Private tutoring - does it help? And if so, how?

(Translation of an article in: IPN Journal No 5)

Are the hopes that so many place in tutoring justified?

Karin Guill

Tutoring is very common. Almost every second teenager receives additional tutoring during his school career. Analyses by the IPN raise questions as to whether students receiving tutoring improve their school performance. But: Does tutoring perhaps have other effects?

The number of registrations for private tutoring increases no later than after the first semester grades are in. According to a now classic definition, this is instruction which takes place both outside the family and outside regular school hours, is used regularly and often temporarily, is given by teachers, university students, (older) secondary school students or other laypersons, is usually fee-based and aims to ensure success in certain subjects. "Usually fee-based" means that it includes some students who receive free tutoring at home, from people their parents know. Research which deals with private tutoring, faces the challenge that questionnaires regarding private tutoring unfortunately often rely more on an intuitive understanding of private tutoring and that one assumes that everyone means roughly the same thing when referring to private tutoring. Thus, a tutoring assistance offered by schools sometimes falls into the category of private tutoring.

According to analyses by the German Institute for Economic Research (DIW), almost one in two young students presently receives paid private tutoring at least once during their school career, and the increase since the beginning 2000s, when not even one in three received private tutoring, is remarkable. These figures seem to reflect a high degree of confidence that the additional support will prove effective. When asked directly, the vast majority of students and parents are convinced that tutoring has improved or "more likely" improved school performance. More likely improved? There remains an uncertainty that is also reflected in a number of studies on the effectiveness of tutoring. These studies compared children and adolescents who either received tutoring or did not receive it. The children and adolescents examined had similar previous achievements, a comparable interest and motivation in the privately tutored subject and a similar family background, and therefore could probably find comparable support at home. In Germany, at least, no positive effect of tutoring on school performance can be demonstrated in the studies, which include the whole range of factually used tutoring.

We were recently able to replicate this finding with data from the National Educational Panel Study (NEPS): Here, 10th grade students were interviewed halfway through the second half of the year. Twenty-three percent (23.6%) said they were presently receiving tutoring. By means of regression analyses we could show that students who received tutoring at the end of the tenth grade nominally even received slightly worse marks in mathematics (b = -0.173, p < .05) and German (b = -0.186, p < .05) than comparable students who did not receive tutoring, even if the half-year grades and some other characteristics were statistically controlled. At least some of the weakest students with poor half-year grades had benefited from the additional lessons (Interaction tuition*Half-year mark in mathematics: b = -0.141, p < .05 and German: b = -0.164, p < .05). Although a large number of the differences between teenagers who received or did not receive tutoring were already controlled for in these analyses, such findings can have many causes: Perhaps there were only subtle differences in performance: The students who received tutoring perhaps more often had a grade of 4- (comparable to an US-American D -) in their half year report card, those without tutoring had a grade of 4 (D). Our analyses cannot reflect if the students then improve to a straight 4. Perhaps the students without tutoring have exerted themselves or received more intensive and better support in their private environment than the tutored students. Nevertheless: The analyses raise doubts as to whether private tutoring can achieve substantial advantages not achievable in the same way by other and more cost-effective means.

Despite these sobering findings, almost all students in NEPS state that they are completely or rather satisfied with their tutoring. Where does this satisfaction come from? And can characteristics of teaching be identified that contribute to performance enhancement?

NEPS permits more in-depth analyses as students were asked to assess the quality of teaching in a private tutoring subject on the basis of eleven statements. Tutoring is strongly oriented towards school lessons: Homework is worked on and discussed, teaching material is worked on, and exercises are practiced for class tests. It therefore makes sense to describe the quality of teaching using the same dimensions as those used to characterize school teaching:

• cognitive activation or challenge, i.e. the mental challenge of the learner by presenting complex tasks or making him justify approaches to solutions
• constructive support, i.e. characteristics of the relationship between student and teacher such as individually supportive teacher behavior, positive and constructive feedback and a positive approach to student mistakes
• structure, i.e. structured material presentation and effective time management.

Can students judge their lessons in such a differentiated way? The question must be answered with a "yes/no", because by means of confirmatory factor analyses we were able to show that their assessments can indeed be described well with a model using these three dimensions.

But all three dimensions are highly correlated, i.e. those who perceive a lot of support from their tutor tend to find their lessons more challenging. The demarcation between the three dimensions is therefore rather difficult for the teenagers.

We also examined whether there were any differential effects of the above-mentioned quality dimensions of tutoring on school grades and students' satisfaction with their school and family situation. This results in an interesting pattern: None of the quality dimensions are related to year-end school grades when the half-year school grades, motivation and family background are statistically controlled. However, research on school teaching has shown that it is difficult for students to assess how cognitively activating lessons are. Exactly this dimension is particularly important for performance development. In this case, interviews with tutors or video analysis would be a better approach.

Students, on the other hand, can assess the teacher's support very well. We found clear effects here: Those who have a supportive and appreciative tutor rate their school and family situation more positively. Thus, if poor school performance becomes a burden, the relationship with a third party who is not angry but supportive can bring substantial relief if the teenagers do not meet the school or parental standards. In the best case, this relief could then be used productively to increase one's own efforts. However, this has not yet been proven.

Tutoring therefore has a positive effect on adolescents' satisfaction with their school and family situation. This may already justify the financial investment for some families. In the case of poor school performance or if private tutoring is subsidized by the state, for example within the framework of education and participation benefits in Germany one would also wish for a substantial positive effect on school performance. Current research raises some doubts as to whether tutoring is the right strategy. The question arises whether one should resort to support measures based on pedagogical theories of subject instruction, positively evaluated and available to all students within the framework of all-day schooling.

>If poor school performance becomes a burden, the relationship with a third party can bring substantial relief<

THE NATIONAL EDUCATIONAL PANEL STUDY

The National Educational Panel Study (NEPS), launched in October 2008, examines the competencies of individuals from different age cohorts on a longitudinal basis over their lifespan. The age range considered in the study ranges from kindergarten to adulthood.

The study was commissioned by the Federal Ministry of Education and Research to be conducted by an interdisciplinary, Germany-wide network of excellence under the auspices of the Leibniz Institute for Educational Trajectories. (LIfBi) at the Otto-Friedrich-University Bamberg. The IPN is part of this network and represents the research area competences in the areas of mathematical competence, scientific competence and computer literacy.

This work uses data of the starting cohort grade 9, doi: 10.5157/NEPS:SC4:9.0.0.