High Rates of Dropping Out of School: The Need for Alternative Solutions

This report aims to summarise the findings by Dominique Goux, Marc Gurgand and Eric Maurin on high school dropouts. Other literature concentrates on altering students’ academic performance to prevent dropouts; however, this study looks at the idea of changing expectations towards future academic choices.

High school dropouts are a significant issue within our society today. The world has become increasingly competitive; therefore it is essential to have a strong educational background in order to ensure employment. Dominique Goux, Marc Gurgand, and Eric Maurin discuss how formulating low achieving student's educational expectations lead to a lower dropout rate and grade repetition. They changed student and parents expectations with a non-intensive experiment in France which was facilitated by the principals of 37 schools.

One of the leading causes of dropout behaviour is the failure to develop realistic educational objectives. Economic literature views school as an investment, if the immediate costs are greater than the expected gains, this leads to a higher amount of adolescents dropping out. The article is important as it highlights how non-expensive intervention facilitated by the principal can result in declining high school dropouts. The parent-teacher interaction is important and the researchers have highlighted this as a way to change the aspirations of low achieving adolescents. The article is also important as it shows if a student changes educational expectations it could possibly prevent a student from dropping out in the future.

In many different countries after 9th-grade, students have to apply to a new type of school. These usually include prestigious academic schools and vocational schools. At the end of middle school in France, students can either enter a three-year highly selective academic programme or enter a vocational school that offers both two and three-year programs. An assignment system in France allocates students to their preferred choice using average marks obtained in key subjects. Both three-year programs give access to higher education for students. Therefore, it is more prestigious than the two-year program. Also, students could drop out of the educational system entirely and take up an apprenticeship.

Generally to reduce dropout rates schools may impose strict activities on students in order to raise academic performance however this is not the case with this study, the researchers conducted a large scale randomised experiment that targeted low achieving students. School principals facilitated the experiment, it allowed students to apply to different tracks (options after 9th grade) based on their academic performance. There was two meetings between parents and teachers, the results show a reduction in grade repetition and dropout rates. The results also show a higher amount of adolescents applied for two-year vocational programs and apprenticeships. The results mean that it may not be useful to try to change a student’s academic record; instead, schools should focus on changing student and parents expectations regarding future choices while considering their academic ability. This is shown as a higher amount of students applied to the two-year vocational course; thus the principal has altered their expectations and preferences based on the academic ability of students.

Other researchers have concentrated more on high ability students who undervalue their potential, whereas this article concentrates on low ability students who overvalue their potential, therefore, bringing new evidence on how altering expectations can result in lower dropout rates. Furthermore, many other dropout prevention policies include financial incentives and academic support therefore the experiment carried out contributes to the literature on school dropout rates. The article is supported by other research such as by Stinebrickner and Stinebrickner (2009) who also found unrealistic expectations regarding university entry led to higher dropout rates in USA colleges. Furthermore, other literature supports the idea that unrealistic expectations lead to poor outcomes for students such as research by Gottfredson 1981.

Methodology

The experiment took place in Paris in the Local Education authority which includes 1.1 Million students. 37 Schools took part which represents 9% of the total schools within the district. Low-income areas within Paris are represented heavily within the sample. Early within the academic year the principles randomly selected 25% of students who are most exposed to the risk of becoming a school dropout. The principals-based their selection on the academic ability of students with the ones selected having the lowest grades. Over half of the selected students had already repeated a school year.

Preselected students were randomised into treatment and control classes by Goux et al. The researchers split half of the classes into the treatment group, and the other half was put into a control group. The randomisation was stratified by the number of pre-selected students, the number of girls and the number of students who had previously repeated a year. In total, they were 97 treatment and 82 control classes. They were no important differences between treated and control classes in terms of characteristics. Only the parents of the 97 treatment classes were invited to attend the meetings. The results are based on a comparison between the treatment and control groups of the pre-selected students.

After each group was split the experimenter invited preselected students’ parents within the treatment group to have two meetings with the principal regarding their children’s future choice of track (options after 9th grade). The reason for the meetings is because it allows the parents to become involved with their children’s future. The meetings took place in January and April before 6 am during the school. LEA experts helped prepare the principals with guidelines for the meetings. These include the academic future for the parent’s children, the procedures of applying for the next track education and how the parents can help. The guidelines included identifying family aspirations; however, only concerning the academic performance of the child. The principles helped adjust expectations based on the academic performance of the adolescents. For example, a principal may suggest the idea of an apprenticeship for a student who is at risk of dropping out of school or repeating a year. The programme allows the parents to understand that two-year vocational qualifications can be a good option and that it may be more suited for their kids who are continually having to repeat the 9th grade.

