Cognitive Information Processing 

 (CIP)

From Chatgpt 

The core claim: the mind works like a computer. Information comes in, passes through stages, gets processed, gets stored. Behaviorists said "don't look inside the black box." CIP says the box is the story.

Three stores:

• Sensory memory: Everything enters here, and stays about half a second. Strike a match in a dark room. Even after it goes out, you hold an afterimage long enough to find the door. Huge capacity, brutally short life. If attention doesn't grab it, it's gone. The key insight is counterintuitive: the limit isn't how much gets in, it's how fast it decays.
• Working memory (short-term): Whatever you're consciously thinking about right now. Two hard limits: it holds roughly seven items, and without rehearsal it's wiped in 20–30 seconds. But "item" is elastic a phone number is hard digit-by-digit, easy in groups. That's chunking, and it's the most practical lever in teaching: break complex material into bite-sized units the learner can actually hold.
• Long-term memory: Unlimited, permanent. Information that gets in here doesn't really get destroyed but it can become unreachable (the file is on the disk; you've lost the folder name).

How does information move between them?

Sensory → working requires attention + pattern recognition. 

    Attention is a limited resource; it decides what gets through the door. Practice something enough and it becomes automatic, costing almost no attention. That's why you can drive and follow a podcast. When traffic gets bad, driving flips back to controlled processing and the podcast vanishes, because driving just ate your attention budget. Same in reading: if decoding letters isn't automatic, there's no capacity left for comprehension.

Working → long-term requires encoding. And here's the distinction that matters most:

• Maintenance rehearsal: repeating something over and over. It holds information in place. It does not make it permanent.
• Encoding (elaborative rehearsal): connecting new information to what you already know. Making it meaningful, giving it examples, imagining it, restating it in your own words.

    This is why the classic "but I studied for hours!" collapse happens: the student reread their notes again and again  pure repetition. Effort is not learning. Repetition is enough only for simple, overlearned material like multiplication tables. Meaningful material has to be encoded.

How is knowledge actually represented in long-term memory? 

    There's no single answer; several models compete. 

    Is memory a network (concepts as nodes, learned relations as links)? 

    A set of feature lists? 

    A web of propositions (not sentences, but meaning units like "a bird has wings". Notice that after reading a passage you remember the ideas, not the exact sentences)? Or a connectionist system where no single unit means anything and knowledge lives in the strengths of the connections? There's also the view that verbal and visual information get two separate codes which explains why "apple" is easier to remember than "justice": the apple has two retrieval routes, justice has one.

    Retrieval: Two forms: recall (generate the answer from scratch) and recognition (pick it from the options). The second is easier because it comes with cues built in. And from this comes the chapter's most usable principle 

    Encoding specificity: the cues you used while encoding are the best cues for retrieving it later. 

    Meaning: if you taught a concept in only one context, the learner won't find it when the question comes wrapped in another context even though the knowledge is sitting right there in memory. 

    The instructional implication is direct: teach the same concept across many contexts and examples. A related finding: recall improves when the learning and testing situations resemble each other the room, the physical state, even the mood.

Forgetting has three causes:

1. Encoding failure -> it was never really learned (the "I studied for hours" case).
2. Retrieval failure -> it was encoded, but there's no cue to get back to it.
3. Interference -> other material got in the way. New learning can disrupt old, and old learning can disrupt new (the veteran tennis player learning racquetball keeps swinging from the whole arm instead of the wrist).

So what does the teacher do? 

    Organize the material, direct attention, make chunking easy, force encoding (make students restate things in their own words rather than copying), and provide lots of practice across varied contexts. And on top of all that: teach students to monitor their own processing  this is metacognition. Telling a student "use this strategy" isn't enough; they have to know when and why it works, or they won't use it.