麻豆淫院

UNSW researchers have shown that poor quality research methods and geographical biases in reviews mean the evidence policymakers may rely on is shakier than it appears.

Starting in the 1960s, a wave of research examined the impact of organochloride pesticides such as DDT on wildlife, the environment, and humans.

Organochlorides are now banned in Australia and elsewhere because of their ability to bioaccumulate in both the environment and within organisms.

Following that first wave of studies, a secondary wave of reviews鈥攐ften in the form of meta-analyses鈥攃ombined results to give a clearer picture of the dangers of these pesticides.

However, new UNSW research published in shows poor quality methodologies in those reviews are common and these studies are frequently cited in policy documents around the world.

The researchers highlighted the quality of more than 100 meta-analyses cited in over 200 policy documents, finding that over 80% were poor quality, including poor literature searches, poor statistical analysis, and a limited use of reporting guidelines.

The researchers say "[Meta-analyses] can be misleading and riddled with subjective bias while projecting the illusion of objective authority, making policymakers believe they are referencing reliable evidence when in fact it may not be the case."

The team found these analyses had been used by organizations including the World Health Organization and the United Nations to build policy.

"This demonstrates that the influence of poor-quality meta-analyses extends to policy documents that may have implications on international policies," says study lead author Kyle Morrison, a Ph.D. candidate from the UNSW School of Biological, Earth & Environmental Sciences.

Australia is a member of both organizations and has representatives at meetings where these policy documents are produced.

The researchers also found no difference in the quality of meta-analyses cited in policy documents and those which were not.

The authors argue that the large influence of poor-quality meta-analysis methodologies reflects a potential larger issue in science.

"Studies tend to gain visibility and citations based on their findings rather than their methodological quality," Morrison says.

"In other words, striking results often receive more attention than well-conducted, transparent research."

The researchers call for a shift in academic incentives away from rewarding only those big results and toward valuing reliability, transparency, and openness.

The researchers also want to see better reporting and methodological standards for meta-analyses developed in environmental sciences, as well as more efforts to improve inclusivity and accessibility of science, especially regarding developing nations.

"Achieving this requires not just changes in individual journals, but a broader cultural shift across science and publishing."

The team argues that stronger guidelines and wider international collaboration are essential if pesticide regulation is to reflect the full weight of evidence.