NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
How Important are Firm Visits for High School Students? Evidence from a Randomized-Control Trial
Initial registration date
December 02, 2016
March 31, 2017 12:28 PM EDT
University of Mannheim
Other Primary Investigator(s)
Additional Trial Information
Firm visit experiences are de-facto established as an important screening tool in the labor market, signaling to the employer the quality and experience of a candidate. Compulsory firm visits (mostly as internships) become an essential part of many school and university curricula, withdrawing the students from education for weeks and months. It is expected that firm visits can support the occupational choice and establish important networks to the employer. At the same time, the labor market literature on the effects of firm visits is almost non-existent. If firm visits have no beneficial impacts on labor market outcomes, and what we so-far observe is driven by self-selection of very motivated students, then keeping students out of school/ university could even have negative impacts on labor market outcomes. In collaboration with UNICEF Montenegro and Giant's Shoulder, the Ministry of Education and the Ministry of Labor we evaluate a program that investigates the effects of a unique labor market pilot initiative called ``Prilike''. Within the scope of the ``Prilike'' project, UNICEF and Giant's Shoulder have established a cooperation with over 50 companies, amongst others with Microsoft, Siemens, Inditex, Ernst \& Young, and KPMG. To evaluate whether the experience of going to these companies changes expectations about the labor market hiring process and expected wages, as well as ambitions to work in the private sector, we set-up a randomized-control trial. From the pool of all schools in Montenegro, we randomly select eight schools to the program. Following, using then individual-level administrative data and a survey responses for students from the 11th and 12th grade, we randomly match a randomly selected group of students to the companies. Overall, we find that the program has made the students' labor market expectations more realistic, optimistic and changed the students' locus of control.
Avdeenko, Alexandra. 2017. "How Important are Firm Visits for High School Students? Evidence from a Randomized-Control Trial." AEA RCT Registry. March 31.
One-day internships to the companies took place between November 16, 2016 and December 2, 2016. In total around 60 companies participated in the initiative, all from different industries and across the country. Adolescents were assigned to at least one 1-day-internship at a company.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
updated beliefs about job market and better educational outcomes; lower likelihood to believe that informal networks are key to find a job; spill-over effects between students
Primary Outcomes (explanation)
Respondent wants to work in the public sector
In your first job, how many EURO per month do you expect to earn?
And when you are 40 years old, how many EURO per month do you expect to earn?
Doesn't know likelihood to find the ideal job at some point in the future.
Doesn't know likelihood to obtain desired education degree.
Expected working tasks: Interesting
Job finding strategy is: Asking parents to contact the employer
Job finding strategy is: Getting good grades in school
Job finding strategy is: Doing an internship
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
The randomization was performed in two steps: (1) we selected the schools and (2) we selected the students. On October 6th 2016 the Ministry of Education shared administrative records of all students in secondary education. Using administrative information to balance characteristics we randomly selected 8 schools out of 50 to be part of the initiative. Throughout the county we implement an online survey to acquire additional background information. Over 40% of students in grades 3 and 4 of secondary schools participated. For further analysis we use a total of 4,003 student surveys from treatment and control schools, Thereof 1,437 were from the 8 selected schools and 2,566 from other schools.
In the 8 schools we randomly selected we select 359 students who were eligible to attend a company and 360 students who were eligible to attend a company three times. 719 students remain in the control group, whereby 359 have already been assigned to a further future intervention.
Experimental Design Details
We used the following stratification variables: region (North, South, Center) and by regional unemployment rates (1 if above median) and randomly assigned treatment and control schools within the strata (generating a random number variable). The selected eight schools teach 2,720 11th[3rd] and 12th[4th] grade students (21.20 percent of the total country population). For the selection of students, we used Stata's build-in randomize command to select four groups: 3 treatment arms and 1 control group. We specified a list of variables for which we tested the balance. Moreover, we
specified that the randomizations were run at least 500 times or until the minimum balance p-value allowable to accept a given randomization exceeded 0.1.
(1) random selection of schools; (2) individual randomization
Was the treatment clustered?
Sample size: planned number of clusters
Sample size: planned number of observations
Sample size (or number of clusters) by treatment arms
8 schools assigned to program vs. 42 control schools.
Within the 8 schools: T1 a student is assigned to 1 one-day-internship (N=359)
T2 a student is assigned to 3 one-day-internship (N=360)
T3 a student will be assigned to a further future intervention (N=359)
359 assigned to control group
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
INSTITUTIONAL REVIEW BOARDS (IRBs)
Post Trial Information
Is the intervention completed?
Is data collection complete?