Hacker News Reader

Comment by lars | Hacker News Reader

There's a huge difference between a) not finding an effect, and b) finding that there is no effect. The former is the default result, and is what you get when your test lacks statistical power (from having a too low sample size). It sucks that the former often gets reported as the latter in the press.

View on HN

I look at it as a series of mis-aligned incentives. Only so much funding is available, and that defines the size and scope of the experiments that can be done. This, in turn, limits the study design (and by implication the downstream breadth of the conclusions that can be drawn). Researchers, by and large, do the research they can afford to do. Once that research is done, career and future funding potential comes from how much publicity and notoriety a study can get. Credit in many research communities goes to the authors of the papers that get the most citations, and you get citations by producing popular papers. Novel conclusions breed popularity.

In the hands of a pop science journalist/communicator, traffic/audience engagement are often used as criteria for success/career advancement/job security. Thus, there isn't proper, unbiased funding for the serious evaluation and consolidation of the conclusions that can be drawn from a new research paper. This is why we get the Buzzfeed-worthy "They compared honey to corn syrup, and you'll never guess which one was healthier" headlines instead of the boring-but-accurate "No variation reported in limited time-series study of sugar intake" that might better represent the data.

On the other hand, honey is mostly glucose + fructose in about the same ratio as table sugar and HFCS.

Seriously. If two things are chemically the same despite being from different provenance, chemistry teaches to expect identical behavior for the most part. Sometimes trace elements change things significantly. Why anyone would expect that of honey though, wishful thinking?

But it's all natural...

It's bad enough when they're talking about one sugar being healthier than another, but it really gets strange when they start comparing table salt.

Reply on news.ycombinator.com