Evidence Check 1: Early Literacy Interventions - Science and Technology Committee Contents

4  Conclusions

85. This first Evidence Check has been a useful exercise. We have discovered that the Government's focus on early literacy interventions and phonics-based teaching is based on the best available evidence. We have also found that the Government's use of Reading Recovery is based on evidence, but a lower quality of evidence than we, as a Science and Technology Committee, are comfortable with. The Government's decision to roll out Reading Recovery nationally to the exclusion of other kinds of literacy interventions was, however, not evidence-based, and we have suggested that the Government should commission some high quality research, such as randomised controlled trials, in this area.

86. We have identified the Government's approach to teaching children diagnosed with dyslexia to read—namely, a structured phonics-based programme—is evidence-based on the best available evidence. But we discovered that the evidence base could be much stronger in this area. The Government's focus on dyslexia, from a policy perspective, was led by pressure from the dyslexia lobby rather than the evidence, which is clear that educational interventions are the same for all poor readers, whether they have been diagnosed with dyslexia or not.

87. In broad conclusion, we found that there was a willingness from the Department to base its approach to early literacy interventions on the evidence. However, we discovered worryingly low expectations regarding the quality of evidence required to demonstrate the relative effectiveness and, in particular, the cost-effectiveness of different programmes.

88. By way of final comment, we reflected on possible reasons for these low expectations of research quality and turned to the Government guidelines on social science research and policy evaluation, which are found in the Magenta Book.[107] We found that the sections on randomised controlled trials only consider an RCT design that places an experimental group against a non-experimental group.[108] Even when considering the ethical problems of assigning people to a non-experimental group,[109] it fails to consider the alternative RCT design where experimental group is compared against experimental group. We, among others, have suggested that directly comparing experimental groups is the most appropriate test for ascertaining the relative effectiveness of different literacy interventions (see Chapter 2). As a yardstick, even Wikipedia is more thorough and informative than the Government's guidelines on randomised controlled trials in that it considers a range of RCT designs.[110] It is hardly surprising that civil servants who rely on a set of guidelines that do not describe the full range of viable research options would have low expectations regarding the quality of social science research and evidence. We recommend that the Government review its Magenta Book with a view to raising its expectations of social science research and evidence in relation to policy.

107   Ev 56; Government Social Research Unit, The Magenta Book: guidance notes for policy evaluation and analysis, 2007 Back

108   Government Social Research Unit, The Magenta Book: guidance notes for policy evaluation and analysis, 2007, sections 1.3.6 and 7.2-7.5 Back

109   Government Social Research Unit, The Magenta Book: guidance notes for policy evaluation and analysis, 2007, section 7.5.3 Back

110   http://en.wikipedia.org/wiki/Randomized_controlled_trial Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2009
Prepared 19 December 2009