**4. Discussion**

The program's potential effects are an essential part of the experience in Buffalo. Our findings on the FAFSA completion project provide strong evidence that a personalized intervention can increase the likelihood students will complete a financial aid application, particularly in a high-need district like Buffalo. This has significant implications for other large metropolitan centers in the U.S.A. and potentially other countries with higher tuition prices for tertiary education [36]. The project was effective in the Buffalo context, because it was developed and implemented in the context of a larger collective impact strategy designed to leverage the power of the school district, local colleges and universities, and participating NGO's. However, the more important implications for researchers, community educators, and NGO's is the collaboration that led to this intervention. We believe there are several important lessons for other local communities, regardless of state or national context, about effective collaborations among schools, colleges, universities, and local community-based organizations.

#### *4.1. Invest Time to Develop Strong Collaborative Relationships*

Perhaps the most important lesson from this program is that collaboration takes time, and it is easiest to cultivate those relationships when the stakes are lower. The opportunities to establish trust are higher. Researchers from UB were already engaged in work around high school dropout prevention well before we began the pilot for the financial aid application process. At that time, dropout prevention was a priority for the district, and our participation helped develop a level of trust necessary when we spent six months shadowing school counselors. The time we spent shadowing counselors was instructive because we had an opportunity to identify problems of practice, all related to the administration of the college choice process. We saw firsthand how much time and energy counselors spent working with students on college applications, essays, admissions test registrations, college visits, fee waivers, and financial aid. We understood the challenges from their perspective and set out to develop solutions that were responsive to the needs of counselors and adapted to the work they were already doing to solve many of these problems. We focused on financial aid in the pilot study because this was part of the college choice process that made counselors most nervous. School counselors were less knowledgeable about financial aid and apprehensive about managing student and family personal data, including social security numbers and tax documents. No matter what problem we hope to address, in partnership with schools, researchers must spend time understanding the experiences of their collaborators and establishing trust.
