Can academic mentoring reduce the time to get a degree? - PDF

Please download to get full document.

View again

of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Art

Published:

Views: 8 | Pages: 8

Extension: PDF | Download: 0

Share
Related documents
Description
Can academic mentoring reduce the time to get a degree? Claudia Marin SFPC University of Bari Anna Rinaldi DEM University of Bari We run a
Transcript
Can academic mentoring reduce the time to get a degree? Claudia Marin SFPC University of Bari Anna Rinaldi DEM University of Bari We run a randomized control trial to assess the impact of a mentoring program aimed at reducing the time needed to get a degree for off-course students i.e. students who have not succeeded in passing all the exams within the prescribed period of time. Literature on graduation rate generally states that students outcomes are affected by financial barriers and liquidity constraints. We find that this mentoring program had no positive effects on students outcomes and that their performances are not related to family income. Keywords: randomized control trial, mentoring JEL Classification: C90, I20. Introduction. The Italian university system In Italy, up to 999 degree programs were organized over four years and conferred their graduates the title of doctor (the same held by Ph.D. in other university systems). Considering this and other significant differences among institutions of European higher education, the Ministers of 29 2 European countries started the Bologna Declaration, in June 999. This aims at creating a homogeneous and transparent academic system across Europe, facilitating the comparison and mobility of European students, creating the European area of higher education. In Italy, the Ministerial Decree 509/999 introduced the 3+2 (-year) system: it consists of a first level course for the completion of which 80 credits are required- plus a second level course for the achievement of which additional 20 credits are needed. In 2004, these two cycles, 3 years and 2 years, were made independent. For this reason there are different kinds of students and graduates that co-exist in the Italian university system: i) 4 years; ii) 5 years; iii) 3+2 years. Except for Medicine (6 years) and Architecture (5 years) 2 ow there are 47 member countries .2. The loss of Universities Italy ranks 34th out of 36 in the OECD list, with only the 9% graduates in the year-old age bracket. Italy is under the European average (34.6%) and the level of many European countries (30.7% in Germany; 40.6% in Spain; 43.4% in France; 45.8% in U.K.). At the E.U. level, the goal for 2020 is 40%, while Italy aims at a more modest 26-27%. This is also due to a peculiar phenomenon: in the Italian University system, 33.6% of students does not pass all the exams within the prescribed period of time. evertheless, these kinds of students have the possibility to enroll in supplementary years. Among those, 7.3% is totally inactive and does not attend any course nor pass any exams (without attending classes, which is also possible) (ational University Council, 203). In other words, in Italy it is potentially possible for a student to remain enrolled in college all life long, without taking any courses nor exams. This is a big loss not only from the student but also from the university overall standpoint, because the so-called Fondo di Finanziamento Ordinario (FFO), the main mechanism by the means of which the Ministry finances each athenaeum, greatly depending on the numbers of off-course students. For example, in 202, the calculation of the aforesaid ministerial loan depended, among others, on the ratio between credits acquired by students and credits expected, divided by degree courses, according to the following formula: A2=CfuPesatiTot/ CfuPesatiTot (of all Athaenei taking part to the distribution) CfuPesatiTot =CfuPesatiA+ CfuPesatiB+ CfuPesatiC+ CfuPesatiD CfuPesatiX=[(Cfu_effX/Cfu_teoX)/median]*Cfu_effX, X=A D where: Cfu_effA, Cfu_effB, Cfu_effC, Cfu_effD are European Credit Transfer System Credits (ECTS) actually acquired in the year 20, divided by cohorts Cfu_teoA, Cfu_teoB, Cfu_teoC, Cfu_teoD are European Credit Transfer System Credits (ECTS) potentially acquirable in the year 20, divided by cohorts. For this reason, the presence of inactive students most of whom are offcourse students- is very detrimental also for universities budget. 