Question types overview

Choice-based questions

Multiple choice is the workhorse for knowledge checks and screening at scale. It comes in single-select and multi-select variants, and single-select options can be laid out in two to four columns — a compact form that works well for Likert-style scales and short answers. Because an answer key drives auto-scoring, choice-based items give you instant, consistent results without evaluator time, and support partial credit and negative marking when you need finer-grained scoring.

Written and structured input

When you want candidates to produce an answer rather than pick one, use the written and structured types: open text for free-form written answers, numeric input for a single number, fill-in-a-table for grid-shaped data, matching for pairing items across two columns, and ordering for arranging steps into the correct sequence.

Most of these auto-score: numeric items check the answer against a correct value with a tolerance, and matching, table, and ordering items score against their keys. Open text can auto-score with exact, contains, or regex rules, or go to rubric-based human review for reasoning and expression that rules cannot judge. Items without a key auto-score as null and wait for manual review — null never converts to zero, so unscored responses are flagged rather than penalized.

Media capture, technical, and evaluator-driven types

Audio recording, video recording, and photo capture collect the candidate's voice, on-camera presence, or a physical artifact as evidence. They suit communication-heavy roles and situational responses where delivery matters as much as content, and where you want the response itself preserved as reviewable evidence.

File upload accepts work products such as documents or spreadsheets, and the code type is built for programming tasks. Both are the natural fit for work-sample exercises where you want to review the actual artifact a candidate produces. A rubric-only item takes no candidate input at all — evaluators score it directly, which suits observed exercises. Beyond individual items, SkillCort also offers AI-led interview questions and case simulations, created through their own flows, where an AI interviewer conducts a real conversation with the candidate.

Four layouts

Every item uses one of four layouts: the regular single-column layout, question-beside-answer, or a sidebar on the left or right of the answer area. Sidebar layouts keep a brief or case study visible while the candidate responds, so they can read the material and answer without scrolling back and forth. Use a sidebar whenever the question depends on a longer text the candidate must reference while answering.

Points and the required toggle

Each item carries a points value. Points set how much the item contributes before section and skill weights are applied, so give heavier items more points. Once an item is placed in an assessment, it also gets a required toggle that signals candidates must answer it. Set both deliberately per item rather than leaving defaults, since they shape scoring and completion behavior.