Skip to content
AP Psychology · Cognition

Levels of processing

Levels of processing appears in 9 reviewed AP Psychology questions and sits in Cognition. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

Levels of processing is Fergus Craik and Robert Lockhart's 1972 idea that how well you remember something depends on how deeply you process it. Shallow processing notes how a word looks, middle processing notes how it sounds, and deep processing focuses on what it means.

Why it matters

It offered a different explanation from the stage model, which says memory depends on moving information through stores. It also explains why linking new facts to things you already know, called elaborative encoding, makes them last.

How the exam tests it

Questions ask which kind of encoding gives the strongest memory, and the answer is semantic encoding. You may also see the self-reference effect, where asking "does this word describe me?" beats asking whether it rhymes or is in capital letters.

Don’t confuse it with

Students confuse it with the Atkinson-Shiffrin stage model, but levels of processing says depth of meaning matters, not how long an item sits in a memory store.

Levels of processing in real questions

Example 1Easier
Craik and Lockhart proposed the levels of processing framework as an alternative to the stage model of memory. According to this framework, which type of encoding produces the deepest, most durable memory traces?
  1. ASemantic encoding, which focuses on the meaning of words and their relationships to other knowledge and personal experienceCorrect
  2. BSyntactic encoding, which focuses on the grammatical structure and sentence position of words in a passage
  3. CPhonemic encoding, which focuses on the sound of words (how they rhyme or what they sound like when spoken)
  4. DStructural encoding, which focuses on the physical appearance of words (their font, capitalization, or visual form)

Why A is right

Craik and Lockhart's levels of processing framework proposes that memory durability is a function of depth of processing, not time in a memory store. Structural/visual processing (shallow) produces weak, rapidly fading memories. Phonological/acoustic processing (intermediate) produces somewhat better memories. Semantic processing — focusing on meaning, associations, and connections to existing knowledge (deep) — produces the strongest, most durable memory traces. Elaborative encoding (connecting new information to existing knowledge) is the deepest form.
Why the other 3 choices miss
B
Incorrect because syntactic encoding (grammatical structure analysis) is not specifically identified as a processing level in Craik and Lockhart's framework. If it were, it would likely be at an intermediate level between phonemic and full semantic processing.
C
Incorrect because phonemic encoding (sound-based processing) is an intermediate level of processing — better than structural but less effective than semantic encoding. While better than shallow processing, phonemic processing does not produce the deepest, most durable memories.
D
Incorrect because structural encoding (focusing on visual/physical features like font or capitalization) represents the shallowest level of processing in Craik and Lockhart's framework, producing the weakest and least durable memory traces.
Example 2Medium
The self-reference effect in memory research refers to which finding?
  1. APeople tend to remember information that flatters their self-image better than neutral or self-threatening information
  2. BMemories about oneself are more susceptible to distortion and false memory effects than memories about other people or external events
  3. CPeople tend to recall their own past behavior as more positive and consistent than it actually was, due to self-serving memory biases
  4. DInformation processed in relation to oneself ('does this describe me?') is remembered better than information processed otherwiseCorrect

Why D is right

The self-reference effect (Rogers, Kuiper, & Kirker, 1977) demonstrates that when information is processed in relation to oneself — asking 'does this word describe me?' rather than 'is this word in capital letters?' or 'does this word rhyme with X?' — memory for that information is significantly enhanced. This occurs because self-referential processing engages elaborate, organized, and emotionally rich self-knowledge structures, constituting one of the deepest forms of elaborative encoding.
Why the other 3 choices miss
A
Incorrect because the self-reference effect applies to information processed in relation to oneself regardless of whether it is flattering or not. The enhancement is due to the depth of self-referential processing, not to a bias toward positive self-relevant information.
B
Incorrect because the self-reference effect refers to enhanced recall for self-related information, not to greater susceptibility to false memories. While self-related memories can be distorted, this is not what the self-reference effect specifically describes.
C
Incorrect because self-serving memory biases (remembering the past as more favorable) are a real phenomenon but are separate from the self-reference effect. The self-reference effect is specifically about encoding depth and recall enhancement, not about the accuracy or positivity of recalled self-relevant information.

Studied alongside

Other concepts from Cognition.