Peer Tutoring as an Improvement Strategy for School Exploitation

To determine the effectiveness of a peer tutorial strategy in the school achievement of English to high school students, whose mother tongue is not English, we made a comparison of two groups; in one, peer tutoring was carried out through the support of high average students and compared with the results of a group that did not receive the intervention. The tutors were 121 students who scored higher than 80 points in the initial exam and accepted to participate in the program. The group of tutees was formed by those who obtained scores lower than 65 and accepted to be tutored (101 formed the experimental group and 112 the control group). The peer tutoring was done in the classroom with the monitoring of the teacher, lasted 11 weeks, and focused on taking class notes and comparing them, giving feedback and clarifying doubts for 30 minutes after the teacher's explanation. The results of the midterm exam and final exam were analyzed through the Student's t-test, showing significant differences in the group that was tutored compared to the low-performing students who did not receive the intervention. The procedure was effective, finding that the use of peer support can improve school performance. It is necessary to monitor the profile of the tutors and systematize the tutorial strategy.


Introduction
English language learning is an imminent necessity in most education systems.In Mexico, institutional efforts have been made to promote certification programs for teachers, in addition to establishing state programs in different entities that incorporate subjects in English since elementary education systems.In those related to Higher Secondary Education (HSE) and Higher Education (HE), courses and diplomas that seek students to obtain the basic skills for the mastery of a basic level or an intermediate level of the language have been introduced; however, the effectiveness of these programs is not known and there is data that reflects a low level on the proficiency of English as a second language in Mexico compared to other countries (Education First, 2014).The analysis of the English level indices (González, Lima & Castillo, 2004;O´Donogue, 2015;Castañedo & Davies, 2004;Davis, 2009) confirm a deficit in the programs of public and private schools for English language learning through its courses, and requires a different approach that improves the use of this subject.
Academic tutoring as a strategy to develop grammatical competences and academic achievement in university students has been implemented by different authors.Aguayo, Caballero & Gómez (2017) observed an increase in sociolinguistic skills through a peer tutoring procedure where interaction and teamwork helped to improve communication through a pleasant environment for students.On the other hand, motivation is an intrinsic factor that is put into play when interacting with others (Topping, 1996;Santee & Garavalia, 2006).
Additionally, in recent years, the characteristics of tutoring programs on the effect on school performance in university students were analyzed, finding a relationship between the number of tutoring sessions and the average of final grades compared to the initial ones (Torrado, Manrique -Hernández & Ayala, 2016).Specifically, Duran & Huerta (2008) developed an innovative strategy through a pair organization class identifying more competent students in English to offer help and guidance to others with fewer skills; the results were directed to the improvement in the autonomy of the peer tutors and in the improvement of the use of the students who received the tutorial accompaniment.
Many advantages have been demonstrated through the strategy of peer tutoring; above all, the benefits have been based on the motivation to study, increment in average grades; in addition to being involved values such as solidarity and other skills such as improved communication (Melero & Fernandez, 1995;Rudland & Rennie, 2014;Durán, 2016).It has also shown promoting responsibility, and having an incidence on self-esteem, especially in students who exercise the role of tutors.
Another finding observed in peer tutoring, is the confidence that is generated between tutor-tutee; since there is no fear of raising doubts or making mistakes; which is usually different in teacher-student relation.However, when certain students are assigned an "expert" or "guide" role, a high expectation is attributed that motivates them to fulfill that role; trying to demonstrate its execution through a high commitment to the task (Robinson, Schofield, & Steers-Wentzell, 2005).On the other hand, in the case of the tutee it is a comfortable and friendly space that facilitates an intimate, less formal contact of camaraderie to solve problems and promote learning.
Therefore, peer tutoring is a positive strategy in institutions of higher secondary education (HSE) and higher education (HE) to reduce the delay and school abandonment, improve grades and acquire motivation to study (Torrado-Arenas, Manrique-Hernández & Ayala-Pimentel, 2016).
The model of peer-tutoring (PT) has its origins in Anglo-Saxon universities (Arbizu, Lobato, & Del Castillo, 2005) and the frame of reference is based mainly on the socio-cultural theory of Vigostky through peer learning it's theoretical model is given through what the author called zone of proximal development (Vigotsky, 1989) to glimpse the formation of intelligence using superior psychological processes such as language, attention and memory; among others; these develop through a social dynamic of exchange with others, acquiring a particular learning.Topping (2005) defines peer tutoring as a teaching and learning mechanism through the acquisition of knowledge and skills that occurs through the help and mutual support of peers with similar contextual and social situations, and which main characteristic is learning while helping to learn within a disciplinary area (Gómez, 2013).
Peer tutoring is done in pairs or small groups that assume an asymmetric relationship by adopting a role of tutor for one of the students whose role is guiding, and another one who is tutored; both work on a common goal; where also dominates a dynamic of interaction regulated and accompanied by the teacher.This tutor-tutee bidirectionality offers a valuable feedback for the formation of the students; especially in relation to the organization of learning and collaborative work strategies (Gairín, Feixas, Guillamón & Quinquer, 2004).
The role of the teacher within PT, focuses in defining clear tasks in order to enhance a differentiated pedagogical structure with criteria of place, time, curricular content, contact format and roles of the participants (Topping, 1996).Now, one aspect to consider is the type of course and age; for example, greater ease of implementation has been observed when students of the same age are given organizational and logical aspects than when they are implemented at different ages (Finkelstein & Ducros, 1989).However, there is another group of researchers who recognize as more valuable than the difference or not of age; the previous preparation and the personal skills of the tutor (Baudrit, 2000); the structuring of the teacher of the roles of the dyad (King, Stafieri & Adalgais, 1999;Jiménez, 2015) and the commitment of the student who acts as tutor to perform a prior preparation of each of the tutorial sessions in order to obtain the necessary competencies to exercise his function with his partner (Durán & Vidal, 2004).
Finally, there are different strategies and techniques designed to implement peer tutoring; some of them are cooperative notes comparison; which consists in comparing the notes that have been made after some reading, class or another instructional strategy (Falchicov, 2001).Peer feedback for non-evaluative purposes, but rather to improve the skills of both peers (Levene & Frank, 1993).Peer-to-peer monitoring; that unlike the other two requires students to perform extra-classroom activities (Beaman, Fraser, Diener, & Endresen, 1977).Another procedure is the three-step interview (Nattiv, Winitzky & Dricker, 1991); in which one of the students interviews another and this, in turn, shares it with the rest of the group.
Think-tie-share and think-tie-fit is another system for monitoring individual responses that are shared later (Mills & Cottell, 1998); while card tutoring is considered a reinforcement, practice focused on the development of basic level cognitive skills where answers to specific questions are prepared with card support.Other techniques with greater degree of preparation and cognitive requirement is the comparison test in pairs (Sherman, 1991); the learning cell; method of problem solving (Koch, 1992); mutual reciprocal mentoring (Griffin & Griffin, 1998); reading and exchange of oral summaries (Spurlin, Dansereau, Larson, & Brooks, 1984); among others.The important thing of all of them is to choose them according to the set objectives and the conditions given for their implementation.
The objective of this paper is to analyze the effectiveness of a peer tutoring strategy applied to high school students, in order to increase academic achievement in the English language learning unit.In this research, academic performance is defined as the level of knowledge a student has reflected through a numerical grade obtained through an evaluation (Torres & Rodríguez, 2006) made by a written exam (Vicente, 2000;Ocaña, 2011).

