The uncertainty of science: In a new study, researchers attempted to replicate and thus confirm the results of several hundred social science papers, and found that the results of only half could be replicated.

Researchers from a variety of universities looked at “164 quantitative papers published from 2009 to 2018 in 54 journals in the social and behavioural sciences,” according to the summary in the Nature article. The team “attempted replications of 274 claims of positive results” but found only about half could be replicated. The researchers found that many published findings did not consistently hold up when tested again, although the exact replication rate varied depending on how success was measured.

This result jives with other reports over the years that found most science research difficult if not impossible to replicate or confirm.

In fact, every study in the past two decades that attempted to replicate earlier work has consistently found that about half the papers published in the scientific literature in the soft sciences (psychology, social sciences, biology, medicine, pharmaceuticals) could not be confirmed.

In other words, a very very large percentage of science research is junk and should be ignored. In fact, no study or research result should be given any credence until someone else has been able to duplicate the results independently. We should especially pay no attention to declarations by any government health officials, as such declarations are routinely completely divorced from any research at all.