2. Formal mentoring In literature several barriers reducing the graduation rate have been identified. Many authors emphasized on financial barriers and liquidity constraints (Dynarski and Deming, 200; Belley and Lochner, 2008), while other scholars underline the importance of students incentives (Angrist et al., 2006). In recent years, more and more universities and education institutions all over the world have carried on interventions in which mentoring, which is normally supplied by third party providers, is considered as an instrument to improve college outcomes and is normally supplied by third party providers. According to Eby et al. (2008), formal mentoring can be defined as a supportive relationship between a more experienced person (the mentor) and a less experienced partner (the mentee), that enables the development of a trust-based relationship allowing for the needs of the mentee to be met. More specifically, one can refer to academic mentoring when the support provided involves academic or vocational assistance from a teacher or another education-related individual (Jacobi, 99). Tutoring and coaching has been the object of some randomized experiments. In the randomized trial aimed at improving academic performances described by Angrist et al. (2009), three different treatment groups were offered, respectively, academic support services, financial incentive for good grades, and both interventions. Their results show that the use of the mentoring service was the highest for women and for subjects in the combined group and that combined treatment also raised the grades and improved the academic standing of women. Bettinger and Baker (20) find that college completion and college success lag behind college attendance, which is improved by coaching that is a particular form of mentoring. In this experiment, students randomly assigned to a coach were more likely to persist during the treatment period, and were more likely to be attending the university one year after the coaching had ended. They also show that coaching is a more cost-effective method of achieving retention and completion gains where compared to previous interventions about financial aid. 3. Assessing the effects of formal mentoring of off-course students in the Athenaeum of Bari. 3.. The Athenaeum of Bari In the academic year 200/, the number of enrolments of the University of Bari corresponds to,632, with a number of missed enrolments equal to 3,78, highlighting a dropout rate equal to 35.2% between the first and the second year. The inactivity rate is, however, 3.%. Enrolments are diminishing at University of Bari and data go from 59,45 students enrolled in the academic year 2009/0 to 50,335 in 202/3. Off-course students that get their degree are more numerous than students who graduate over the prescribed time: in 202, out of 7,826 students getting a degree, 4,386 were off-course. 3.2 The experiment Mentoring in higher education has a long tradition in many nations, while in Italy is a relatively new practice. To fight the plague of the off-course students, the Senate of the Athenaeum of Bari appointed a committee made up professors, one from each faculty (faculty manager for off-course students). evertheless, the Senate did not impose a standard protocol to achieve the goal of the reduction of this category of students and left each faculty manager free to decide the way to do it. We found that some faculty managers decided not to approach the problem, because the program would bring no direct returns and had no budget. Faculty managers that decided to approach the problem did not agree on a common method and designed many different programs. Still, we can infer some common patterns within faculties: - for each degree course a list of the most difficult exams, that students were not able to easily pass, was created; - in some degree courses with a small number of students (and off-course students), ad hoc recovery courses, for each subjects on the list, were created; - in more populated degree courses, with many supplementary year student, the faculty manager preferred to appoint one or two mentors for each subject of the list. We have been able to run a randomized evaluation program in the Faculty of Law of the Athenaeum of Bari, which belongs to the second aforementioned category 3. The program in the Faculty of Law started in It targeted older students, belonging to the 4-year degree course. There were 2490 and were divided into subsets by number of missing exams. Because of budget constraints, the faculty manager decided to treat only students with up to six missing exams and to encourage students with more than six missing exams to move to the shorter 