Method
Given the particular conditions of the present study, it was used a quasi-experimental quantitative design due to the characteristics of non-probabilistic sampling and the manipulation of the independent variable (peer tutoring intervention).The evaluation of the participants was carried out in 3 moments, in the first one a measure of the baseline of the students' grades was obtained; followed by two measurements in which the strategy of peer tutoring was applied.The same tests were performed on a control group.Once the measurements were made, comparisons were made after the previous test for independent groups (c 1-3) and an additional comparison for related groups (c 4-5) was made, as shown in Figure 1.The context of the application as well as participants in the present study are described below.

Participants
It should be clarified, that for both, the selection of the tutee and non-tutee participants an informed consent was made where each student was explained the objective, type of participation, activities to be carried out and benefits of the study, making explicit their interest to participate voluntarily.
The participants were chosen considering the following inclusion criteria for each of the study groups;  Students with high academic performance.High school students with a grade higher than 80 in the first partial exam in the English learning unit were invited to be part of the group of tutors during the biannual period August 2014.This same process was continued during the following 5 semester periods.In total, a sample of 121 tutors enrolled in the last year of high school was obtained.
 Students with low academic performance.The sample consisted of 101 students of the last semester of baccalaureate who obtained grades lower than 65 in the first partial exam.This group is called tutees.
 Control group.Students who obtained a grade lower than 65 in the first partial exam and who did not accept to receive the support of peer tutoring.The total number of students in this group was 112.

Instruments
The exam is an instrument used to evaluate learning.Its elaboration was done in a collegiate manner among the professors of the English academy and consisted of 50 items.The exam application was done in 3 moments: initial exam, first partial and indicative exam.The exam application was the same for all students.

