3 ms·

Misinformation here refers to "links to untrustworthy sites". Maybe the term misinformation and the designation of certain sites as "untrustworthy" are develop

Misinformation here refers to "links to untrustworthy sites". Maybe the term misinformation and the designation of certain sites as "untrustworthy" are developed to be politically useful in the US? That would explain that they apply less to politicians in other countries.

RichardCNormos 4y ago
They are unclear about the definition of "misinformation", making it completely subjective by designating sites as "untrustworthy" because it allows them to fit the "study" to the outcome they want.
giuliomagnifico 4y ago
> To determine the trustworthiness of information shared by the politicians, we extracted all links to external websites contained in the tweets and then used the NewsGuard database to assess the trustworthiness of the domain being linked to > We repeated our analyses using a second database (https://github.com/JanaLasser/misinformation_domains https://github.com/JanaLasser/misinformation_domains) of news website trustworthiness instead of NewsGuard. This robustness check was important to minimize the risk of possible partisan bias in what is considered “untrustworthy.” >The second database was compiled by academics and fact checkers such as Media Bias/Fact Check (https://mediabiasfactcheck.com https://mediabiasfactcheck.com). Reassuringly, the results matched our primary analyses and we found the same trends.
- amadeuspagel 4y ago
  Even setting aside the idea of political usefulness and partisan bias (not quite the same, something might be politically useful by discrediting other factions within one's own party), the fact that both of these databases are made by americans is in itself a completely sufficient explanation for the main result of this study. Databases made by americans are going to disproportionally contain websites that tend to be linked by american politicians. But, for the record, I don't think it proves much that two databases lead to similar results.
- marcusverus 4y ago
  This is why we should teach logical reasoning in school... Bob and Jill both tweet about politics. Bob is 7x more likely to tweet from sources which experts say are more likely than average to spread misinformation. Which of the following is true. A) Bob tweets more misinformation than Jill. B) Jill tweets more misinformation that Bob. C) The presented data is insufficient to make a determination either way. The correct answer is C. In order to select A or B, you'd need to offer up data about the individual articles being tweeted by Bob and Jill. By relying on data about the underlying domains at large, you rely on the ASSUMPTION that Bob and Jill tweet randomly from their respective sources, without any agency whatsoever. By pretending that the presented data is sufficient to say that Republicans tweet more misinformation, Neimanlab is, in fact, spreading misinformation, based on poor logic and lazy analysis.
  - nescioquid 4y ago
    > By pretending that the presented data is sufficient to say that Republicans tweet more misinformation... Or, you'd also have to admit that the conclusion may yet be correct, even if it isn't justified by the data they've collected and the analysis. Even an unjustified true belief is still true (I'm not saying the article/study is correct, just based on your analysis we also don't know that it is actually a false conclusion). But your point is an interesting one in that when dealing with "misinformation", you'd expect some case-by-case assessment rather marking a tweet misinformation because it linked to a domain that is "more likely than average to spread misinformation". I imagine this is really a proxy for whether the outlet depends primarily on official sources (e.g. access to government officials) or whether it contradicts them. In which case, you'd have to ask yourself why a politician would choose one outlet or another?

RichardCNormos 4y ago

They are unclear about the definition of "misinformation", making it completely subjective by designating sites as "untrustworthy" because it allows them to fit the "study" to the outcome they want.

giuliomagnifico 4y ago

> To determine the trustworthiness of information shared by the politicians, we extracted all links to external websites contained in the tweets and then used the NewsGuard database to assess the trustworthiness of the domain being linked to > We repeated our analyses using a second database (https://github.com/JanaLasser/misinformation_domains https://github.com/JanaLasser/misinformation_domains) of news website trustworthiness instead of NewsGuard. This robustness check was important to minimize the risk of possible partisan bias in what is considered “untrustworthy.” >The second database was compiled by academics and fact checkers such as Media Bias/Fact Check (https://mediabiasfactcheck.com https://mediabiasfactcheck.com). Reassuringly, the results matched our primary analyses and we found the same trends.

amadeuspagel 4y ago
Even setting aside the idea of political usefulness and partisan bias (not quite the same, something might be politically useful by discrediting other factions within one's own party), the fact that both of these databases are made by americans is in itself a completely sufficient explanation for the main result of this study. Databases made by americans are going to disproportionally contain websites that tend to be linked by american politicians. But, for the record, I don't think it proves much that two databases lead to similar results.
marcusverus 4y ago
This is why we should teach logical reasoning in school... Bob and Jill both tweet about politics. Bob is 7x more likely to tweet from sources which experts say are more likely than average to spread misinformation. Which of the following is true. A) Bob tweets more misinformation than Jill. B) Jill tweets more misinformation that Bob. C) The presented data is insufficient to make a determination either way. The correct answer is C. In order to select A or B, you'd need to offer up data about the individual articles being tweeted by Bob and Jill. By relying on data about the underlying domains at large, you rely on the ASSUMPTION that Bob and Jill tweet randomly from their respective sources, without any agency whatsoever. By pretending that the presented data is sufficient to say that Republicans tweet more misinformation, Neimanlab is, in fact, spreading misinformation, based on poor logic and lazy analysis.
- nescioquid 4y ago
  > By pretending that the presented data is sufficient to say that Republicans tweet more misinformation... Or, you'd also have to admit that the conclusion may yet be correct, even if it isn't justified by the data they've collected and the analysis. Even an unjustified true belief is still true (I'm not saying the article/study is correct, just based on your analysis we also don't know that it is actually a false conclusion). But your point is an interesting one in that when dealing with "misinformation", you'd expect some case-by-case assessment rather marking a tweet misinformation because it linked to a domain that is "more likely than average to spread misinformation". I imagine this is really a proxy for whether the outlet depends primarily on official sources (e.g. access to government officials) or whether it contradicts them. In which case, you'd have to ask yourself why a politician would choose one outlet or another?