How do you conduct a credible study? Well, if you’re looking into the effects of a company’s practices, it might look a little more kosher if that company doesn’t fund the research.

This story in today’s Herald reports on a Dow-funded study into the health effects of a Dow weed-killer factory in Taranaki.

The study concluded that “there is no evidence of increased cancer or disease” related to dioxin exposure at the factory.

But buried at the bottom of the story (page two on the web version), is the study leader’s acknowledgement that the research “incorrectly suggested that health effects had been studied”.

The research could only really comment on death rates and levels of dioxin in the blood. And even that data was extrapolated.

Rather than taking samples from all 1700 people they deemed related to the case, they sampled just 346, and drew their conclusions from that.

Dow’s a big company and I don’t imagine it’s short of cash. Surely it would have been more conclusive to sample everyone concerned, and look at their medical history.

As it is, it looks to me like they did a little study and extracted big conclusions. What do you scientists out there think?

Other posts by Katie Llanos-Small