The effectiveness of an intensive individual tutoring programme (Numbers Count) delivered individually or to small groups of children: a randomised controlled trial

Abstract

Introduction: In this article, we present the results of two small randomised controlled trials (RCTs) investigating the relative efficacy of a one-to-one numeracy programme, Numbers Count (NC). This intervention was developed as part of the Every Child Counts programme to specifically target the lowest achieving children at key stage 1 (KS1) mathematics. The research question focused on the relative efficacy of NC delivered either individually (as originally intended by the developers of the programme) or in an adapted version to small groups of two or three children. Design and methods: In 15 schools 75 children in year 2 identified by the schools as being eligible to receive NC were randomised to receive it individually or in pairs and in 7 schools 54 children in year 2 identified by the schools as being eligible to receive NC were randomised to receive it individually or in triplets during the school year 2009–2010. The design of the trial required five children, for the pairs sub-trial, or seven children for the triplets sub-trial in each school to receive NC individually or in pairs/triplets in autumn term 2009 or spring term 2010. Results: The primary outcome measure Progress in Maths 6 (PIM 6) was undertaken and marked blind to group allocation by independent testers. We found no statistically significant difference between the scores of the children taught individually or in pairs in terms of PIM 6 scores although a slight difference in favour of pairs was observed. We also found no statistically significant differences between children taught individually or in triplets in terms of PIM 6 scores. We pooled the effect sizes for individual versus pairs and triplets delivery in a meta-analysis of individual versus small group teaching. The pooled effect size was −.26 (CI −2.18 to 1.65) which demonstrates no statistically significant difference between individual and small group teaching. Conclusion: There was no evidence of a difference between the groups, which was not unexpected as our sample sizes were relatively small. This is the only robust evidence from RCTs of the promise of the intervention. However, given the small sample sizes of the trials, we recommend further larger trials should be undertaken comparing small group mathematics teaching with one-to-one educational interventions, using both this programme and others. These two trials could also usefully be included in meta-analyses comparing small group and one-to-one teaching in mathematics education.

Keywords:

• experimental design
• Numbers Count
• mathematics education
• education research

Acknowledgements

The trial described in this article was funded by the DCFS. We also acknowledge the University of York and Durham University for additional funding to support the trial. The corresponding author held posts at the Universities of York and Birmingham during the period when the research took place and currently holds a position at Durham University.

Additional information

Notes on contributors

on behalf of the Every Child Counts independent evaluation team

Torgerson, C.J., Wiggins, A., Torgerson, D.J., Ainsworth, H., Barmby, P., Hewitt, C., Jones, K., Hendry, V., Askew, M., Bland, M., Coe, R., Higgins, S., Hodgen, J., Hulme, C. and Tymms, P.