Columbia University researchers are blaming child development problems on maternal exposure to the ubiquitous class of chemicals known as phthalates. What do we blame?
Table 6 of the study indicates that the explanation for the researchers’ conclusion is more likely to be the multiple comparisons fallacy. That is, if you test for enough correlations, eventually you will find a few.
As to the observed child development deficiencies, those are probably more likely linked to the fact that almost 85% of the study subjects came from families earning less than $30,000 annually.
Children need to be rescued from poverty, not chemicals.
Special interest groups are trying to get phthalates banned. They need the science to back up their claims.
@ pockets64: in Greek ‘phth’ is 2 consonants – ‘ΦΘ’ – pronounced ‘-fth’ as in ‘fifth’.
Once again a study is using “comparisons of ratios” to emphasize a ‘comparison’ between incidence rates of phenomena (observed versus putative ‘normal’) when a chi-squared statistic is called for to establish confidence limits. Once again data mining is used to identify ‘shapes in the clouds’ while examining data – pure pareidolia.
Any substance that starts with 4 consonants has to be bad.
Anyway, recycling sick people for statistical correlations again. Stats, not facts.
Does anyone believe anything that EHP publishes any more? If so, why?