Is color psychology science or marketing?

Short answer

Both, but in different proportions. There is a serious scientific field studying how colors affect people's emotions, thoughts, and behavior; even this field's own researchers say it's still young and that many of its findings aren't solid. The charts circulating online claiming that "blue inspires trust, red stimulates appetite" mostly come not from this research but from experience, tradition, and marketing copy.

Where do color-meaning charts come from?

Search online for "the meaning of colors" and you'll run into nearly the same chart every time: red for energy and passion, blue for trust and calm, green for nature and peace, yellow for happiness, black for luxury. These charts get presented as guides for choosing a brand's colors, painting a house, even picking an outfit. Most cite no source at all; when one is given, it's usually another blog, a design book, or an old historical theory.

Some of these meanings really do come from old cultural associations: the link between red and fire and blood, or between green and vegetation, is thousands of years old. Some come from the personal observations of artists and consultants who wrote about color in the 19th and 20th centuries, and some come straight from the language of the advertising and branding industry. The result is that information of wildly different reliability ends up presented in a single chart, with the same air of certainty.

A frequently cited number is a good example of this. A marketing article published in 2006 states that people make up their minds within ninety seconds of encountering a product, and that 62 to 90 percent of that judgment rests on color alone. The number has been repeated in countless presentations and online articles ever since. The question you should ask as you read it is this: on which products, with which people, by what method was that range measured, and what exactly does "resting on color" mean?

Experimental research vs. anecdote

An anecdote is a single observation or story: "We painted our café green, and customers stayed longer." An observation like that can be interesting, but it doesn't explain much. The menu might have changed around the same time, the season might have turned, a nearby business might have closed. The only way to separate the effect of color from everything else is an experiment that changes only the color and holds everything else constant.

The sneakiest confound in color research is that color itself has three separate dimensions. As DT-01 distinguishes, every color carries a hue, a lightness, and a saturation value at the same time. When an experiment compares red and blue, the two colors usually differ not only in hue but in lightness and saturation too. If a difference turns up in the result, there's no way to know whether it came from hue, from brightness, or from vividness.

The second important distinction is between association and effect. A person saying "blue reminds me of calm" is an association; actually being calmer in a blue room is an effect. People can report consistent associations about colors, but whether those associations change behavior is a separate question. PS-02 examines this distinction in detail, with findings.

The field's own assessment

Two of the most cited researchers in color psychology assessed the state of the field in 2014, in one of psychology's most respected review journals. Their conclusion was neither dismissive nor glowing: there had been recent theoretical and methodological progress, but the field was still at an early stage. It wasn't known under what conditions many findings held, who they worked for, or how far they carried over into real life.

Opposing view

The other side of this picture deserves a fair hearing too. Some researchers stress that color–emotion associations turn out to be surprisingly similar across many countries, and that this can't simply be a random marketing invention. Practitioners also point out that the signaling function of color, for instance in traffic and safety colors, is proven and vital. Marketers argue that even without believing in the miraculous effect of a single color, following the color codes established in a category, or deliberately breaking from them, has measurable consequences. In other words, the problem isn't that color psychology exists, but that it gets exaggerated and presented without sources.

Five questions for testing a claim

When you come across a claim about color, asking these five questions usually reveals how much weight it can bear, quickly.

The first question is the source: is the claim based on research, or is it just a repetition of another piece of writing? Without a cited source, a claim is at most an opinion. The second question is what was measured: an association, or an effect on behavior or the body? "People associate blue with trust" and "more money gets deposited at banks with blue logos" are two very different claims.

The third question is how the color was defined: were the lightness and saturation of the colors being compared equalized, and under what light, on what screen, was the experiment run? The fourth question is sample size and replication: how many people took part, and was the result replicated by another research group? A single study with a small sample is weak evidence, no matter how striking it looks; how this problem plays out in the field is covered in PS-05 and PS-06. The fifth question is context: in which culture, which age group, which situation was the result obtained, and can it be carried over to another context?

