the children who weren't expecting it?

5. A priori hypothesis?

6. Confounding variables ignored and unaccounted for.

7. Conclusions not justified by the data collection method.

8. No double blind methodology used.

9. No validity and reliability info provided on the screening tool they used or if or how it had been piloted on AS children previously.

10. Limitations of a study need to be made clear, and recognised in reputable research

11. Post hoc analysis?

12. Poor randomisation. These children were all visitors to the clinic. That is NOT a random sample.

13. They seem to think that size of a sample compensates for poor methodological design. It doesn't.

If you don't follow the rules, then the findings are 1) useless 2) misleading 3) invalid.

This seems to be the "test" they used:
http://www.noahsarkinstitute.org/files/ ... pdq-1r.pdf

As a basis for future investigations, perhaps, but the inflated claims that it identifies AS is not credible, and I couldn't find any reliability or validity scores for the test. If there are none, they shouldn't be using it in this way. It's unethical to do so.

I have yet to read the full (9 page) paper, but it does seem that the method and sample leave the study open to all kinds of criticism, as B19 listed.

This is simply a questionnaire for parents to see if their children could benefit from an evaluation.

Shouldn't be much criticism.

Over the past 5 years, while I have been closely observing it, I have been very surprised to learn that lower standards generally apply in AS research. I ponder why this is, and why often very poor papers get publicity at all.

I have different theories about this. Generally, the standard of what is medical research is fairly good. Possibly AS is seen in the glow of that halo, although AS is not a medical illness. Another theory is that people who go into AS research may do so - possibly - because they think it is an easy option. They know that low standards are not questioned, and not a high level of research skills are needed to make progress. Or something else.

Could it be that they believe the AS population and their families don't question the sticky issues, that they assume them to be a naive audience? I hope not, but given the stigmas, maybe that is part of it. Most research these days, especially on children, needs ethical approval before it can be done. I wonder if this convention is applied to AS studies in the same rigorous way.

I have a lot of thoughts and questions about the whole scene, not many answers though.

Good research is complex, at each step and stage, and validity issues are ultra important. In AS research there are glaring omissions - little discussion (if any) about confounding variables, and sometimes you get several researchers over a few years doing similiar studies and each claiming it is new and "groundbreaking". Only some of this groundbreaking stuff was done years before by someone else, and no credit given. This is odd, because it's convention to do a literature search before embarking on "new" research.

Some studies are ok, though they seem the clear minority of the whole. It's a strange phenomenon.