Hacker News new | ask | show | jobs
by VLM 4712 days ago
Bad story design... trying to talk about the typical result, then late in the article drop the bombshell "The team considers Jaimee and her mother, Karen, among their best success stories.". If they "have to" go anecdotal, they could have at least picked the median subject. Like doing a report on the health effects of smoking, and intentionally selecting the healthiest 99 year old smoker in the world rather than the most likely outcome.

The most interesting part of the whole situation can be summed up by one of the lines describing the babies, "nearly all were African Americans." Even since the first days, It never was about medical issues or science, just a sorta-stealthy way to bash black folks. You'll note there was carefully no outage at the time, or long term medical study, at white coke snorting suburban women, although the blood levels of coke the babies experienced probably were about the same in the end. By analogy it would be like creating a social meme and scientific study of the negative pregnancy impact of malt liquor consumption (by urban black women), carefully ignoring the consumption of fruity margaritas (by suburban white women). Because you can't bash black people unless you can "other" them first.

I guess the two startup lessons are if you're trying to make median situation analysis, policy, and decisions, and you must use an anecdote, don't chose an extreme outlier, use a median... unless you've got an axe to grind and you're trying to mislead people, in which case unusual sample selection can be a powerful tool to mislead people. Startup lesson two is one popular way to scam people is to play word definition games as a strategy for divide and conqueror, so look out for that gameplay technique, and/or use word redefinition as a weapon of your own.

3 comments

I found the success story important to heavier weigh the fact that cocaine exposure doesn't stunt brain development to a greater degree than general poverty. The author made plenty of other points about the tested groups not behaving other than expected given poverty as a common factor among them.
Bad story design...

I disagree. If you're trying to tell a compelling story, you want to reinforce the story that you're trying to tell in every way that you can. That makes it more compelling, and sells more page views.

The danger being, of course, that making reasoned decisions based on compelling story lines is dangerous. But this article isn't in that business - it wants to sell page views.

Journalists have a responsibility to the public that goes beyond page views. That is what makes them journalists.
If you wish to be consistent and say that the authors of the articles in virtually every blog of note are not journalists, then I will agree.

Unfortunately lots of people who have nothing resembling "journalistic integrity" make a living that way. If you're bored, you can turn on Fox News and see how quickly you can identify one.

"By analogy it would be like creating a social meme and scientific study of the negative pregnancy impact of malt liquor consumption (by urban black women), carefully ignoring the consumption of fruity margaritas (by suburban white women)"

If suburban white women consuming fruity margaritas were much more responsible with their drinking, so that they never consumed while pregnant, then we could study the effects of alcohol on pregnancy solely by focusing on the effects of malt liquor on pregnancy.

Crack use was widespread and used by poor pregnant women. Cocaine use has never been as widespread in suburban white populations. So researchers aren't being racist, they're investigating the problem where it actually exists.