Policy Quality Control

Policy Quality Control for User-Submitted Content

Rotten Apples implements a variety of quality control measures when incorporating user-submitted data and reviews. While it is impossible to guarantee 100% accuracy for every individual review or data point, these measures aim to: 1) reduce the impact of potentially erroneous or inaccurate data, and 2) ensure that aggregated data accurately represents a given school.

Submission of User Content

User-generated content on Rotten Apples is collected through our school surveys and account registrations, both of which are accessible via the navigation menu at the top of any page and through various links on the site. To submit content (including surveys and reviews), users must first register for a Rotten Apples account. During registration, we gather information to help identify potentially fraudulent users or reviews not associated with the school. While we do not require a school-affiliated email address—since many schools do not provide consistent access to these addresses, and we want alumni to be able to submit reviews—we ensure users' identities are legitimate and relevant to the school they are reviewing.

Quality Control Procedures

Our quality control process consists of three main phases, each designed to enhance the accuracy of self-reported data and reviews.

1. Automatic Quality Control Measures

These systems automatically filter out erroneous and fraudulent data before it reaches our site or affects our scores. These checks detect outlier data (such as birth dates in 1900) and flag unhelpful content (e.g., swear words or responses like "I don't know").

2. Ongoing Community Flagging/Feedback

Users can report inaccuracies or offensive content by clicking the "Report" link next to any user-generated content. When this happens, our team manually reviews the flagged content to verify its accuracy or appropriateness.

3. Regular Data Updates and Audits

We regularly conduct audits to review the validity of user-generated data. Our automated systems identify irregularities such as an unusual pattern of consecutive 1-star or 5-star reviews, multiple reviews from the same IP address, or discrepancies between a user's location and the school's characteristics. In addition, we employ statistical methods to minimize the impact of extreme reviews from users who may be either overly satisfied or dissatisfied.

Quality control is an ongoing process at Rotten Apples, and we are committed to continuously improving these measures as we receive feedback from users, schools, and the data itself.

Effective: February 6, 2025