The principals only invited parents of the preselected children within the treatment group, not the control group. This is shown through the attendance sheet. It shows that the take-up is only large for pre-selected students within treatment groups. Goux et al. found 52% of Parents attended one of the meetings and 21% attended both. Only a tiny portion of families who were not invited turned up. This shows the principal made an effort to call and invite the parents of the preselected students within the treatment group.

Data

The researchers collected data taken from schools during the 2010-11 academic year. The data includes the student’s grades in key subjects such as Mathematics and English. This is very important because those two subjects are key indicators of whether a student will receive his preferred choice of school/apprenticeship after the 9th grade.

The researchers also collected the attendance and the mark taken at the end of year national exam for everyone in the 9th grade. The exam is not compulsory and does not determine if the students will achieve one of their preferred school choices after the 9th grade. The exam is held on the last day of the academic year, and some students are absent on this day even though on this day most students may know if they got accepted onto their preferred track choice. The researchers found absence on this day was higher amongst dropouts.

The next part of data is a questionnaire that parents are asked to complete. The questionnaire included questions about how the parents are involved in their children's future and how they are able to get in touch with school teachers. One of the questions also included the educational expectations of their children etc. In order to increase the response rate, the researchers called non-responders of pre-selected pupils and asked them the same set of questions.

Also, the writers took administrative exhaustive data which gives information about the application and the assignment process for each pupil. The data includes applications for all of the four preferred choices at the end of the year, and the researchers know the rank of each choice and type of school/programme. Also, they know whether a student appealed a decision or decided to retake the 9th grade. The data also includes student's whereabouts one year after the treatment (meetings) year and even two years after. This uses each student's national identification number. The researchers can then tell who is still in school, apprenticeship and who has dropped out of the system altogether.

Results

Y_(is = αT_i + X_is β+ U_S +V_s )

The writers analysed the effect of meetings using the following model above. They focused on preselected students. T_i Is the dummy variable (0 or 1) indicating if a student is in a treatment class or isn’t. X_is Is a vector of control variables which include dummies for a variety of different characteristics of the students. The other two variables are fixed effects of school and the potential error term.

Parents Altered Preferences

At the end of the treatment, the parents became heavily involved with their children’s academic careers. Academic aspirations changed tailored to what the adolescent’s academic record was. This is shown on the options that the adolescents took at the end of the third term.

The results of the experiment suggest that parents adapted their expectations in relation to their children’s academic performance. The researchers found they were an 8% reduction in the number of parents expecting their child will complete a three year academic/vocational program leading to higher education (comparing treatment and control classes). The researchers also found a 3.4 percentage point increase in parents expecting a two-year vocational course. This highlights that the results show a change in academic expectations through meetings with the principal. The article results also suggest very few parents expect their children to drop out of school regardless if they were in the control or treatment group, which agrees with the idea that parents expectations of their child dropping out is extremely rare. This is because parents understand the problems in finding employment if their child was to drop out of high school.

End of Exam Examination Presence

End of Middle school examination C T-C

Fail but present on exam day 44.8 +6.7**

Not present on examination day 10.6 -5.4**

The researchers collected information on the number of students who turned up and took the end of middle school national examination. The exam is not compulsory, and results are not taken into account. Most students know what academic track they are on for the next two/three years before this examination. This allows the schools to find out what academic track each pupil is on, on this examination day. This means that students who do not turn up to the examination are likely to be potential dropouts. Table 1 has two columns, C which is the value of the preselected control group and T-C which is the treatment dummy value. The results of the article show that preselected students in the control group have a higher absent attendance (10.6%) compared to the treated group (5.2% as 10.6 – 5.4 = 5.2). They were no increases in academic performance by people who attended compared to those who did not. This shows that students who did turn up wanted to know about what academic route they are on in the future meaning they are less likely to drop out of education. Furthermore, it shows the experiment had an effect on possibly preventing high school dropouts for the preselected students in the treatment group.

