A STUDY'S TRANSPARENCY
[PT: A study's transparency]
Okay, I've just had an issue of someone in my own comment section correcting me about something showing me a study.
A study that had the worst formatting, and methodology/source transparency I've ever seen in a study about a demography from a, what it looks like, a source with good intentions. And I'm saying this as someone who's not an expert on seeing this kind of shit, I'm just passionate, so I will miss stuff.
SO TRANSPARENCY 101:
[PT: So transparency 101:]
The methodology has to be explained. If they're going through the news to take the information what makes a news oultet count/be included is important. What outlets were used is important and should be referenced. Even the ones that had no information that was used, if they were used for the pool of information it should be there. Knowing what outlets were taken out shows you why they didn't (which in had studies shows the bases and what they didn't want to count).
Everything, every bit of information that comes from somewhere (a social media post, a Newspaper, another study, etc) should have the source linked in the bibliography.
If theres more than one source of information —different kinds like more right wing or left wing, different types like social media posts/news outlets/people reaching out to them and reporting it/etc— the statistics of what parts of the demographic coincide with each outlet is important. And the amount of effort to used each source is also very rellevant. As if group A is very reported but they're mostly from news outlets, and group B is almost insignificant of how low they are and they were mostly from people reaching out, and for the news there was and algorithm but for the reporting it wasn't very advertised/known, data on group B wont be showing what they're actually living through.
Any ambiguos wording about how something is done is a part of information they're actively not giving you for a reason. Again, things have to have a source.
If most of the sources they give are linked to themselves? Bad. If you have to go through the links to themselves to then inside there find the original source most of the time? Bad. Or they're hiding something and want you to take too much to do It so you don't look deep, or they're very badly administrized and formatted which if added to it there's a lack of sources and methodology transparency means it's very probable it's the same for how they did the study itself, which is not a good luck for a study to be badly administrized if it's about data/statistics. What if they accidently changed things? What if they accidently deleted information? What if they didn't organinize the methodology right and didn't actually fet a good sample because of a bad algorithm (like not talking into account other wordings or types of news outlet)?
Every demographic intersection should be added from the pool of data. And if demographic 1 is shown to be intersected in a specific way with demographic A, demographic 2/3/... should also be shown in relation to A. Even if to shown there's no significant importance with one intersection but there is with another.
Do not believe a study if one or more of these things is not done right.
A big red flag is also not having a version of the report in an academic study format.
If anyone wants to add anything I've missed I'd love it! We news to fet educated on media literacy, especially about studies done in bad faith that hide behind progressive/academic language.
