🧠Levels of Processing Test

Does thinking about what a word means make it easier to remember than thinking about what it looks like?

Take this free Levels of Processing Test to compare shallow visual encoding with deeper semantic encoding.

You will judge 8 words by a surface feature—whether they appear in capital letters—and 8 different words by meaning—whether they refer to something living. Afterward, a surprise recognition test checks which set you remember better.

The usual research prediction is that semantic processing produces better later recognition than shallow structural processing.

How to Take the Levels of Processing Test

1. 1Press Start Test.
2. 2For the shallow items, answer the capitalization question.
3. 3For the deep items, answer whether the word refers to something living.
4. 4Respond to the question you are given rather than deliberately memorizing every item.
5. 5Complete all 16 study trials.
6. 6Continue to the surprise recognition phase.
7. 7Select the words you remember seeing earlier.
8. 8Compare recognition for shallow and deep items.

If you intentionally use a strong mnemonic strategy for every word, you may reduce the difference between conditions.

The experiment is designed to change how you process the words during encoding.

What Are Levels of Processing?

The Levels of Processing framework was introduced by psychologists Fergus Craik and Robert Lockhart in 1972.

Their central argument shifted attention away from thinking of memory only as information moving through separate storage boxes.

Instead, they proposed that later memory depends importantly on the kind of processing performed when information is encountered.

A surface-level judgment may leave a different memory trace from an analysis involving meaning.

This became one of the most influential ideas in cognitive memory research.

Shallow Processing: Focusing on Appearance

A structural or shallow task directs attention toward physical features.

In this test, the question is:

“Is this word shown in capital letters?”

You can answer without thinking deeply about what the word means.

For example, if you see:

ELEPHANT

you can identify uppercase letters even if the task never requires you to think about elephants.

The word still receives some processing.

But the judgment can be completed mainly from its visual form.

Deep Processing: Focusing on Meaning

A semantic task requires access to meaning.

In this test, the question is:

“Does this word refer to something living?”

To answer for:

ELEPHANT

you need to identify the concept represented by the word and evaluate its category.

That semantic analysis connects the item with existing knowledge.

Under ordinary explicit recognition conditions, this richer processing tends to produce better later memory than a purely structural judgment.

Craik and Tulving's 1975 Experiments

Craik and Endel Tulving tested levels-of-processing predictions in a major 1975 paper.

Importantly, that paper contained a series of 10 experiments, not one single fixed task.

Participants answered orienting questions that required different kinds of analysis of words, including structural, phonemic, and semantic judgments, followed by unexpected memory tests.

Across the early experiments, semantic processing generally produced substantially better retention.

Your site's capitalization-versus-living design is best described as a close adaptation of the levels-of-processing paradigm.

It should not be described as a literal reproduction of one single “exact” Craik and Tulving procedure, because their paper used multiple question types and experimental variations.

Recognition vs. Recall

This page ends with a recognition test.

You see candidate words and decide whether each appeared earlier.

That is different from free recall, where you must generate studied items without seeing answer choices.

Recognition provides retrieval cues.

Levels-of-processing effects have been found in explicit memory tasks, but the size of the deep-processing advantage can depend on how memory is tested.

Do not compare a recognition percentage directly with a free-recall percentage from another experiment.

Why Semantic Processing Often Helps

Meaning-based encoding can create a richer representation.

A semantic judgment may connect a word with:

• ✓categories;
• ✓concepts;
• ✓prior knowledge;
• ✓related words;
• ✓images;
• ✓or contextual associations.

Those relationships can later provide multiple routes to retrieval.

A visual feature such as capitalization can be processed accurately without creating the same network of semantic connections.

This is why simply staring longer at information is not always the same as learning it more effectively.

“Deeper Is Better” Is Useful—but Too Simple

Levels-of-processing findings are robust, but modern memory science does not reduce everything to one ladder where semantic processing always wins.

A major qualification comes from transfer-appropriate processing.

Memory can benefit when the processing performed during learning matches the processing needed at test.

In a classic 1977 study, semantic encoding produced better performance on a standard recognition test, but rhyme-based encoding could outperform semantic encoding when the later test specifically required rhyming information.

The lesson is:

meaningful processing is usually excellent for remembering meaning, but effective memory also depends on what you will need to retrieve later.

Transfer-Appropriate Processing

Suppose you are studying pronunciation for a language exam.

Thinking deeply about a word's conceptual meaning is valuable.

But if the test specifically asks you to distinguish sounds, practicing sound-based processing is also important.

Likewise:

• ✓learn concepts semantically for conceptual questions;
• ✓practice retrieval in the format you will actually use;
• ✓solve problems if the exam requires solving;
• ✓recall information without notes if the exam requires recall.

Encoding and retrieval work together.

This is a more complete study lesson than “always process everything as deeply as possible.”

Why Incidental Learning Matters

Classic levels-of-processing experiments often used incidental learning.

Participants were told to answer questions about words but were not initially told that their memory for those words would later be tested.

This is scientifically useful.

If semantic items are remembered better even when participants were not deliberately trying harder to memorize them, the result shows that the encoding operation itself matters.

Your site's surprise recognition phase follows that general logic.

Once you know the test structure, however, a second attempt is different because you may intentionally memorize every word.

What If You Remember the Shallow Words Better?

That can happen in one short session.

You have only 8 words in each condition.

A few items can change the percentages dramatically.

Possible reasons include:

• ✓some shallow words were unusually distinctive;
• ✓you accidentally elaborated on their meanings;
• ✓attention varied;
• ✓some deep judgments were difficult;
• ✓you intentionally memorized certain items;
• ✓or normal sampling variation.

A single person's result does not need to reproduce the average research effect perfectly.

The theory is supported by patterns across many participants and trials.

Levels of Processing and Studying

The practical lesson is stronger than:

“Highlight less.”

Try to transform material.

Instead of rereading:

• ✓explain the idea in your own words;
• ✓ask why it is true;
• ✓connect it to something you already know;
• ✓generate an example;
• ✓compare it with a related concept;
• ✓apply it to a problem;
• ✓retrieve it without looking.

These activities require meaningful engagement.

However, match practice to the final goal.

If you need to perform a procedure, practice the procedure—not only explain it.

Can This Test Measure Your Memory Ability?

Not globally.

This short experiment compares two encoding conditions.

A deep-minus-shallow difference is not a general “memory score.”

It does not measure:

• ✓working-memory capacity;
• ✓long-term memory capacity;
• ✓intelligence;
• ✓dementia risk;
• ✓or learning ability across every situation.

Use it to experience an experimental effect.

If you are concerned about real-world memory decline, this browser task cannot diagnose the cause.

Frequently Asked Questions

Do More Than Look at Information

The deepest lesson from this experiment is not that capital letters are bad for memory.

It is that what your mind does with information while learning it matters.

Notice appearance and you may remember appearance.

Analyze meaning and you create richer connections.

Then practice retrieving the information in the way you will actually need it.

That combination turns a classic laboratory finding into a practical learning principle.