It would be great if we also had some idea what fraction of low-achieving people make resolutions, since then we would be able to draw conclusions about whether making resolutions is a useful thing to do.