1. This note presents supplementary evidence
to the Treasury Sub-committee as promised in Sir John Kingman's
oral evidence on 16 October 2002. It explains the background to
the Statistics Commission's consideration of the extent to which
policy and practice in relation to key stage result statistics
are grounded in the results of research and refers to information
provided to it by the Department for Education and Skills (DfES).
2. Professor Peter Tymms and others raised
concerns about "drift" in standards in key stage 2 literacy
tests and the consequential effect on National Statistics based
on these. The suggestion was not that the figures were
being consciously distorted but that the statistics were not on
a precisely consistent basis from year to year because of the
intrinsic difficulty in standardising across years. In particular,
they were concerned that a small but consistent bias in the method
used for standardising from one year to the next would lead to
a lowering of standards over time.
3. The Commission recognised the wide range
of existing research evidence in this area covering educational
as well as statistical questions. It asked for information about
the extent to which policy and practice in key stage result statistics
as a whole (ie not purely key stage 2 literacy) were grounded
in such research.
4. DfES provided the Commission with a range
of information about the standard-setting process and research
which was relevant to this. The Commission concluded that although
there was some evidence of relevant research being taken into
account in formulating policy and practice, this area would benefit
from a systematic and rigorous review as part of the National
Statistics Quality Review process. The chairman wrote to the National
Statistician on 1 October 2002 to request this.