The falling standards of education in American public schools have increased demands for accountability and performances in schools. Consequently, school administrators evaluate ways of improving academic performances, such as tutoring programmes. There are a number of tutoring programmes like peer, volunteer, one-on-one and cross-age tutoring.
There have been growing evidences with regard to volunteer tutoring. Consequently, this paper presents a systematic review of The Effectiveness of Volunteer Tutoring Programs for Elementary and Middle School Students: A Meta-Analysis by Ritter and other authors.
The review of meta-analysis seeks to establish the effectiveness of volunteer teaching programmes that aim to enhance the academic skills of student in public schools of the United States. The review also aims at establishing who can benefit from tutoring, and factors that influence the effectiveness of tutoring. These constitute the main of objectives of the studies by Ritter and other authors.
The authors used 21 studies consisting of 28 distinct cohorts in their studies in providing guidance for assessments of the effectiveness of volunteer tutoring programmes. Through these assessments, the authors concluded that volunteer tutoring had a positive effect on students’ performance.
They discovered that students who had tutors got higher marks in specific sub-skills such as oral fluency, letters and words, and writing. However, students who had no tutors scored relatively low marks compared to their counterparts with tutors.
The authors used several search strategies to gather evidences to support their studies. These sources included academic sources, primary sources, middle search sources, collections of professional studies, research in the field of psychology and behavioural science, literature in sociology, education information database such as ERIC, and proquest database.
At the same time, these authors used selection criteria based on randomised field trials. The authors used these methods to enable them make comparisons with the control groups that did not participate in the studies from the year 1985 to 2005.
Meanwhile, the authors were careful to include only studies that had academic effects, and ignored other vital factors such as emotion and behaviour. The study included programmes for students in grades of K-8 with the adult non-professional volunteers.1
The authors based the results of the review on information from the 28 cohorts in the study using 21 articles and reports. The results of the analyses, mainly consisting of small samples, indicate that volunteer tutoring and coaching have positive effects on reading and language in students. Examinations of the overall consequences of volunteer tutoring programmes on reading outcomes of student showed improvements.
The authors used 25 studies to assess the reading outcomes and effects of volunteer tutoring programmes where they found an effect size of 0.23. However, when they eliminated an outlier study that disproportionately influenced the results, the effect rose to an average of 0.30.
The study meta-analyses also looked at specific academic areas. The volunteer tutoring also made significant improvement on reading where the effect size was 0.26. Letters and words had an overall effect of 0.41 while oral fluency had a significant effect size of 0.30, and writing effect was 0.45. The study also found positive effects on reading comprehension and mathematics. However, the effects of these two were not significant.
The meta-analyses also revealed that there were no significant differences with regard to grade level, programme of focus, or tutor type. Detailed volunteer tutoring programmes had vital and greater effects on reading outcomes than lowly structured programmes.
However, these did not affect other academic domains. The authors also established that there were no statistical significance and differences in published reports and non-published studies. They also found out that the published texts had no biases.
There are some grey areas in this work. Therefore, the authors must update their works and conduct further researches, and possibly contact experts in volunteer tutoring. At the same time, they should also extend their studies to other geographical areas in order to establish consistency in their conclusions and recommendations.
These authors also used studies based on quasi-experimental designs. Quasi-experimental designs may lead to biases in publications. This is because such studies may test effects on an intervention without using appropriate assignment to create groups of the study.
Ritter and others used small samples in their studies. There were total of 28 cohorts in meta-analysis. However, 19 out of the 28 cohorts had less than 25 students in the group. Only 3 study cohorts had a full sample of over 100 students for both treatment and control groups.
This creates small chances of detecting significant programmes’ effects on students’ achievements. However, meta-analysis enhances this shortfall by using cumulative mean and standard deviations in enhancing statistical strength. The authors also did not show homogeneity in volunteer trainees and learner characteristics as these may affect the emotions or behaviour of the treatment groups.
The meta-analyses of the study established that volunteer tutoring programmes have positive effects on students’ performances especially in literacy. These authors also showed the significant of using meta-analysis in conducting studies. Meta-analysis enabled the authors to perform multiple analyses and establish the statistical significance in 5 areas out of 7 areas studied. These results were not possible in standalone studies.
Data collection and analyses reveal details of methodology and procedures such as design, outcomes of analysis, duration of study, and sample background variables such as age, gender, and race, among others. The authors coded all these aspects of studies for analysis of various volunteer tutoring programmes, and for study characteristics.
According to Littell et al, “we should train coders and assess the reliability in coding. At the same time, we must pay special attention to the problem of study quality and assessment. In other words, we must be keen on how we identify variations in quality among the studies included in a systematic review and meta-analysis”.2 For instance, these authors must review all the outcomes that the studies aim to measure.
They also selected a suitable effect size for meta-analysis. In order to avoid biases, they used Hedges’ unbiased measurement of g in the standard mean difference on effect size statistic. These represented differences between the control groups and the study groups mean on the outcome variable that they divided by cumulative standard deviations for every outcome of measurement.
In meta-analysis, we must ask questions relating to variations in effects. Ritter and others do not discuss deeply how publication bias, study design, sample characteristics, intervention characteristics, and assessment of outcomes influenced their studies. The authors did not establish whether the effects had been consistent over time or not. Publication bias arguably is the most potent source of bias in meta-analysis.
