How People Actually Learn
The 10 most effective methods for teaching complex ideas to anyone, ranked by cognitive science research. This determines how every evidence card in God Examined is structured.
Why Most Teaching Fails
Most explanations fail for the same reason: they start with the abstract concept and expect the reader to build a mental model from scratch. Cognitive science shows the opposite works. The brain learns by attaching new information to things it already understands. If there's nothing to attach to, the new information slides off.
The average person can hold 4 items in working memory at once (Cowan, 2001 — not Miller's famous 7, which has been revised downward). Every technical term you drop without explanation consumes one of those 4 slots. Drop 3 jargon words in a sentence and you've filled 3 of 4 slots with confusion, leaving only 1 for the actual idea.
The 10 Methods — Ranked by Effectiveness
These are ranked by research-backed effect sizes on comprehension and retention, drawn from Hattie's meta-analysis (1,400+ studies), Mayer's multimedia learning principles, and cognitive load theory (Sweller, 1988).
| Rank | Method | Effect Size | Why It Works | Example |
|---|---|---|---|---|
| 1 | Concrete Analogy First | d = 1.32 | Hooks new info to existing knowledge | "Infinite regress is like asking who's first in a line with no beginning" |
| 2 | Comparison to Known | d = 1.23 | Makes abstract concrete via familiar reference | "Plato has 7 copies. Jesus has 25,000." |
| 3 | Visual Before Verbal | d = 1.14 | Brain processes images 60,000x faster than text | SVG diagram showing the trilemma before explaining it |
| 4 | Spaced Repetition | d = 0.94 | Revisiting the same idea in different contexts cements it | Same argument appearing in summary, then bullets, then deep dive |
| 5 | Story / Narrative Structure | d = 0.93 | Brains retain stories 22x better than facts (Stanford) | "Three suspects, two alibis, one left standing" |
| 6 | Socratic Elimination | d = 0.82 | Process of elimination feels like reader's own discovery | "Option A fails because X. Option B fails because Y. Only C remains." |
| 7 | Concrete Numbers | d = 0.78 | Specificity signals credibility and aids recall | "42 sources" not "many sources." "89,708 people" not "a large study." |
| 8 | Objection-Response Pairs | d = 0.76 | Addresses the voice in the reader's head saying "but what about..." | Steel-man the objection, then dismantle it with specific evidence |
| 9 | Multi-Sensory Encoding | d = 0.73 | Text + image + table = 3 encoding channels instead of 1 | Same data shown as paragraph, SVG chart, AND table |
| 10 | Emotional Anchor | d = 0.68 | Emotion triggers hippocampus, converting short-term to long-term memory | "His enemies picked up stones. They understood exactly what he was claiming." |
1. Concrete Analogy First
Effect size: d = 1.32 — the single most effective teaching method
The brain does not learn by absorbing new concepts in a vacuum. It learns by mapping new information onto structures it already has. An analogy provides the structure. Without it, the concept has nowhere to land.
Gentner's Structure Mapping Theory (1983) showed that analogies work because the brain transfers relational structure from a known domain to an unknown one. When you say "infinite regress is like asking who's first in a line with no beginning," the reader instantly imports everything they know about lines (they have starts, middles, ends) and sees that this "line" is broken.
Gentner, D. (1983). Structure-mapping: A theoretical framework for analogy. Cognitive Science, 7(2), 155-170. | Richland et al. (2007). Analogical reasoning in science education. Cognition and Instruction.
2. Comparison to Known Quantities
Effect size: d = 1.23
Numbers by themselves are meaningless. "25,000 manuscripts" means nothing unless you know how many other ancient texts have. The brain evaluates by comparison, not by absolute value. This is called anchoring (Tversky & Kahneman, 1974).
When you say "Plato has 7 copies, Caesar has 10, Homer has 1,757 — Jesus has 25,000," the reader's brain instantly ranks them. The comparison does the argument for you. You don't need to say "this is a lot." The reader's own comparison system says it.
3. Visual Before Verbal
Effect size: d = 1.14
The brain processes images 60,000 times faster than text (3M Corporation / Zabisco research). When a reader sees a bar chart showing Jesus' 25,000 manuscripts next to Plato's 7, they understand the scale difference in under 1 second. The same information as text takes 30+ seconds to process and may never be fully grasped.
Mayer's Multimedia Principle (2009): people learn better from words AND pictures than from words alone. But the picture must come FIRST — it creates the spatial framework that subsequent text fills in. This is why every evidence card opens with an SVG illustration.
4. Spaced Repetition
Effect size: d = 0.94
Ebbinghaus's forgetting curve (1885) showed that people forget 70% of new information within 24 hours — unless they encounter it again. Each re-encounter resets the decay curve and makes the memory stronger. This is why God Examined presents the same argument at three levels of depth:
- Level 1 (Summary intro on main page): One-line claim with key number
- Level 2 (Evidence card summary): 3 paragraphs + proof bullets + SVG
- Level 3 (Deep-dive sections): Full detail with rebuttals, tables, sources
The reader encounters the same core argument three times in increasing depth. By Level 3, the neural pathway is well-established.
5. Story / Narrative Structure
Effect size: d = 0.93
Stanford neuroscientist Uri Hasson (2012) showed that during storytelling, the listener's brain synchronizes with the speaker's brain — neural coupling. Facts alone don't trigger this. Stories do. People retain stories 22x better than isolated facts (Stanford research).
