ABSTRACT
This study reviews the European evidence on the impact of professional development (PD) of pre-school educators on child outcomes. A meta-analysis investigates how PD of pre-school educators in formal pre-school centers in Europe affects child outcomes. The European studies are quite recent and limited in numbers, and our results show a significant positive effect of PD on child outcomes with an overall effect size of 0.35 (with a 95% confidence interval from 0.20 to 0.51). The magnitude is slightly smaller than corresponding results based on US studies, but indicates a general positive effect of PD on child outcomes.
Acknowledgements
This study was conducted as part of the Curriculum Quality Analysis and Impact Review of European ECEC (CARE) project. We thank the European Commission for financial support within the Seventh Framework Programme. We would like to thank Lars Lund-Thomsen and Anne Catharine Andersen for assisting with the systematic search for studies, and Yvonne Anders, Bente Jensen, Paul Leseman, Ted Melhuish, Hannah Ulferts and the rest of the CARE-group for providing very helpful comments. Finally, we would like to thank four anonymous referees for their helpful suggestions and comments. The usual disclaimer applies.
Disclosure Statement
No potential conflict of interest was reported by the authors.
ORCID
Peter Jensen http://orcid.org/0000-0002-5179-6884
Astrid Würtz Rasmussen http://orcid.org/0000-0003-2540-4178
Notes
1 An overview of coaching-based PD interventions is provided by Powell and Diamond (Citation2010).
2 We thus do not compare pre-schools with, for example, home care.
3 See also the later discussion of Inclusion Criterion 2.
4 These databases were chosen based on recommendations by university librarians. The databases cover a wide range of studies in education research, psychology, economics, and so on.
5 We have decided not to include Russia and Turkey in our definition of Europe as both countries are large and span two continents, Europe and Asia.
6 Many studies on PD in early childhood only investigate teacher outcomes. In Schachter (Citation2015), for example, only 49% of the included studies contain measures on children’s outcome; 40% of the studies use measures of children’s learning; while only 11% use measures of children’s behavior.
7 See earlier discussion of start- and end-dates under “search strategies”.
8 For studies with unequally sized treatment and control groups, the sample sizes are used to calculate the within-groups standard deviation, pooled across groups.
9 This add-on command on meta-analysis provides similar results and outputs as, for example, the meta-analysis program developed by Borenstein. It provides forest plots, tables with effect sizes and the overall effect, tests of common hypotheses, and so on.
10 For some studies this strategy was not possible due to lack of information. For example, we had to exclude the study by Ahsam, Shepherd and Warren-Adamson (Citation2006) because it was impossible to calculate an effect size based on their results.
11 To investigate the importance of (1) the domain and (2) the intensity of training, we conducted a suggestive moderator analysis on each of these parameters, and we discuss the results in the final Discussion and Conclusion section.
12 Many UK studies do not rely on interventions with pre- and post-intervention tests as is, for example, the case for the British Effective Pre-School and Primary Education (EPPE) project study (see Sylva et al., Citation2010).
13 Reception classes (ages 4 to 5) as well as multiyear classes (ages 3 to 7) were included, but we could not separate the 7-year- olds from the rest of the children. Since they do not dominate the groups (the mean age of children in both groups is less than 5 years), we decided to include this study.
14 Three of the four effect sizes were above 0.8 and thus quite large.