3-year degree course. The program lasted for one and half academic years and involved only six missing exam students. End-line data were collected for the year 20. Students missing six exams were randomly divided by a simple lottery in control group and treatment group. The control group was made up of 7 students and the treatment group, 69. Table Distribution of students divided by sex. Groups Sex Total Male Female Control group Treatment group Total Treatment group students were contacted one by one via or telephone by the faculty manager staff. The first contact was aimed at explaining the program (that they had the possibility to be tutored) and at obtaining the participation of students in the program, establishing, where appropriate, by which exam and mentor to start. It was up to the student to make an appointment with the mentor indicated. The role of the mentor generally was to provide clarification on the syllabus and to answer questions about the contents of the program. The results, unfortunately, turned out to be quite disappointing. In fact, one can observe a paradoxical effect: students belonging to the control group took more exams than treated students (table 2), while the number of graduations (table 3) was exactly the same (only four in both groups!). 3 We also collected data and evaluated ex post the programs that took place at the Faculty of Economics, that will be the next stage of this research. Table 2 umber of taken exams, control group and treatment group (differences ) DIFF GRP Students GRP Students 0 C 33 T 34 C 7 T 5 2 C 6 T 6 3 C 5 T 5 4 C 6 T 5 5 C 4 T 3 Tot C 7 Tot T 68 Table 3 umber of graduated students, control group and treatment group (differences ) GRP grad T C This clearly means that the mentoring program adopted by the Faculty of Law was not effective at all. Interviewed students explained that they did not find the program useful for two main reasons: mentors were not motivated and were not able to really train mentees and meetings between mentors and mentees (generally one or two) were too scanty to affect students knowledge. This also explains why in less populated degree courses, mentoring programs seem prima facie to have been more successful. We wanted to be sure that those results were not related to other factors normally mentioned in literature, such as liquidity barriers or financial constraints. To verify this, we used the family income of each student involved in our experiment. We also tried to understand if other variables -such as year of enrolment, year of birth, sex or high school graduation score- could affect outcomes in tables 2 and 3. In the following regression, the dependent variable is given by the exams passed in the three-year period. The independent variables were sex, grade of maturity, income, year of birth, year of enrolments and groups. The sample consisted of 39 students. Figure Model Summary Model Model Summary Std. Error Adjusted R of the R R Square Sq uare Estimate,256 a,066,020,54 a. Predictors: (Constant), YEAR OF EROLMET, ICOME, YEAR OF BIRTH, SEX, GROUPS, HS GRADUATIO SCORE Data in Figure indicate that the model has not a good fit to the data, since the value of R Square is,066 and that of Adjusted R Square,020: in Model Regression Residual Total practice, the regression model explains almost 7% of variability of the dependent variable. The small difference between the two values, however, is a further finding of no redundancy (collinearity) in the predictors. There are no clear guidelines for evaluate these indexes, because R Square depends on the number of subjects and the number of predictors. The following figure, AOVA, shows the significance test that checks the null hypothesis that R Square is different from zero. Figure 2 Anova AOVA b Sum of Mean Sq uares df Sq uare F Sig. 20, ,422,44,204 a 292, ,375 32, a. Predictors: (Constant), YEAR OF EROLMET, ICOME, YEAR OF BIRTH, SEX, GROUPS, HS GRADUATIO SCORE b. Dependent Variable: EXAMS I THE 3-YEAR PERIOD Model Table 4 Regression coefficients. (Constant) GROUPS ICOME YEAR OF BIRTH SEX HS GRADUATIO SCORE YEAR OF EROLMET a. Dependent Variable: EXAMS I THE 3-YEAR PERIOD Coefficients a Standardi zed Unstandardized Coefficie Coefficients nts Collinearity Statistics B Std. Error Beta t Sig. Tolerance VIF -22,338 53,26 -,420,675,75,35,056,554,580,736,360 2,58E-06,000,038,426,670,970,03,20E-02,027,045,445,657,755,325 -,696,299 -,207-2,328,022,958,044,22E-02,0,9,47,253,70,427 9,33E-05,000,095,072,286,975,026 Table 4. presents the output (table of regression coefficients) in