Experimental Group Control Group
The participants were chosen for equal conditions (nonprobabilistic sampling) The validation and reliability of the evaluation instrument with which the students were evaluated was analyzed by the University Evaluation Committee, starting with the global application made to approximately 60,000 baccalaureate students.The evaluation committee is made up of experts in the field of English, as well as statisticians who analyze the type of items, included or not, in each of the exams applied.

Procedure
Tutors detection phase.Once the students with high academic performance were identified, they were summoned to a group meeting in which the guidelines and activities that should be done in case of accepting to be tutors were explained; for example: attending classes, making class notes; accompanying their tutee and supporting them during the practice exercises, explaining the topics seen in class and clarifying doubts.For the conformation of the tutor-tutee dyad the students are offered a list with the names of the possible tutors, this facilitates the relational dynamics between them.
Peer tutoring was done during the class hours of the English learning unit, which lasted 50 minutes, of which the first 20 were used for the teacher to explain the class and the last 30 minutes the tutor did a sequence of monitoring through the comparison and complementation of class notes (Falchicov, 2001), explain the basic concepts and dispel doubts when necessary.The duration of this stage was 11 weeks.The teacher strategically made the accommodation of tutors and tutees in the classroom so that they were close to each other in order to monitor them.
Evaluation of the effect of peer tutoring.We proceeded to detect the grades of the first partial (week 5 of class) and the evaluation of the indicative exam (week 11 of class).Once they were obtained, a statistical analysis was performed to compare the scores between the group that received peer tutoring (experimental group) and the group that did not receive it (control group).

Data Analysis
The SPSS statistical package was used to analyze the data.First, tests of normality were used according to the size of the sample, and once the assumption of non-normality was identified, the Mann-Whitney U statistical analysis was performed to compare the measurements before and after the intervention.A subsequent analysis was performed through the Wilcoxon sign test for related samples (experimental group according to the comparison of the measurements before and after the intervention).The type of design can be seen in Figure 1.

Results
In the case of the sample of 213 students identified as "underperforming" and according to the classification of groups of tutees (n = 101) and non-tutees (n = 112), a comparative analysis between both groups was carried out.First, a goodness-of-fit test of Kolmogorov-Smirnov (KS) was performed to determine normality of the data (See Table 1).Table 2 describes the groups compared, the sum of ranks as well as the average range, finding that the initial qualification is equal (z = -5.76,p> 05) between the group of tutees and non-tutees (See Table 3).Contrasts statistics for the grouping variable: low performance.
When making the comparative statistical analysis of the implementation of the peer tutoring program, we can observe a statistically significant difference in the grades obtained in the first partial exam between the control group and the experimental group.Table 4 reflects the analysis made through the rank comparison test, note that the low-performing group that received tutoring obtained an average range of 139.62 while the group that did not receive the tutoring was 77.58.Table 5 shows the analysis with the Mann-Whitney U test (z = -7.34,p <0.05), resulting in a statistically significant difference.Contrasts statistics for the grouping variable: poor performance in the first measurement of the intervention.
With the objective of greater reliability in the results, a second measurement was carried out in both groups, as shown in Table 6.Once the 11 weeks of peer tutoring intervention had been completed, the indicative examination qualifications were identified (final grades) for both groups.In the case of the experimental group, the average range obtained was 140.21, while in the control group was 77.05 (See Table 6), finding a statistically significant difference between the two measurements (z = -7.491,p <0.05) according to the analysis with the Mann-Whitney U test, as shown in Table 7. Contrasts statistics for the grouping variable: poor performance in a second measurement during the intervention.
It is evident that the peer tutoring program received by the group of underperforming students had a favorable effect on their grades.Table 8 shows that the average range of ratings during the initial measurement is lower for both the first partial and the final measurements; and these grades are maintained during the intervention; this can be evidenced by using the Wilcoxon sign test and comparing the measurements indicated, finding a significant difference between the initial qualifications (z = -8.576,p <0.05) with the first partial and the final exam (z = -8.142,p < 0.05); while no difference is assumed between the grades during the peer tutoring intervention (z = -0.74,p> 0.05).Shows the ranges obtained by the group that received peer tutoring when comparing the baseline measurement with the first and second measurements during the intervention..000.000.941 Signals range test when comparing the measurement before and after the intervention in the experimental group.

