The student had read the chapter three times. They had highlighted it. They had reviewed the highlights twice more. By the night before the exam, the material felt familiar.

What they had done was maintenance rehearsal - passing the same words through their awareness repeatedly. What they had not done was process the material at a level that would build durable memory. The familiarity was real; the retention was shallow.

What it is

The levels-of-processing effect describes how the depth of mental processing during encoding influences later memory retention, with deeper (semantic) processing leading to stronger memory traces than shallower (phonological or structural) processing.

Shallow processing means attending to the surface features of information - its sound, its visual appearance, whether it fits a simple category. Deep processing means engaging with its meaning - what it implies, how it connects to what you already know, what examples it brings to mind, how you would explain it in your own words.

Deep semantic processing leads to better recall than shallow phonological or structural processing. The effect holds across different types of material and is not limited to simple word lists - it applies to concepts, procedures, historical events, and anything that can be processed at varying levels of meaning.

The mechanism

When information is processed deeply - semantically, in relation to meaning and context - it becomes linked to a rich network of existing concepts, experiences, and associations. Those links act as retrieval cues. When you try to recall the information later, the network provides multiple access routes.

When information is processed shallowly, it gets few such links. The trace is isolated. There are fewer routes in when retrieval is needed, and the trace is more susceptible to interference and decay.

The quality of the encoding moment matters more than the duration of the rehearsal. Spending ten minutes genuinely engaging with the meaning of a concept creates a stronger trace than spending an hour repeating its definition without thinking about what it means.

The common mistake and the better move

The common mistake: studying by re-reading, highlighting, and reviewing already-familiar notes - all forms of maintenance rehearsal that feel productive but primarily deepen familiarity rather than durable recall.

Highlighting a text gives the brain a signal that this is important. It does not require engagement with what the highlighted content means, how it connects to adjacent ideas, or what an example would look like. The highlight stays; the understanding may not.

The better move: replace passive review with elaborative encoding - processing that connects new information to meaning.

Specific forms of elaborative encoding that trigger the levels-of-processing effect:

Self-explanation. Reading a passage and then closing it to explain what it said in your own words. The gap between what you thought you knew and what you can actually produce reveals where understanding is thin.

Personal examples. When encountering a concept, stopping to generate a concrete example from your own experience. The act of generation encodes the concept more richly than reading an example.

Connection to prior knowledge. Asking: Where does this fit? What does it remind me of? What does it contradict? These questions force semantic processing that reading alone does not.

Spaced retrieval. Attempting to recall information after a delay - without looking - before reviewing. The effort of retrieval, even partially, produces a deeper trace than re-reading the same material.

None of these require more time. They require a different kind of engagement during the time already being spent.

What it is not

The levels-of-processing effect does not mean rote repetition is always useless. Maintenance rehearsal keeps information in working memory for immediate use. The effect concerns long-term retention: for durable learning, the quality of the encoding moment matters more than simple repetition.

It does not only apply to word lists. Earlier discussions of the effect focused on word pair experiments, leading some to assume it is limited to simple material. The effect extends to complex conceptual content, procedural knowledge, and applied domains - anywhere that shallow and deep processing can be distinguished.

Deeper is not always harder or slower. Elaborative encoding can become habitual. The additional time required for self-explanation or example-generation is often less than the time spent on repeated surface review - and the retention is substantially better.

Sources