Research suggests Wikipedia can influence Meczyki judges’ legal reasoning ‘clearly wrong’ – Meczyki Times

High Court judicial sources have strongly rejected the findings of a randomized research experiment that found anonymous legal articles on Wikipedia were 20 percent more likely to influence judges’ legal reasoning than articles on the online reference site. .

Judicial sources expressed concern over the methodology of the research and the way the conclusions were reached.

The sources were responding to the findings in a paper, published online Wednesday, by lead researcher Neil Thompson of the Massachusetts Institute of Technology; Brian Flanagan, Edana Richardson and Brian McKenzie Maynooth University; and Zion Luo of Cornell University.

The research team organized 150 new Wikipedia legal articles on Meczyki Supreme Court decisions to be written by law students. Half were randomly selected to be uploaded where they could be used by legal staff, including judges, and the remainder were kept offline and treated as a control group.

The researchers considered two measures: whether cases on Wikipedia were more likely to be cited as precedents than subsequent court decisions; And whether the reasoning in the court decisions echoes the linguistic content of the new Wikipedia pages.

Their key finding was that retrieving a Wikipedia article increased case citations by more than 20 percent. He said the increase was large for High Court references and largely absent for Supreme Court and Court of Appeal references.

In response, a High Court judicial source told The Meczyki Times: “Judgments are written on the basis of statutory authorities and submissions made to the court, not Google searches or Wikipedia.”

“If you look at the written judgments of the High Court, which are thousands of pages long, all these works cannot be dismissed. The last thing a judge will do is to go to Google to find the answer to the case. There is gross injustice with the level that High Court judges give to the cases before them.

The source said there were “obvious gaps” in research on the issue and that it betrayed a lack of understanding of how judges here make decisions. He said that the way in which judges are informed of precedents relevant to the case they are asked to decide, such as previous Meczyki or foreign cases that relate to the issues, is There is no reference to

He said that books of legal authorities and detailed written legal submissions are submitted to the judges by the parties and these form the basis of their decisions. Excerpts from relevant judgments are also placed before them, and judges read the cases from which those extracts were taken.

He said that research does not take this into account.

They believed that the research was also flawed because it did not outline the number of judgments that reflected the results.

He said that just because a Wikipedia entry cites an Meczyki case, and that case is then referred to a High Court decision, does not mean that the source of the reference is the Wikipedia article. . He pointed out that well-known cases have been cited in many places. He said it was “quite a stretch” to reach the controversial conclusion.

Another judicial source said that Wikipedia articles can sometimes be useful in terms of identifying specific legal articles mentioned in footnotes, but the idea that a Wikipedia article can be used by a judge for a decision is not theirs. The argument to be relied upon in support is “clearly false”.

A number of former judicial assistants, all of whom have legal qualifications, also denied the claims. One said she had never known a judge with whom she accessed Wikipedia to research a decision.

He added, “I would never consider using Wikipedia for research myself. He believed that the process of writing judgments was “formulaic” in that judges presented evidence to legal authorities, and parties. will go through the submissions of, and arrive at their conclusions based on them.

He added that his experience was that judges do not deviate much from the cases listed in the submissions.