However, the authors tried to counteract these biases, particularly where the mean and standard deviations were missing in estimating the effect sizes. They adopted the procedure of Wilson and Lipsey and computed a cumulative mean for effect size.
The study achieved it systematic objective of gathering, integrating, and summarising the empirical work on the effects of volunteer tutoring on students. The authors put up a strong research base compared to other previous studies on educational interventions. This is due to the fact that the reviews covered 21 articles and 28 cohorts dated back to the year 1985 and 2005 in randomised field trials.
Ritter and other authors use a systematic approach to present their study meta-analysis and results. They gathered several articles that effectively used experimental designs to assess the effects of volunteer tutoring on students. The systematic analysis shows that all reading outcomes had positive and statistically significant effect size of 0.30 as standard deviations.
Carole Torgerson notes that systematic review and methodology has the potential to manage potentially unmanageable amounts of information, and rationalise existing evidence efficiently by determining whether research findings are consistent and generalisable, and to explain why if they are not.3
The researchers can group similar studies together statistically in meta-analyses since they know that they can use meta-analyses to increase the power, and precision in the measurement of effect sizes. The application of scientific methods in systematic reviews also enabled them to reduce random and systematic errors of bias.4
The authors applied this method in grouping the specific outcomes in six specific domains of study. Meta-analysis enabled them to create broad domains through the use of standardised assessments in mathematics and reading. Conversely, they also created specific four domains focusing on sub-skills of language and reading.
Through these groups, the authors established that volunteer tutoring programmes have potential influences on academic performances of students and can be used to improve performances. The authors also establish consistency with references to the effect sizes of 0.26 to 0.45 standard deviations. Further, the authors established whether particular interventions were also effective or not.
They demonstrated that every intervention is unique with individual studies using small units. The use of small units was effective in controlling replicas of other works. This enabled the authors to establish the characteristics of most effective interventions.
They were able to establish this through computing differential mean effects sizes using different elements of study variables, such as detailed interventions, type of tutors, and age of tutors.
The authors concluded that the outcomes of their studies should act as vital statistics for educators to use volunteer tutoring in order to improve students’ achievements. They also advised that school administrators should consider the use of structured and reading-based volunteer programmes as suitable strategies in enhancing reading and language skills.
The researchers have effectively demonstrated the power and effectiveness of meta-analysis. Through grouping of various studies, as opposed to standalone studies, they were able to establish statistical significance of relatively large data in their domains.
The research also expands the knowledge we have on the use of systematic reviews and meta-analysis about using data in enhancing the use of interventions in promoting the learning outcomes. These researchers provide useful information that educators can use to implement effective volunteer tutoring programmes to improve students’ achievements.
This is a positive review of the works of Ritter and others. The authors use systematic review and meta-analysis to demonstrate that volunteer tutoring can positively transform achievements of learners, particularly in language and reading skills.
They emphasise that their results should serve as an essential piece of information for educators who want to use volunteer tutoring to improve performances in learners. They further advise that educators should consider structured and reading-based volunteering systems to improve their yearly students’ performances.
We must acknowledge the fact that systematic review and meta-analysis can go some way towards addressing the problem of underdeveloped trials. However, they will not give a true estimate of effectiveness if the trials within the study analyses have methodological flaws. At the same time, meta-analyses may give unduly optimistic results if there is substantial publication bias.
In other words, if studies have null or negative effect remain unpublished, they may not form part of any review. Ritter and other authors address the issue of trials flaws through systematic review by using inclusion and exclusion criteria. They avoided publication bias by ensuring that all data were consistent and avoided unpublished studies that they could not identify.
The authors treated their trials with a high degree of caution by defining all their trials, sources of data among other elements of both treatment and control groups. They did these to avoid chances of overestimating the effectiveness of the volunteer tutor programmes. However, these authors must update their studies regularly and include large samples in their studies in order to yield positive and moderate results.
Therefore, researchers seeking statistical significance must use large sample sizes. The probability of an ‘educationally significant’ difference being statistically significant is partly a function of sample size.
Cooper HM, Synthesizing Research: A Guide for Literature Reviews Applied Social Research Methods Series, vol. 2, Sage Publications, Inc., London, 1998.
Littell JH, J Corcoran & V Pillai, Systematic Reviews and Meta-Analysis, Oxford University Press, Inc., New York, 2008.
Ritter GW, JH Barnett, GS Denny and GR. Albin, ‘The Effectiveness of Volunteer Tutoring Programs for Elementary and Middle School Students: A Meta- Analysis’, Review of Educational Research, vol. 79, no. 1, 2009, pp. 3-38.
Torgerson C, Systematic Reviews, Continuum International Publishing Group, London, 2003.
1 GW Ritter, JH. Barnett, GS. Denny & GR. Albin, ‘The Effectiveness of Volunteer Tutoring Programs for Elementary and Middle School Students: A Meta-Analysis’, Review of Educational Research, vol. 79, no. 1, 2009, pp. 3-38.
2 JH Littell, J Corcoran & V Pillai, Systematic Reviews and Meta-Analysis, Oxford University Press, Inc., New York, 2008.
3 C Torgerson, Systematic Reviews, Continuum International Publishing Group, London, 2003, p. 63
4 HM Cooper, Synthesizing Research : A Guide for Literature Reviews Applied Social Research Methods Series, vol. 2, Sage Publications, Inc., London, 1998 p. 104