 Have you ever seen a TV ad claim 4 out of 5 dentists use this brand of toothpaste? Why do marketers like to cite experts? The reason is because when it comes to complicated issues, people rely on expert views. But here's the thing, everyone, including scientific experts, can have their views. But when it comes to science, they need to back up their views with evidence and rigorous analysis. And then, as a reward for their efforts, their work gets scrutinized by other experts. This is the peer-review process. Peer-review is when scientists who are experts in the field scrutinize a paper before it gets published. It's like a spam filter on steroids. It's not perfect, but what it strives to do is weed out the errors, ensuring that the science is rigorous and evidence-based. This idea of scientific research being based on common standards of evidence is referred to as social calibration. Once a paper has gone through the peer-review process and published in a scientific journal, then it faces the hardest test of all, the test of time. You get an idea of the quality of scientific research from how frequently other scientists cite that research in their own papers. So what peer-review papers, which have been scrutinized over time, have to say about climate change. In 2004, Naomi Arezquez examined nearly 1,000 peer-reviewed papers from 1993 to 2003, matching the search global climate change. She was particularly interested in the number of papers which rejected the scientific consensus that humans are causing global warming. What she found was nothing. Not a single paper rejected the consensus. This wasn't to say that there were no papers in existence rejecting human-caused global warming, but her analysis indicated that rejection papers had a negligible presence in climate research. Nearly a decade later, I led a team of researchers at Skeptical Science in conducting a follow-up analysis to the Uresca study. We thought it was time to have another look at the level of agreement in scientific papers. In order to capture the range of topics in climate change research, we expanded the search from global climate change papers to include global warming papers. In 21 years from 1991 to 2011, this added up to more than 12,000 papers. We had at least two members of our team read each abstract, the summary at the start of each paper. We categorized each paper on whether it affirmed human-caused global warming or not. In the end, we identified around 4,000 papers stating a position on human-caused global warming one way or the other. Among these 4,000 papers, 97.1% endorsed the consensus that humans are causing global warming. What about the other 8,000 papers? Well, abstracts are short summaries that introduce a scientific paper. If they don't need to, they don't waste time stating something they assume their readers will already know. Most astronomy papers don't think it's necessary to explain that the earth revolves around the sun. Similarly, nowadays most climatology papers don't see the need to reaffirm the consensus position. When a paper's abstract does mention human-caused global warming, there's overwhelming agreement. Replication is a fundamental part of the scientific method. It should be possible to repeat an experiment and get the same result. So how could we check our results? We invited the scientists who wrote the papers to rate their own research. This offered another way of appraising the work. 1,200 scientists responded to our invitation, and over 2,000 papers were self-rated. Among the papers stating a position on human-caused global warming, 97.2% endorsed the consensus, according to the scientists who did the research. As far as cross-checks go, this was a strong validation of the 97.1% consensus that we found from reading the abstracts. So both the early work of Naomi Oreskes and our more recent analysis found overwhelming consensus dating back to the early 1990s. An additional way to look at how scientific understanding has changed over time is to look at the way scientific studies cite other studies. If a paper makes a valuable scientific contribution, other scientists will refer to it in their own research. If there was an ongoing debate within the scientific community on climate change, then you would expect to see two schools of thought competing in the published research. But as a consensus forms, the presence of the dissenting viewpoint gets proportionally smaller. One study looked at this by analysing the way climate papers reference each other. This graph shows a measure of scientific disagreement over the past few decades. They found that scientific consensus had already formed in the early 1990s. Coming at it from a completely different angle, they found exactly the same result as Oreskes and our own analysis. The scientific research tells a clear story. In 1995, a landmark IPCC report found a discernible human influence on global climate. But even before that report came out, there was already overwhelming agreement in the scientific literature. 18 years later, the evidence has become so strong that the most recent IPCC report stated that it was more than 95% likely that humans have caused most of the observed warming since the mid-20th century. 4 out of 5 dentists endorsing a brand of toothpaste is a strong statement, but that's just 80%. 97% of relevant climate papers affirm the scientific consensus that humans are causing global warming. And that's not just a view of scientists. It's a position expressed in peer-reviewed scientific papers.