While scientific studies can collect valuable information, the conclusions drawn from those studies are always open to alternative explanations. Scientific studies are limited by the methods available at a given time, and are based on assumptions that might or might not be correct. Even more tentative are the conclusions that we reach based on our imperfect scientific investigations. Even where there is consensus among most scientists, it is always possible that future investigations will offer new insights and new conclusions. In other words, when it comes to science, scientific inference is never settled.
Science Fact Versus Inference
Science is about using systematic ways to collect evidence and draw conclusions from that evidence. This can be seen in a typical scientific peer-reviewed report which is broken into four sections.
- Introduction: The first section explains the purpose of the investigation and provides context and background.
- Method: This section explains the way in which the evidence was collected. What was measured and how it was measured.
- Results: The data are analyzed. In my field, most studies involve collection of quantitative data that is analyzed with statistical tests.
- Discussion: The results are interpreted and conclusions are drawn.
The results are about the collection of information or data. Suppose I conduct a simple survey study where I ask employees in a company to rate their job satisfaction on a scale from 1 (hate it) to 10 (love it). If the mean winds up to be 8.5, that observation is a “scientific fact”. There is little interpretation other than the mean is 8.5. There are underlying assumptions that the survey platform accurately collected the responses, that there were no glitches when I download the dataset, and that I did the analysis correctly. In theory we could have 10 researchers download the data, and they would all get the same mean, giving us confidence is this is a verifiable “fact”.
Scientific inference is the interpretation we give. I can claim that most people in this company are highly satisfied with their jobs. That assumes that my respondents were able and willing to give an honest response. If the sample includes people with limited English proficiency, many might have misunderstood the question. If many in the sample are concerned that their supervisors would see their responses, they might have been dishonest and rated high when their true feeling is low. We can confidently say that the mean is 8.5, but our interpretation or inference is up for debate. I might make one claim, but someone else might have an alternative and perfectly reasonable explanation.
Scientific Inferences Can Change
Scientific inferences change over time, and it is not unusual for the meaning of “scientific facts” to be reinterpreted. At one time it was believed that Japanese were more depressed than Americans because they scored higher on depression scales. A Japanese researcher, Iwata Noboru, was skeptical. He noticed that depression scales have a mix of items where some were depression symptoms (I feel my life is hopeless) and some reflected the opposite (I find joy in life). When he investigated the two types of items he found that Americans and Japanese responded the same to depressive symptoms, but not to the more positive items where Americans scored higher. One interpretation is that due to Asian modesty, Japanese are hesitant to give high positive ratings to positive things about their lives because it seems like bragging. This is but one potential explanation, but others are feasible.
Scientific Inference Is Never Settled
It is unfortunate that science has become so politicized. As we saw during the COVID19 pandemic, some people demanded acceptance of “settled science”, while others rejected those conclusions. Over time what was considered settled changed as more information became available. This changing of conclusions is a normal part of science as scientists collect new data and reinterpret old. Unfortunately, it can undermine the public’s confidence when those demanding acceptance are less than transparent about what the available information can and cannot tell us.
Science is about collecting and weighing evidence. It is about lively debate among scientists about what results mean and what is most reasonable to conclude. Scientists often disagree, and that disagreement motivates new research and new evidence to be debated. There are many inferences that enjoy consensus, but that does not make them correct, only accepted at a point in time. Science is similar to court cases in that it is all about weight of evidence. New evidence can overturn a guilty verdict just as new evidence can change scientific conclusions. This is why scientific inference is never settled. To accept things unquestioningly as “settled science” is not how science is supposed to work.
Image generated by DALL-E 4.0. Prompt “image of a scientist looking puzzled. older and wider aspect.”
SUBSCRIBE TO PAUL’S BLOG: Enter your e-mail and click SUBSCRIBE