The last block was only what came back. This one is me trying to understand it. Three results are worth digging into, and I will take them in order: the trust surprise, the confusion both versions caused, and the clean split between instruction and feeling. The sample is small, so this is interpretation, not proof.
1. The trust surprise
People trusted the illustration more as an instruction, even though the photograph is the more realistic image. That is the result I expected least, and honestly I am still not certain why it happened. But I can see a few possible reasons, and they probably work together.
One is that realistic and clear are not the same thing. A photograph shows one specific person, one specific door, one specific moment, full of detail that the viewer has to sort through. A drawing has already removed all of that. Nothing in the frame is accidental, so it reads as someone deliberately telling you the rule, not as a photo of a thing that happened once. The reduction itself might be what makes it feel like guidance.
Whatever the exact reason, the real world clearly agrees. Airplane safety cards, IKEA manuals, medication leaflets, and exit signs are almost always drawn, never photographed. Not because photographs were unavailable, but because for “do this, then this,” a clean diagram feels clearer and more authoritative than a real picture. My ten people landed in the same place that decades of instruction design already did, which makes me trust the signal even with a small sample.
2. The confusion is itself the finding
Both versions made people reach for words like stupid, confusing, and a plain “huh?”. My first instinct was to treat that as a problem with my images. But the more I look at it, the more I think the confusion is the actual finding, not a flaw to fix.
The story has no words at all. So when something is even slightly unclear, the viewer has nothing to fall back on, no caption, no label, no sentence to rescue them. That is exactly the condition I set out to study: communication when words are removed. The confusion is the cost of going wordless, made visible. It shows where a pure image starts to wobble and where a single word would have instantly steadied it.
You can see the same thing in real pictograms. A toilet sign works perfectly, but plenty of public symbols leave people guessing, which is why airports and hospitals so often pair the symbol with a word. The “huh?” my viewers felt is the same “huh?” everyone has had standing in front of an unlabelled sign. That reaction is not noise. It is the edge of wordless communication showing itself.
3. Instruction versus feeling
Putting the first two together gives the cleanest result of the pilot. The drawing was stronger for the instruction: clarity, trust, ease of reading. The photograph was stronger for the feeling: emotion, calm, the sense of a real person in a real moment. Comprehension was a tie, so neither medium was simply better. They were better at different jobs.
This lines up with how the two are used in the world. Instructions are drawn. Manuals, signs, and diagrams strip a task down to its logic. Emotion is photographed. Advertising, journalism, and portraits use real images to make you feel a real presence. A charity poster uses a photograph of a real face to move you, then often switches to simple icons to explain what to actually do. The medium follows the job.
Where this leaves me
The honest headline is the one I guessed at the start. There is no single winner. The drawing carries the instruction, the photograph carries the feeling, and which one is right depends entirely on what you are trying to say. The confusion sits underneath both, marking the limit of saying anything at all without words. Ten people cannot prove this, but the pattern is clean enough that I want to test it properly on the shower and the phone, with more people, in the next block.
