One of my sayings is “Where there is a debate, there is research” – but there are problems in the research.
To discuss this point I must first explain how science works. In universities, where much of all published research is done, research is broken up into small teams. Each team is led by a primary investigator. Their main job is to plan the research and get funding for it. They write up an application for research money, often to a public source, but sometimes a private source. For this, the application must sound fantastic: never heard of, could cure something, could find out something amazing etc. etc. Once they get the funding, they will hire staff on short term contracts and rope in students such as Ph.D. students to do the research. The primary investigator must then report back to the research funder on how things are going. Once they have results, they will write up the research in a journal article and submit it to one of the many journals out there – you may have heard of Science or Nature.
After you submit the article, the editor of the journal will send your article to other scientists in the field to ensure the research is good; the article will then be returned to authors with the corrections and, whether it was perfect research or not, there will be corrections. Once corrected, the process is repeated and then it is published with the authors’ names on it (the person who did the most work coming first and the primary investigator last). Now this is important. For the careers of all people in science, there is a saying: “PUBLISH OR PERISH.” This means get as many journal articles as possible with you as the first author or your career will perish.
The first problem with this is availability. I regularly cringe at what the media and the public say with the illusion of having knowledge. They’ll spin some message with jargon, statistics and – worst of all – personal anecdotes, to sound scientific; sound like they know something; sound like they read the research. I think these morons are spreading ignorant propaganda. But, on second thought, it’s not their fault. To publish something without doing any research takes three minutes. To do the research and then get it published in a journal can take years.
Then, after all that effort to get published in a journal, who reads it? Not the public – because most journal articles cost money to read. It would cost the average internet user $60USD to read one paper, and to do a basic review of the current literature on a topic would require reading upwards of 100 papers. So what do they read instead? “Yahoo! Answers” or something like that. This means that information today is owned by the journals and ignorance is free. Soooo, what gets more hits? What is read more? What is BELIEVED more? IGNORANCE!
Another problem is positive publishing bias. This is basically the fact that things that work get published in journals and things that fail get rejected. Here is an example: imagine that I come up with the idea that humming loudly cures cancer, and I do a very thorough experiment to test this idea. If I was wrong, there is no way in the world any scientific journal would publish this research – and they would think I was a moron for thinking this. However, if I was right, I would be published in possibly a very good journal. This is an extreme example, but it does occur all the time. Now you wouldn’t think that would be a problem if the research is good and the result is positive. That means you’ve cured something and the world should know – if it doesn’t work, who cares? Why does the world need to know that something doesn’t work? This comes down to statistics.
If you have two groups that vary, but their averages are different, how do you know whether this difference is real or just because of the variation in the population compared to the sample size? For example, if two people had two coins and then flipped them at the same time, and one landed on heads and the other tails, does this mean the coins are different? No – because this could have occurred by chance. If they flipped them five times each and one always landed on heads and the other always landed on tails, this also could have occurred by chance. However, you can calculate how confident you are that this didn’t occur by chance (for one coin it would be 0.5 X 0.5 X 0.5 X 0.5 X 0.5 = 0.0031, or a 3 in 1000 chance). Once you’ve flipped them 100 times each and one is still only heads and one is still only tails, it seems clear that the coins are different. In science a 1 in 20 (0.05) chance that something occurred by chance is seen as a “significant” difference – as real and not by chance.
Now back to the positive publishing bias. What if twenty groups in the world are working on seeing if the same drug can cure cancer (that is entirely realistic), and 19 of them find that a drug has no positive effect on a disease? These may get rejected from journals and we would never know about their research. But if 1 of those groups, purely by chance, found that the drug seemed to have a significant effect on cancer, they would be more likely to get published. Then, when the average researcher went and read up on the drug, they would only find the one article on it – saying that the drug worked – and they would never know about the 19 other groups that found it did not. This drug that actually had no effect on the disease would now, according to the literature, be seen as a future remedy for cancer.
Now, when you add the PUBLISH OR PERISH pressure to the situation, this positive publication bias situation gets worse. They must get published and to get published it has got to be positive. Now scientists aren’t coming from an unbiased thirst-for-knowledge situation; they are coming from an “I hope this works” point of view. This isn’t good for research, and it isn’t good for knowledge.
OK, again, not the complete story – just some things to think about. Comment below; I would love to discuss further.