6.3 Encoding and Retrieval in Short-term Memory

Encoding and Retrieval in Short-term Memory

Short-term memory is your brain's temporary storage system. It holds onto information for a brief period, allowing you to process and manipulate it. This system is crucial for tasks like remembering a phone number long enough to dial it or following a conversation.

Three key processes drive short-term memory: encoding (getting information in), maintenance (keeping it active), and retrieval (getting it back out). Understanding how each works helps explain both the power and the limitations of our moment-to-moment memory.

Encoding Processes in Short-term Memory

Encoding in short-term memory

Before information enters short-term memory, it passes through sensory memory. Iconic memory (visual) lasts only about 250 ms, while echoic memory (auditory) persists for roughly 3–4 seconds. Other senses like touch and smell have their own brief sensory stores as well. Only information you attend to moves forward into short-term memory.

Attention allocation is what determines which sensory input actually gets encoded. You can direct attention selectively to one stimulus (like listening to a single voice in a noisy room) or divide it across multiple tasks, though dividing attention reduces encoding quality for each task.

Once information has your attention, several encoding strategies shape how well it's stored:

• Chunking groups individual pieces of information into larger meaningful units, effectively expanding capacity. A 10-digit phone number (3-1-0-5-5-5-1-2-3-4) becomes three chunks (310-555-1234).
• Pattern recognition matches incoming sensory data with knowledge you already have, which speeds up encoding. A chess expert encodes a board position faster than a novice because they recognize familiar configurations.
• Elaborative rehearsal connects new information to existing knowledge, deepening the encoding. For example, learning the word "gregarious" sticks better if you link it to a friend named Greg who's very sociable.
• Maintenance rehearsal simply repeats information mentally or out loud to keep it active. This holds information in the short-term store but doesn't necessarily create a lasting memory.

Maintenance and Retrieval in Short-term Memory

Attention and rehearsal for retention

Attention acts as a limited-capacity resource. Without it, information can't transfer into short-term memory in the first place. But attention also plays a role in keeping information active once it's there.

There are several maintenance mechanisms:

• Maintenance rehearsal involves subvocal repetition, like silently repeating a phone number over and over. It extends how long information stays in short-term memory but doesn't do much for long-term retention.
• Elaborative rehearsal processes information at a deeper level and increases the likelihood of transfer to long-term memory. Creating a mnemonic device is a classic example.
• Attentional refreshing briefly redirects focus to each item in memory, maintaining activation without verbal rehearsal. Think of it as mentally "touching" each item to keep it alive.
• The articulatory loop (a component of Baddeley's working memory model) combines a phonological store with a subvocal rehearsal process, and it's the main system supporting verbal information retention.

Serial vs. free recall

How you're asked to recall information changes what you remember:

• Serial recall requires items in their original order. It produces both primacy effects (better memory for early items) and recency effects (better memory for recent items). Errors tend to be position errors (putting an item in the wrong spot) rather than item errors (recalling a completely wrong item). Remembering a sequence of digits is a typical serial recall task.
• Free recall lets you report items in any order. The recency effect is usually more pronounced here, and people tend to cluster semantically related items together. Recalling items from a grocery list is a good example: you might recall all the fruits together even if they weren't presented that way.

The key difference is that serial recall relies heavily on temporal-positional information (where each item appeared in the sequence), while free recall allows more flexible retrieval strategies like semantic grouping.

Retrieval factors in short-term memory

Several factors determine whether you can successfully pull information back out of short-term memory:

Decay theory proposes that information simply fades over time without active maintenance. This is one reason short-term memory is so brief.

Interference is often a bigger factor than pure decay:

1. Proactive interference occurs when older memories disrupt recall of newer information. If you just memorized one phone number, it can interfere with learning a second one.
2. Retroactive interference occurs when new learning disrupts recall of older information. Learning that second number might make you forget the first.

Similarity effects make interference worse. Phonologically similar items (like "man," "can," "ban") are harder to keep straight in verbal short-term memory, while visually similar items cause more confusion in visual short-term memory.

Context and cue effects also matter:

• Environmental context can facilitate retrieval. You're more likely to recall information when you're in the same room where you learned it.
• Retrieval cues provide partial information that helps you access the full memory. Hearing the first letter of a forgotten name, for instance, can trigger recall of the whole name.

Output interference is a subtler factor: the act of recalling early items in a list can interfere with your ability to recall later items, because retrieval itself creates interference.

Finally, capacity limitations constrain retrieval. When information exceeds short-term memory capacity (classically estimated at $7 \pm 2$ items, though more recent research suggests the true limit is closer to 3–5 chunks), retrieval accuracy drops sharply.