The authors misinterpret their own statistics.

Daniel Västfjäll with several writes in his op-ed that the results of their study shows that reasoned thinking on immigration and brottsbenägenhet occurs among the participants, that it is found both in the group of “swedes” and in the group of “global citizens”, and moreover, that there is “as much to both sides”.

We believe that the last of these statements is incorrect: the study does not show that there is “as much to both sides', but on the contrary it is present in significantly lower degree, among “världsmedborgarna” in the study.

In particular, the u.s. studies have shown that values affect people's ability to interpret statistical data: participants draw conclusions of statistical material that is incorrect but that fits better with their political views. We fully agree with Västfjäll with several to a high frequency of reasoned thinking among citizens in a democratic society is problematic. We also agree that it should be studied more, and also on the american bottom. This latter has Västfjäll with several creditable done. However, we believe that they have misinterpreted their own statistics.

the Authors write: "When the figures showed that crime increased most in the areas with refugee reception had people who saw themselves as 'more of an american' [...] 18 percentage points higher probability to interpret the information correctly compared with people who saw themselves as 'more of a citizen of the world'. [...] When the figures showed the opposite correlation had people who classified themselves as 'more of a citizen of the world' 20 percentage points higher probability to interpret the numbers right, compared with those who classified themselves as 'more of an american'. The conclusion of this is not only to reasoned thinking occurs, but that there is as much to both sides."

“the Swedes” draws about 27 percentage points more correct conclusions when the numbers agree with their views, than when they do not match. The corresponding difference for the “världsmedborgarna” is close to 12 percentage points.

the Results show indeed that there is reasoned thinking, but they don't show how it motivated thinking is distributed between the groups. Each group is about 20 percentage points better than the other when the numbers are in accordance with the group's views, of course, is consistent with reasoned thinking is distributed equally between the groups. But it is also compatible with that it occurs more in one group or with only one (let for example, the “swedes” have 60 percent right in both cases — no reasoned thinking at the level of the group — while the “världsmedborgarna” has been 80 per cent right in one case and 40 per cent right the opposite). The conclusion that there is “as much to both sides“ is therefore not supported by the results, the authors invoke. They draw the conclusion in its scientific article.

“the Swedes” draws about 27 percentage points more correct conclusions when the numbers agree with their views, than when they do not match. The corresponding difference for the “världsmedborgarna” is close to 12 percentage points. Should these results be compared with results in control tests, but the control tests reported in the scientific article does not suggest that the differences between the groups are smaller. In this study thus appears to opinions affect “the” statistical interpretation more.

the Reasoned thinking is a form of kunskapsmotstånd. Kunskapsmotstånd constitutes a serious social problem, and therefore, it is important not to give the impression that it is more widespread than it is.

the Authors participating in the research programme ”Kunskapsresistens: Causes, consequences and countermeasures”, funded by Riksbankens jubileumsfond, and is led by Åsa Wikforss. The aim of the programme is to investigate the kunskapsresistens, its nature and causes. Researchers in philosophy, psychology, political science and media and communication research are cooperating in the program.