“Mentorship, Entrepreneurship, and Labor Market Opportunity in Saudi Arabia.”

2019-04-01

2020-12-31

Whether students in the Mentoring Network, compared to those who were not selected into the Mentoring Network:
-Are more likely to be employed
-If already were employed are more likely to remain employed
-If had entrepreneurial aspirations are more likely to maintain those aspirations or start their own business
-If job seeker already owned their own business, if they are more likely to still have their business open after the period of the intervention.

Whether high-school students treated, compared to those who were not treated:
- Are more likely to graduate high-school
- Are more likely to pursue further studies
- Are more likely to express entrepreneurial aspirations
- Are more likely to express interest in being employed
- Are more likely to report good relationships with their male relatives

I expect mentees whose families attended the Mentoring Network event to have greater effects of the mentoring network (through a channel of family/household support of the mentoring relationship)

Students and job-seekers will be recruited to a Mentoring Network through a baseline survey which has an information intervention to incentivize recruitment to the Network. Students who opt into the Network will be randomly selected into the Network and matched with a Mentor. Employment, entrepreneurship, and well-Being outcomes will be assessed through intermittent surveys in order to evaluate the impact of membership into a Mentoring Network on employment, entrepreneurship, and well-being outcomes.

High-school students were recruited through a baseline survey.

In partnership with our policy partner we encouraged randomized priority invitation to their Mentoring Program so that we could have a control group with similar background.

In the midline, high-school students will be randomly treated with an entrepreneurship information intervention. Interest in employment, entrepreneurship, and well-being outcomes will be assessed in the endline survey.

Where possible, aspects of the Mentor will be also randomized such as whether the Mentor is employed, an entrepreneur, had the same college major, etc. This will be done to explore alternative explanations for the impact of Mentorship aside from the Network effect such as role-modeling, sponsorship, and information provision.
Together with our NGO partner, we will randomize an information treatment signalling female entrepreneurship to see if this intervention leads students to immediate signing up for an entrepreneurship workshop and later affects their likelihood of graduating high-school and accessing the labor market.

Randomization will be done in office by a computer.

Randomization into the network and then assignment to a mentor will occur at the individual level.

Randomization of treatment for high-school girls will occur at the individual level.

No

We plan to recruit 1000 students or job seekers (500 male, 500 female) and 200 Mentors.

We recruited 200 high-school girls.

We hope to have 1000 students and 200 Mentors filling out baseline surveys, weekly meeting data for one year (52 short surveys), and outcomes surveys at 3 months, 6 months, 9 months, and 12 months. We have 200 high-school girls surveyed in three waves: baseline, midline and endline.

We hope to have 500 students/job seekers in the Treatment group, 500 in Control; 150 Mentors in Treatment and 50 in Control.

We hope to have around 100 high-school girls in the treatment condition and about 100 in the control condition.

Previous studies on mentoring ranged from 400 to 1150 participants needed to see a treatment effects of mentoring and with my previous experience in email recruitment, I plan to work with university partners to email information intervention surveys electronically to the entire student listserve and existing databases. If a standard 4% response rate is expected from email solicitations, we would need to email at least 29,000 students to obtain the required sample size to see a treatment effect. Luckily the universities I am working with have at least this many students enrolled and I expect to detect treatment effects through our recruitment efforts.