At least in the first instance it looks like there's a minor overestimation, but it depends on what you're looking at. Quoting my comment from previous submissi

At least in the first instance it looks like there's a minor overestimation, but it depends on what you're looking at. Quoting my comment from previous submission: The general effect still exists, yes. It's just reported on poorly in this instance. The 'relative overplacement' in the first table (the extent that participants over-estimated their own performance) is a correlation of 0.04, while the 'actual rank' is a correlation of 0.13, both in comparison to 'subjective social class'. There are a lot of correlations of higher magnitude than 0.04, so it's disappointing that the article attempts to focus on nothing else.