Since you didn't include the options, I will answer the question as well as I can: the one factor I would be considering is whether the journal is known to be biased, for example whether it's a peer-reviewed journal ( those are less biased) or whether they journal is paid by some interest group.
Otherwise I would look at whether the data is reliable, for example whether the author cites the sources for their data
