# Hi. I'm **Samantha**.

I write about not getting stuck with your academic assignments.

Trusted Academic ServiceI write about not getting stuck with your academic assignments.

Trusted Academic ServiceMany scholarship search services make exaggerated claims about the number of scholarships listed in their database. This creates confusion concerning the actual number of available awards. This section of FinAid reviews the most reliable data concerning the number and dollar amount of available private sector non-college-controlled scholarships.

A *scholarship* is a form of student financial aid that does not need to be repaid. Selection of scholarship recipients is usually based on a set of criteria, such as academic, athletic, or artistic merit.

A *grant* is a form of student financial aid that does not need to be repaid. Selection of grant recipients is usually based on financial need.

The terms *scholarship* and *grant* are often used synonymously.

A scholarship *sponsor* is an entity that offers a scholarship to qualified candidates. Sponsors can include individual philanthropists, non-profit organizations, for-profit businesses, and private foundations.

A scholarship sponsor may operate several scholarship programs each of which may offer several scholarships. For example, the Ayn Rand Institute sponsors three essay contests, the Anthem Essay Contest, the Fountainhead Essay Contest and the Atlas Shrugged Essay Contest, which offer 229, 244 and 49 awards, respectively. An individual student can apply to only one contest, depending on year in school, and can win at most one scholarship in the contest.

Scholarship search services usually advertise the number and total dollar amount of scholarships listed in their database, as opposed to the number of award programs (distinct addresses). These counts are prone to several sources of error:A better approach would involve multiplying the number of programs by the median number of awards per program. (A third of all scholarship programs have just one award, and three-quarters have four or fewer awards. The median number of awards per program is two.) One could also use the average number of awards per program after removing statistical outliers.

For example, the Ayn Rand Institute's Fountainhead essay contest has a first prize of $10,000, three second prizes of $2,000, five third prizes of $1,000, 35 finalists of $100, and 200 semifinalists of $50, for a total of 244 awards. If a scholarship database multiplies the award count by the maximum award of $10,000, they will report $2.44 million for this program, even though the actual figure is $34,500.

Another problem with the counts published by the various scholarship databases is that they are not independently verifiable.

The most reliable method of comparing scholarship databases is to count the number of distinct addresses listed in each database. This gives an indication as to the number of different scholarship programs listed in the database, and correlates well with the number of good matches a typical student can expect.

Since the scholarship databases are unwilling to release their proprietary lists of sponsor addresses, the only available option is to use statistical sampling techniques to estimate the expected number of matches for each database:

The result is the expected number of valid matches for a typical user.

This result doesn't give any information about the total number of awards or the total dollar amount available, but can be used to derive two quality measures for comparing scholarship databases: precision and recall.

__Precision__ is the percentage of initial matches that turn out to be valid matches. It is a measure of whether the match results will actually be relevant to the student. A database with a low precision score will force users to read through hundreds of irrelevant awards in order to find the awards for which they are qualified. Precision is reported on a scale from 0% to 100%, with 100% being best.

__Recall__ is the percentage of relevant awards that appear in the search results. It is a measure of the coverage of the database, indicating whether the results include all awards for which the student is qualified, or whether some awards are missing from the database. To arrive at this figure, one compares the matches from one database with the union of valid matches from all databases. A score of 100% indicates that all valid matches are included in the database. Recall is reported on a scale from 0% to 100%, with 100% being best.

The following table ranks the major scholarship databases according to recall, and also reports the precision of each database.

The low precision of the Peterson's, FindTuitionand College Board scholarship matches yielded a significant number of irrelevant awards. (The NextStudent and SRN Express databases also had low precision, but yielded fewer irrelevant awards due to the smaller size of their databases.)

The National Postsecondary Student Aid Study (NPSAS) is a large survey of college students to gather data on how they paid for school. It is conducted every 3-4 years by the National Center for Education Statistics (NCES) at the US Department of Education.

The NPSAS is the largest survey of its kind. The 2003-2004 NPSAS surveyed about 80,000 undergraduate students and 11,000 graduate and professional students. The 2007-08 NPSAS surveyed about 114,000 undergraduate students and 14,000 graduate and professional students. This is a statistically significant representative sample of the student population.

The NPSAS's PRIVAID variable (PRIVAIDR in 1989-1990) contains information about private sector scholarships from "outside" sources. It excludes government, employer and institutional aid, such as athletic scholarships. The following tables provide statistics for graduate and undergraduate students based on this variable.

Undergraduate students includes students at 2-year schools in addition to 4-year schools. If the data is restricted to students at 4-year schools, 9.6% of undergraduate students received private scholarships worth $2,184 on average in 2003-04 and 8.3% of undergraduate students received private scholarships worth $2,796 on average in 2007-08. Students at 4-year colleges represent 71.5% of scholarship recipients in 2003-04 and 77.0% of scholarship recipients in 2007-08. If we further restrict the data to just the students who were enrolled full-time at 4-year institutions, 12.1% of students, or about 1 in 8, received scholarships worth $2,223 on average in 2003-04 and 10.6% of students, or about 1 in 10, received scholarships worth $2,815 on average in 2007-08. Full-time students at 4-year schools represented 63.2% of scholarship recipients in 2003-04 and 69.4% of scholarship recipients in 2007-08.

Fields marked with * have to be filled.