How To Filter Good Science From Bad Science In Minutes

• comments
• 
• 

Every day, you find hundreds of new science papers and research papers making your way to you through online and offline media. Many of them seem to have blockbuster results, except that do not really qualify as good science. Even studies published by well-ranked Science journals might turn out to be fraudulent later.

Officially, only 700 science paper retractions happen every year. It is just the tip of the iceberg of the science frauds we get exposed to a daily basis. Thankfully, there are some quick ways to spot credible science papers from untrustworthy ones that count as bad science - in a matter of minutes.

Michael J.I. Brown, Associate Professor of the Monash University, says that good science is meticulous. It takes months, and even years, to identify and address a problem. He writes in The Conversation, "If you're taking the time to do meticulous science, why not take the time to prepare a good manuscript?... It seems obvious enough, which is why a sloppy manuscript or poor grammar can be a warning sign of bad science."

A paper draft with poor formatting, figures with captions on the next page, and incorrect grammar are some of the signs of bad science that you can spot in the first few minutes.

An article published by the BigThink contains a number of red flags that can quickly help you filter through science news and research papers and figure out which ones are not worthy of your time:

1.       Sensational headlines that misrepresent findings or over-simplify them

2.       Misinterpretation or distortion of research findings to increase their interest factor

3.       Research results that favor or malign certain products or brands

4.       False assumptions, for example, if two factors are correlated that does not mean that one causes the other

5.       Speculation or use of words like 'may', 'could', 'might' etc.

6.       Too small sample size

7.       Sample that does not represent larger population (especially in human trials)

8.       No control group

9.       No blind testing (Here, subjects should not be told that they are in test group or control group.)

10.   Using 'selected' results from other experts to support research conclusion, ignoring complete results

11.   Results that are not replicable

Even papers published by major journals or having a number of citations can be wrong or flawed. So keep your mind open.