ComPAIR vs PeerEval

ComPAIR is a self-hosted web application for peer review in which students compare pairs of work and give feedback with both answers in view. Instructors can enable adaptive scores based on comparison outcomes and use them to assemble later pairs. The application has student, instructor, and system administrator roles with distinct permissions, plus rubric scoring and moderation tools. It supports LTI login, course membership, and grade updates; administrators configure LTI consumers with a key and secret. Documented authentication options also include username and password, CAS, and SAML 2.0. Learning analytics can be sent to a Learning Record Store through xAPI or Caliper, though both are disabled by default. The default attachment limit is 250 MB, with PDF, audio, video, and image extensions allowed. Administrators can manage whether instructors see student email addresses and CAS/SAML usernames; those fields are disabled by default. ComPAIR is open source under GPL-3.0 and freely available to download, install, and use. Its README documents deployment with Docker Compose, with access at localhost:8080. The tool is described as suitable for higher education and K-12 communities.

PeerEval is a peer assessment tool for classroom evaluation sessions. Instructors use a browser interface to enroll students, build evaluation rubrics and download results for analysis. Rubric scoring and moderation tools are supported, along with learning management system integration and gradebook synchronization. Students enter an instructor-provided session name to access an evaluation, add written comments and modify submitted work while the class session remains active. The service runs on the web, and an iPhone app is listed as free on the Canadian App Store. PeerEval is described as paid, but pricing is available only on request and the maker’s pages do not state plan amounts. The App Store identifies Thomas N Robb as the developer and seller. Apple’s listing says the developer has not provided privacy-practice details to Apple. Instructors therefore get browser-based setup and result handling, while students complete and revise rubric-based peer evaluations during class activities.