This is why "Three suspects, two alibis, one left standing" works better than "Autogenesis is the only logically viable option for the origin of reality." Both say the same thing. The first one your brain remembers. The second one it forgets.
6. Socratic Elimination
Effect size: d = 0.82
When you tell someone the answer, they receive it passively. When you walk them through a process of elimination where they watch two options fail and arrive at the third, they feel like they discovered it themselves. Self-generated conclusions are retained 2-3x longer than received conclusions (Slamecka & Graf, 1978, generation effect).
This is the structure behind the autogenesis argument: "Option A fails because X. Option B fails because Y. Only C remains." The reader doesn't feel lectured. They feel like they solved a puzzle.
7. Concrete Numbers Over Abstract Claims
Effect size: d = 0.78
The human brain doesn't process "a lot" or "very" or "extremely unlikely." It processes 42 and 89,708 and 1 in 1017. Specific numbers signal credibility (the speaker actually knows the data), enable comparison, and create memorable anchors.
"A lot of manuscripts" → forgettable. "25,000 manuscripts vs. 7 for Plato" → unforgettable. The number IS the argument.
8. Objection-Response Pairs
Effect size: d = 0.76
Every reader has an inner skeptic. If you don't address it, the skeptic gets louder and louder until the reader stops listening. Inoculation theory (McGuire, 1961) shows that presenting and refuting counterarguments makes the original position stronger — not weaker. Like a vaccine: exposure to a weakened objection builds resistance.
The key: steel-man the objection first. Make it stronger than the critic would. Then dismantle it with specific evidence. The reader sees you're not dodging — you're engaging honestly. Trust goes up. Persuasion follows trust.
9. Multi-Sensory Encoding
Effect size: d = 0.73
Paivio's Dual Coding Theory (1971): information encoded in both verbal and visual channels is recalled 2x better than information in one channel alone. Add a third channel (tabular/spatial) and you get even more retention.
This is why the same data in God Examined appears as: (1) a prose paragraph, (2) an SVG chart, and (3) a comparison table. Three channels, one argument. The reader who skims gets the chart. The reader who reads gets the paragraph. The reader who analyzes gets the table. All three get the argument.
10. Emotional Anchor
Effect size: d = 0.68
The hippocampus converts short-term memories into long-term storage. Emotional arousal activates the hippocampus (McGaugh, 2000). Facts without emotion are stored briefly and discarded. Facts with emotion are stored permanently.
"His enemies picked up stones to kill him on the spot. They understood exactly what he was claiming." That sentence has the same informational content as "Jesus' claim to divinity was recognized by his contemporaries." But the first one you remember tomorrow. The second one you've already forgotten.
Optimal Deep-Dive Section Order
Based on the 10 methods above, the deep-dive sections in each evidence card should follow this order for maximum comprehension and retention:
| Order | Section | Method(s) Used | Purpose |
|---|---|---|---|
| 1 | The Analogy | Concrete Analogy (#1), Story (#5) | Hook the reader with something familiar. Build the mental framework before any detail. |
| 2 | The Evidence | Concrete Numbers (#7), Comparison (#2), Visual (#3) | Present the specific data with comparisons and visuals. This is the meat. |
| 3 | The Elimination | Socratic Elimination (#6) | Walk through alternatives that fail. Let the reader arrive at the conclusion themselves. |
| 4 | Objections & Rebuttals | Objection-Response (#8) | Address the inner skeptic. Steel-man, then dismantle. Build trust. |
| 5 | Comparison Tables | Comparison (#2), Multi-Sensory (#9), Visual (#3) | Same evidence in tabular format. Third encoding channel. Scannable. |
| 6 | Falsifiability | Objection-Response (#8) | "What would disprove this?" Shows intellectual honesty. Builds trust. |
| 7 | Convergence | Spaced Repetition (#4), Cumulative Force | How this evidence combines with the others. The whole is greater than the parts. |
| 8 | Verdict | Emotional Anchor (#10) | One sentence the reader takes away. Make it land. |
Applied to God Examined Evidence Cards
Every evidence card in God Examined should follow this structure, derived from the science above:
The Summary (always visible)
- Title + SVG illustration — Visual before verbal (#3). Sets the spatial framework.
- Opening paragraph with analogy — Concrete analogy first (#1). Hook to known.
- Inline data SVG — Multi-sensory (#9). Second visual channel.
- Proof bullets with evidence type labels — Concrete numbers (#7), comparison (#2). Each bullet explains how it fits.
- Closing paragraph — Convergence + emotional anchor (#10).
The Deep Dive (collapsible sections, in order)
- The Analogy (expanded) — Full version of the opening analogy with additional parallels
- The Evidence (full detail) — Every source with: who said it, when, the quote, "so what?", what it eliminates
- The Elimination — Walk through every alternative. Show each one failing. Let the reader arrive at the conclusion.
- Objections & Rebuttals — Steel-man → Response → Counter → Final. 4 stages per objection.
- Comparison Tables — Same data in scannable tabular format
- Falsifiability — "What would disprove this?" + "Has it been found? No."
- Convergence — How this combines with evidence from other cards
- Verdict — One sentence. Make it land.