Discussion
Tutoring nowadays has become a support strategy for the accompaniment of students through a personalized relationship that guides on academic, personal and / or professional aspects (Alañón, 2000, Almajano, 2002, Coriat & Sanz, 2005) by a tutor.The role of facilitator, counselor and guide can be assumed by the expert teacher or by an agent that covers the paper, in this case by another student (Pedicchio & Fontana, 2003, Arbizu, Lobato & Del Castillo, 2005).The process of accompaniment between two students seems to be an effective strategy for school achievement, as indicated by Baudrit (2000); Duran & Vidal (2004) and Duran & Monereo (2008).The conditions required for the proper functioning of peer tutoring is basically that there is a difference in attitudes and abilities between the two (Arco and Fernández, 2011); Chullén-Galbiati, 2013); because one is the one that guides the other regarding the knowledge in which the tutee has difficulty.In this way, in the first instance, the tutee perceives an unfavorable situation such as the failure of the English learning unit as a great challenge difficult to overcome; especially for those who lack basic skills for learning a second language.It is evident that learning a second language is a complex process.Ordorica (2011) suggests that expectations regarding bilingualism combined with attitudes towards the second language as well as the cultural aspect are the basis for the attitude of learning a second language; without leaving aside variables of intelligence, aptitude and motivation (Dickinson, 1987, Arnes, 1990, Krashen, 2003).The peer tutoring strategy implemented in the present research identifies attitudinal aspects from the beginning in both the tutor and the tutee.In one instance, in the tutee a perspective of solution towards the problem of reprobation is generated, establishing an extrinsic motivation to obtain a personal benefit of approvement with support of the school's own resources; without having greater demand of investment of time or cognitive effort; and on the other hand, for the tutor, positive elements are attributed from the beginning about their performance, encouraging their academic self-concept as good students; and this in turn results in greater commitment and motivation for the benefit of his partner (Robinson, Schofield andSteers-Wentzell, 2005, Duran, 2009).
The procedure implemented by the tutors in the present investigation was effective given the results in the grades obtained by the students with underperforming English learning in the first partial and in the final exam, compared with the students who did not receive the support of the peer tutoring.The structuring of the class session was a key element for the proper functioning of peer tutoring since the tutors had to prepare in advance the topics to be seen; they needed to take notes and compare them with those of the tutees and provide feedback on the doubts they had.This planning made possible to assume a listening position both for the tutor and for the tutorship promoted in an atmosphere of solidarity and openness (Rudhland & Rennie, 2014); Above all, this environment promotes closer interactions given that they are students of the same age and in a similar context (Jiménez, 2015).Maynard & Almarzougi (2006) suggest that the construction of trust environments between students encourages feelings of security through spontaneous expression through questions, discussions and the resolution of doubts.
On the other hand, as Falchicov points out ( 2001), the peer feedback strategy fosters cooperative learning through nonevaluative aid interactions and, together with the monitoring carried out during the classes, suggests that the time dedicated by the tutors to the tutees has a positive effect for learning (Fraser, Beaman, Diener & Kelem, 1977).In this way, the model of peer tutoring could be outlined as an option that respects learning styles and preferences and takes advantage of psychological factors to develop skills for learning the English language; and it also potentiates the use of time in learning environments, given the scarcity of this on the part of the teacher (Viáfara & Ariza, 2008).
Although the results found reflect a significant effect in the strategy of peer tutoring, it is necessary to have greater rigor and control of variables, such as the timely training of the tutors prior to the implementation of the peer tutoring model; besides focusing and detailing the particular tasks of both agents in said model; taking care to identify the profiles of the tutors in order to make the tutorial model more efficient.
On the other hand, it is important to monitor the academic performance of high performance students throughout the intervention process in order to improve future choices of tutors.In this way, under the specifications and delimitations identified, it is concluded that the peer tutoring strategy in high school students has a significant effect on the improvement of academic achievement in the English learning unit.

Table 1 .
Kolmogorov-Smirnov Goodness-of-Fit TestShows a summary of the null hypothesis test to observe normality in the measurements.
213Shows the average ranges of the control group and experimental group in the initial grade.

Table 3 .
Comparative analysis through the Mann-Whitney U test in both groups

Table 4 .
First partial grade comparison of groups during the interventionShows the average ranges for the control group and experimental group obtained in the first partial grade.

Table 5 .
First intervention comparative analysis through the Mann-Whitney U test

Table 6 .
Comparison of groups in a second measurement during the intervention 213Shows the average ranges for the control group and experimental group in the final grade.Table 7. Final grade comparative analysis through the Mann-Whitney U test

Table 8 .
Ranges obtained by the experimental group when comparing the evolution before and after the intervention

Table 9 .
Comparative analysis through the Wilcoxon sign test