Overprecision is the excessive confidence that one knows the truth. For reviews, see Harvey
[13] or Hoffrage.
[14] Much of the evidence for overprecision comes from studies in which participants are asked about their confidence that individual items are correct. This paradigm, while useful, cannot distinguish overestimation from overprecision; they are one and the same in these item-confidence judgments. After making a series of item-confidence judgments, if people try to estimate the number of items they got right, they do not tend to systematically overestimate their scores. The average of their item-confidence judgments exceeds the count of items they claim to have gotten right.
[15] One possible explanation for this is that item-confidence judgments were inflated by overprecision, and that their judgments do not demonstrate systematic overestimation.
The strongest evidence of overprecision comes from studies in which participants are asked to indicate how precise their knowledge is by specifying a 90% confidence interval around estimates of specific quantities. If people were perfectly calibrated, their 90% confidence intervals would include the correct answer 90% of the time.
[16] In fact, hit rates are often as low as 50%, suggesting people have drawn their confidence intervals too narrowly, implying that they think their knowledge is more accurate than it actually is.