Below are some of the shortcomings of the data model so far, which we are working on rectifying.
Some of the data we use are based on IPEDS 2015-6 data - the most recent "final" data set available.
Some of the country's most selective colleges are moving towards awarding aid only for financial need, rather than awarding a combination of merit and need-based scholarships. In some cases our model may project some merit-based aid for students at those institutions, meaning the estimate is over-generous.
Relatedly, often the mix of merit-based and need-based financial aid for a given college is opaque or undefined and is not consistently reported publicly. Our assumptions about this mix relies on our understanding of key institutional characteristics as well as primary data from students and college leaders, but there is extrapolation involved.
Some of the public data that is self-reported, such as the average net price, only requires that colleges include students that receive some form of aid in their calculation - thus omitting the students that pay in full. Our estimates rely heavily on those net price averages, meaning we may be overestimating aid for students who do not apply or qualify.
As stated above, the portion of our financial aid estimates which correspond to financial need is currently based solely on the zip code provided by the student, despite there being clear variability in household financial situation within one zip code - in particular with respect to factors such as assets, parental situation, and siblings in college which can impact your EFC.
Room and board is not included in either the published tuition price shown, nor the net tuition estimate. Some schools (for example, those that meet full financial need) may provide scholarships that also contribute towards the cost of room and board - that is not reflected here.
The estimates shown apply to first-time dependent freshman and are prices for that first year. Some scholarships and aid may not be guaranteed for all four years, and some may make certain requirements as to continued eligibility (e.g. meeting a minimum GPA or course load).
Our estimates cover "institutional aid," meaning scholarships that usually don't require a separate application (either need-based or merit-based) and are offered by the school directly rather than an outside party. There are many scholarships offered by third parties, and some state-sponsored (vs. college-sponsored) scholarships which are not reflected here and could add additional aid to your package.
Remember that tuition increases every year unless a college says the tuition is locked in. Our estimates of individual students’ merit are based primarily on standardized test scores and unweighted GPA (converted if the student enters a weighted GPA). These measures omit other components of merit such as rigor of high school coursework (core vs. honors or. AP/IB), athletic ability, extracurricular achievements, etc. Students who are exemplary along these dimensions may be evaluated differently by the institutions they apply to.
Student-reported data is central to the algorithm, and our starting data set may have inherent biases: Our data set may be skewed towards students who are of higher merit than the general applicant population if those students are more likely to report their awards to others. This may bias our financial aid estimates upwards.
Our data set is skewed towards students who are more price-sensitive to the cost of college attendance (and have received some aid), such that they are more willing to provide information on their awards in exchange for meaningful comparisons/information. This means that students with higher household incomes are underrepresented in our data, and that our estimates of financial aid may be less accurate for students at higher income levels.
Our data may also be skewed towards larger and more selective colleges (those that receive more applications and thus where there is greater demand for information on financial aid awards. This may mean that our estimates for less selective colleges may be less accurate.
Do you have others to add? We are eager to hear from you - whether you’re a student, parent, administrator, or researcher. Submit feedback using this form and we will follow up with you shortly.