It is worrying that there continue to be such large discrepancies between pupils' actual attainment, which has stagnated since the 2005 Scottish Survey of Achievement, and teachers' judgment of their attainment - discrepancies which have been found throughout the curriculum, and seem to be getting worse.
There seem two likely explanations. The first is that teachers have an imperfect understanding of assessment criteria. If assessment is truly to be for learning, as hoped for in the Curriculum for Excellence, this suggests further training and advice is needed before the system achieves that aim.
Some might claim the fault lies in assessment criteria (of national tests and the Survey of Achievement). But considering the care that has gone into refining these over the years, it is incumbent on them to show why teachers' judgments are better.
The second is that objective, independent assessment finds pupils performing at a lower level than on tests they have a chance to prepare for and repeat in class. Confirmation is found in the assessment of writing, which has a lower discrepancy than reading.
Writing is assessed here by pieces of work submitted by teachers, then independently assessed. Unlike reading, this is not an exam-like test. If this is part of the explanation, then serious doubts are cast on the validity and reliability of continuous assessment of the kind that has dominated 5-14 in Scotland since 1991.
That so few pupils enjoy reading - around one in five at S2 - is perhaps the most serious finding. We know from much international research that reading for pleasure in the teenage years is excellent preparation for high attainment across all domains, and for life as a critical citizen in a democracy.
If Scottish pupils are not developing this habit, then their intellectual and cultural awareness will be stunted, and their capacity to exercise their democratic responsibilities severely limited.