End of Ninth Grade Options

Applications C T-C

At least One two year vocational program 15.8 +4.9

Two year vocational – first choice 11 +3.8

Two years vocational – not first choice 4.8 +1.1

It was found that the experiment had weak effects on the academic records of pupils. The actual aim of the treatment was not to change academic records but to change expectations of pupils which adhere to their academic record. They were no increase in preselected students applying for a three-year program however this may be expected as the expectations of the children may have been changed through these meetings. Table 2 has two columns, C which refers to the average value for preselected students in the control group and T-C which is the treatment dummy value. There is a 4.9 percentage point increase in the number of preselected students who included a two-year vocational course in their applications. They were also a decline in students who applied for three-year vocational courses and students who repeated a year. Two-year courses are easier to be accepted on; therefore this shows the meetings facilitated by school principals may have allowed students to alter expectations and apply for two-year courses as a potential back up. The track choice results show students are realistic about their academic future and their chances of being accepted into a three-year program. The results show that the experiment made students more aware of other choices rather than either applying to three-year programs or just resitting, as fewer students resat and applied to three-year programs.

Student progress after 1st and 2nd year

The number of students who retook grade 9 decreased. The intervention resulted in a higher number of students who received apprenticeships one year later along with less grade repetition. This is a result of the principal changing unrealistic expectations of the students and the parents, by clarifying to them that obtaining a three-year study course is more difficult than they think. For students with poor academic records, the principal convinced parents that an apprenticeship might be their best option in order to get employment rather than a three-year course. The researchers found that the principals managed to induce potential dropouts to consider an apprenticeship and potential repeaters to consider a two-year course. The actual experiment did not induce potential students who wanted to stay in the school to drop out, instead just changed their line of thinking to a more suited two-year course.

The major effect of the study within the article is to prevent people from repeating the ninth grade or dropping out. The program resulted in a higher number of students obtaining an apprenticeship within the treatment group however two years later some students could potentially be disappointed with their choice; thus they may have been better off repeating the ninth grade. The school principal may have convinced students of realistic options only in the short run. In order to test this hypothesis, the writers compared grade advancements between treated and controlled students two years after the intervention. If the experiment just delayed grade repetitions and dropouts, then the results must show weaker differences in dropouts and grades. Goux et al found that two years after the experiment the number of students who completed and moved onto their second year of high school is almost the same as the initial gap for access to high school programs between treatment and the control groups. This shows that students were happy with their decision to take part in two-year vocational courses rather than apply for a selective three-year course.

The results show that the difference in dropout rates between the treatment and control group is larger after two years. The experiment resulted in fewer dropouts after the treatment year and fewer dropouts for those who resat the 9th grade (-1.5 percentage points). Overall the experiment reduced dropout rates through two different ways. The first was to help potential dropouts look at alternatives such as apprenticeships after the treatment year. This increased the gap between dropout rates between the treatment and control group and has continued to persist over time. The second way the experiment reduced dropouts is by helping underachieving students look at alternatives such as the two-year vocational programs. One year later the majority of treated students completed their first year of the high school program.

Conclusion

Across the world, students have to pick different educational routes at the end of the 9th grade. The only problem with this is low ability students often aim too high and have unrealistic expectations. Students in France have a negative perception of two-year vocational programs and generally, prefer to go for three-year paths even though they are not academically ready for it. The researchers showed that a simple experiment which is facilitated by school principals could help low achieving students to form realistic expectations based on their academic record. The meetings led to a decrease in the dropout rate by 25%.

The writers show that by adjusting preferences over the academic future of the students, this allowed students to apply and enter better academic tracks that are suited for their abilities. This means that if the program is carried out on a wide scale, it has the potential to reduce dropout rates seriously and to increase future employment. Overall the experiment shows it is possible to influence student choices in a way that will improve their lifelong chances at a very small cost. Preferences and expectations are changeable, and principals are able to alter them for less able students.

Goux et al. also showed that non-intensive meetings between parents, students, and principals could prevent future dropouts from occurring in high school. Furthermore, the paper shows that educational outcomes of low ability students can be improved with a similar approach to other pieces of literature which considered low aspirations of high performing students. To conclude the paper highlights the alternative of changing student expectations rather than academic performance to reduce dropout rates.