How to get the result you want from a poll

Keep an eye out for news articles reporting overwhelming support worldwide for human embryonic stem cell research.

An online questionnaire was prepared asking 60 questions about IVF ethics; it was anonymous and in a multiple-choice format. For this study, two of the sixty questions were examined: 1) “I believe it is morally WRONG to use embryonic stem cells for research”; and 2) “I believe it is morally WRONG to use embryonic stem cells for medical treatment”. The respondent answered the questions based on their current knowledge. The questionnaire was available online at as well as the patient support sites and .

Let me get this straight:

* The questions don’t specify human embryonic stem cells (though I’ll grant that was probably assumed by most respondents);
* The questions don’t mention that embryos are typically destroyed to obtain stem cells;
* The survey was conducted on three IVF support/advocacy sites. IVF doesn’t have to be done in a way that deliberately destroys embryos, but in many countries it is, so this selects for a group of people who are not opposed to the destruction of human embryos;
* It’s an online survey, and there’s no indication that data were weighted to account for the difference between this self-selected sample and a random sample.

That pounding noise you hear is George Gallup headdesking from beyond the grave.

Overall, we can conclude that people who responded to the survey expressed an overwhelming support for the use of embryonic stem cells for both medical and research purposes, independent of their background.

Sure. What they can’t conclude is that this has any relevance to anyone except the population who took their survey.

This finding is not reflected accurately in public communications or in the socioeconomic environment.

I’m not quite sure what it would mean for a survey finding to be reflected accurately “in the socioeconomic environment” — I think they’re bothered because “It is widely perceived that the public is equally divided in their views toward this controversial subject.” Mostly, it sounds to me like the authors would like to generate media reports claiming that “78% of people worldwide support human embryonic stem cell research!”

And look, they’ve already gotten some.