Mastery learning
A way of teaching that breaks a course into short units and refuses to let the class move on until most students have actually learned each one. Students take a low-stakes, ungraded test at the end of a unit; those who fall short get extra help and a fresh attempt, and the cohort advances together once nearly everyone has reached the bar. The wager is that most failure comes from too little time and feedback rather than too little ability, so giving slower learners more of both pulls them up to the same standard.
An instructional approach, developed by Benjamin Bloom and his Chicago doctoral students from the late 1960s, in which a curriculum is broken into short units, each unit ends with a non-graded formative test, students who have not yet reached a mastery criterion receive corrective instruction and re-take the test, and the class advances only when most students are at criterion. The two premises are that classroom failure is mostly a function of insufficient time and missing corrective feedback rather than insufficient ability, and that a well-designed loop of test → correction → re-test will move most students to mastery on most material, given enough time.
Etymology§
The term learning for mastery (LFM) is Bloom's coinage in Learning for Mastery, a 1968 booklet from the UCLA Center for the Study of Evaluation. The intellectual lineage runs back further: John Carroll's 1963 Teachers College Record paper A Model of School Learning proposed that aptitude be re-conceptualised as the amount of time a student needs to reach a given criterion rather than as an upper bound on what they can achieve. Bloom's programme took Carroll's reframing and built the instructional apparatus around it. The empirical case for the approach was developed across Bloom's doctoral students in the 1970s and consolidated in the 1984 Two Sigma Problem address.
The structural form of mastery learning is a feedback loop the classroom rarely runs. The class progresses through a unit; the unit ends with a short test that does not count toward any grade; the teacher uses the test to identify which students have reached the mastery criterion (typically set at 80–90% correct) and which have not; the unmastered students receive a corrective — additional explanation, different examples, peer help, supplementary problems — and re-take a parallel form of the test. The class moves on when most students have passed. The two non-standard features are the non-graded formative test and the willingness to spend additional time on a subset of students rather than advance the cohort uniformly.
The construct's theoretical claim is John Carroll's: that the variable to manage is time to mastery, not level reached in fixed time. In Carroll's model, students differ in how long they need to reach a criterion under given conditions, and a classroom that delivers the same instruction to all students for the same duration produces variation in mastery as an artefact of holding time constant. Mastery learning relaxes the time constraint and, the empirical work argues, the variation in mastery collapses toward the criterion.
The effect-size evidence accumulated through the 1970s and 1980s. Bloom's doctoral student James Kulik and his collaborators produced meta-analyses through the early 1990s reporting average effects of 0.5–1.0 standard deviations on post-instruction tests when mastery learning was implemented with fidelity. The Two Sigma Problem situated mastery learning among the candidate group methods approaching the tutoring effect, with effect sizes in the 0.5–1.0-sigma range when implemented with fidelity.
The construct's reach into present-day teaching is mixed. The instructional logic is widely cited and appears in most teacher- preparation programmes; the structural form — non-graded formative testing, corrective instruction, deferred grade-book entry — is rarely implemented because school calendars and grade-book requirements push against the time-relaxed structure. Where the approach has survived intact, it is most often in domains with a clear mastery criterion (programming, mathematics drills, language vocabulary) and in computer-mediated instruction, where the adaptive loop can be automated.
Several of the contemporary grading-reform moves — specifications grading, ungrading — adopt the mastery- learning position that the grade is not the feedback loop, and that running the feedback loop requires either decoupling assessment from grading or doing the assessment differently. The vocabulary differs from Bloom's; the structural commitment to time, feedback, and a non-graded formative test is recognisably mastery-learning's.