Example — testing the claim that "blue inspires trust"

A design agency claims that a blue logo inspires trust for a new finance app. Let's imagine an experiment to test this. Participants are randomly split into two groups. Both groups see the same app promo; the only difference is the logo's color: blue for one group, another color for the other. The critical point is that the two colors are equalized for lightness and saturation; otherwise, any difference could be attributed to brightness rather than color. The colors are shown on calibrated screens, or printed and viewed under the same light.

Participants then aren't just asked "did you find it trustworthy?"; where possible, a behavior is measured instead, for instance the tendency to download the app or provide information. The experiment is run with a large enough sample and repeated across different age groups, ideally in different countries. An experiment like this either supports the sentence "blue inspires trust," or turns it into something far more honest, such as "blue can create a small difference in some people, under some conditions."

Skin tone also has a bearing on experiments like this: the skin tones of the faces used in an app's or an ad's visuals can create a stronger trust or affinity effect than the logo's color does. A good experiment keeps these visuals identical across both groups, separating the color's effect from theirs.

Common misconception

Summary

1. The color-meaning charts online are mostly a mix of cultural association, personal observation, and marketing language; the research-based part is small and usually presented without sources.
2. Color psychology is a real research field, but according to its own researchers it's still at an early stage; many studies have conflated color's hue, lightness, and saturation dimensions.
3. Testing a claim with questions about source, what was measured, how the color was defined, sample and replication, and context lets you separate association from effect and evidence from anecdote.

Test yourself

1. (Easy) What are the two concepts that describe the difference between a color reminding you of an emotion and that color actually creating that emotion?
2. (Easy) According to the review published in 2014, what stage is the field of color psychology at?
3. (Medium) How would you design an experiment to test the claim that "blue inspires trust"?
4. (Medium) In an experiment comparing red and blue, why isn't it enough to change only the hue?
5. (Hard) When you come across a number like "62–90 percent of the judgment rests on color," what questions do you ask?
6. (Hard) Does the fact that color–emotion associations turn out similar across many countries justify color psychology charts? Justify your answer.

Answers

1. Association and effect; association is the thought or feeling a color evokes, effect is the color actually changing behavior or the body ("Experimental research vs. anecdote").
2. At an early stage; there has been recent progress, but much work is still needed on boundary conditions and generalizability to real life ("The field's own assessment").
3. I'd split participants randomly into two groups and show both the same promo, differing only in logo color; I'd equalize the two colors for lightness and saturation and present them on a calibrated screen or under the same light. I'd measure a behavior, not just an opinion, with a large enough sample, and replicate it across different groups ("Example — testing the claim that 'blue inspires trust'").
4. Red and blue usually differ in lightness and saturation too; if these dimensions aren't equalized, there's no way to know whether a difference found comes from hue or from brightness ("Experimental research vs. anecdote").
5. I'd ask on which products, with which people, by what method it was measured; what "resting on color" actually means; how large the sample was; and whether the result has been replicated ("Where do color-meaning charts come from?", "Five questions for testing a claim").
6. Not entirely. Associations turning out similar shows they're not random, but association isn't effect; similarity also doesn't erase cultural differences, and charts erase these distinctions ("Opposing view").

Sources

• Elliot, A. J., and Maier, M. A., "Color psychology: effects of perceiving color on psychological functioning in humans," Annual Review of Psychology 65: 95–120, 2014.
• Elliot, A. J., "Color and psychological functioning: a review of theoretical and empirical work," Frontiers in Psychology 6, 2015.
• Singh, S., "Impact of color on marketing," Management Decision 44(6): 783–789, 2006.

Next step

PS-02 fills in the distinction this lesson sets up with findings: it looks at what multi-country color–emotion studies have found, the distance between association and effect, and how lightness and saturation can be more decisive than hue. PS-06 zooms in on the methodological problem: it covers small samples, publication bias, preregistration, and stimulus controls like screen calibration through the lens of the replication crisis.