5 ms·
So while I haven't read the PDF (I don't have an account/institution), I wonder how robust this study is given that it's based on a questionnaire on how often p
by azinman2 12y ago
So while I haven't read the PDF (I don't have an account/institution), I wonder how robust this study is given that it's based on a questionnaire on how often people talk to their babies. This kind of self-reporting is often troubled as the assumptions as to what 'always' means is very relative on a per person basis. It wouldn't surprise me that people who read frequently to their baby might under report how much they speak to their baby when given a line of questioning about it, and vice versa.
I also question how much you can really test a baby for cognitive development that is robust against time to say anything about adulthood. The Ages and Stages questionnaire used is about assessing delayed development versus adulthood, and the sensitivity is as low as 51% for the 4-month ASQ and 75% overall, and the specificity is 86% [1]. These numbers themselves aren't super great which only will compound with the cohort study in question.
It's probably better to do both, especially as when the child ages reading to them helps them gain practice reading themselves. I fear that many people will look at a headline like this and feel like they can avoid reading to their kids by turning on the radio.
[1] http://www.psychwiki.com/dms/other/labgroup/Measufsdfsdbger345resWeek1/Lindsay/Squires1998.pdf http://www.psychwiki.com/dms/other/labgroup/Measufsdfsdbger3...
- Gimpei 12y agoAgreed. Reporting behavior could be correlated with unobservables that are driving the results. Interesting study but far from conclusive. IMHO correlations like this are useful for driving future, more rigorous research, not for drawing broad conclusions.