which they were applied indices of collinearity. By the indices of tolerance and VIF we can observe that none of the variables considered in our test presents problems of collinearity: in fact more than 60% of the variance of each variable is not in common with the other independent variables in the analysis. Table 5 Collinearity diagnostics. Collinearity Diagnostics a Model Dimension Variance Proportions HS Condition YEAR OF GRADUATIO YEAR OF Eigenvalue Index (Constant) GROUPS ICOME BIRTH SEX SCORE EROLMET 5,708,000,00,0,0,00,00,00,0,549 3,223,00,54,08,00,00,00,03,386 3,846,00,02,7,00,00,00,23,266 4,630,00,07,3,00,02,02,73 5,944E-02 9,800,00,09,0,00,68,33,00 3,049E-02 3,682,00,2,07,00,28,54,00 3,23E ,926,00,06,00,00,02,09,00 a. Dependent Variable: EXAMS I THE 3-YEAR PERIOD In the table of Collinearity Diagnostics (Table 5) additional indexes of collinearity are shown. In particular, the eigenvalues are obtained by performing the principal component analysis of the matrix of scalar products between the independent variables, and explain the correlation between the independent variables. If many eigenvalues are close to 0 the variables are strongly correlated. The index of collinearity (condition index) derived from the eigenvalues: if it is between 5 and 30 indicates possible collinearity issues, if it is greater than 30 collinearity is severe (this index, however, is less important than the VIF and tolerance, Pedhazur, 997). The last index is the proportion of variability explained by the principal components ( Dimensions ) associated with each eigenvalue. Collinearity is a problem if a component ( Dimension ) with a high degree of collinearity contributes significantly to the variance of two or more variables. In Table 5 none of the components with collinearity index greater than 5 contributes substantially to more than two variables; there are therefore not present problems of collinearity. Table 6 Correlations Correlations GROUPS ICOME YEAR OF BIRTH SEX HS GRADUATIO SCORE SCORES(ITERVAL) YEAR OF EROLMET EXAMS I THE 3-YEAR PERIOD **. Correlation is significant at the 0.0 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). YEAR OF HS GRADUATIO YEAR OF EXAMS I THE 3-YEAR GROUPS ICOME BIRTH SEX SCORE SCORES(ITERVAL) EROLMET PERIOD,000 -,099 -,369** -,05 -,463** -,444** -,085 -,035,,26,000,859,000,000,320,683 -,099,000,027 -,26 -,08,002,022,059,26,,765,55,842,985,804, ,369**,027,000,80*,402**,383**,089,083,000,765,,034,000,000,297,33 -,05 -,26,80*,000,073,036,060 -,34,859,55,034,,39,673,483,6 -,463** -,08,402**,073,000,806**,43,8,000,842,000,39,,000,094,65 -,444**,002,383**,036,806**,000,75*,42,000,985,000,673,000,,040,095 -,085,022,089,060,43,75*,000,0,320,804,297,483,094,040,,235 -,035,059,083 -,34,8,42,0,000,683,502,33,6,65,095,235, The correlation table, unfortunately, do not show high correlations, no value is close to and many data are negative. Conclusions In this paper, we showed the results of a randomized control trial on some off-course students at the University of Bari. We found that the mentoring program that should have had to reduce the time for off-course students to get a degree, did not in fact give any good results. We also observed a paradoxical effect regarding the number of exams passed. When we tried to understand whether these outcomes could be explained by other factors, first of all income or age, we found that students (un)success was not related to income, as the literature on graduation rate usually states (financial barriers). References Angrist, J., Bettinger, E., and Kremer, M. (2006) Long-Term Educational Consequences of Secondary School Vouchers: Evidence from Administrative Records in Colombia, American Economic Review, 96(3): Angrist, J., Lang, D., Oreopolus, P. (2009) Incentives and services for college achievement: Evidence from a randomized trial. American Economic Journal: Applied Economics, I(), Belley, P., and Lochner, L., The Changing Role of Family Income and Ability in Determining Educational Achievement, Journal of Human Capital, (), Winter 2007: Bettinger, E., Backer, R. (20) The effect of student coaching in college: an evaluation of a randomized experiment in student mentoring, BER Working Paper 688. Eby, T., Allen, T.D., Evans, S.C., g, T. Dubois, D. (2008), Does Mentoring Matter? A Multidisciplinary Meta-Analysis Comparing Mentored and on-mentored Individuals Journal of Vocational Behaviour 72(2): Pedhazur E. J. (997), Multiple regression in behavioral research, Wadsworth Publishing.
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks