After Mehrabian: Nonverbal communication research

In the communications and presentations industry there are generally just two research studies that are quoted when talking about the importance of nonverbal communication. They were carried out by Albert Mehrabian and his colleagues in 1967. Forty-two years ago.

Mehrabian misinterpretation

In the second of those studies, Mehrabian proposed the 7-38-55% formula. He suggested that the perception of a speaker’s feelings was based 7% on the word used, 38% on the tone of voice and 55% on the facial expression. In two previous posts, I showed how this has often been misinterpreted to apply to the meaning that is derived from a communication, or to the feelings that are engendered in the listener. (The studies are in fact restricted to the judgement of the speaker’s feelings).

After Mehrabian

Based on the fact that these are just about the only studies on nonverbal communication ever quoted by communications consultants, you would think that these were the only pieces of research ever done on nonverbal communication. Not so. In a 1985 literature review I found 33 other studies specifically on the relative importance of nonverbal communication (Noller, 1985).

In this post, I look at the other research on nonverbal communication. There is a rich and complex literature on the issue.

Many factors affect the importance of nonverbal communication

The issue is far more complicated than Mehrabian’s figures lead us to believe. Research carried out after Mehrabian’s 1967 studies has shown that:

  • The channel carrying the negative connotation is the one likely to be used in deciding whether the message is positive or negative. (Bugental, 1970, 1972, 1974)
  • The more verbal material there is, the more important the verbal channel. (Cline, 1972)
  • When observers were asked to make a cognitive judgement about the speaker (as opposed to a judgement about how the speaker feels) the verbal channel was more important. (Freidman, 1978)
  • When observers were asked to judge whether a speaker was being honest or deceptive, the verbal channel was most important. (Kraut, 1978; Krauss, 1981)
  • The sex of the speaker, and the sex of the observer affect the importance of the channel. (Argyle, 1970; De Paulo, 1979, Noller 1980)

Little support for the findings of Mehrabian’s research

In addition many later papers specifically state that they found no support for the claim that nonverbal communication is dominant. Here’s just one quote:

“Claims from the previous literature about the primacy of the visual over the verbal channel and the particular importance of the face as compared to body or speech [Mehrabian (1967) is cited here along with others] were not supported.”  O’Sullivan, Ekman, Friesen and Scherer (1980)

Even when it came to the judgement of the speaker’s emotions, researchers concluded:

“Nothing we have found would lead us to conclude that nonverbal information, whether it be transmitted via the vocal or the visible channel, is in any sense the primary basis for the perception of another’s affect [emotions]. ” Krauss, Apple, Morency, Wenzel and Winton (1981)


In any field of scientific research, you’re likely to find some variability in results, dependent of many factors including the design of the experiment. It’s tempting to look at just the research that supports your point of view, and ignore the rest. It’s so common that it has a name: cherry picking. My conclusion is that in the presentations industry, we’ve been guilty of just that.

This is my third post in my series on Mehrabian’s research. If you haven’t already checked them out, see:

Why the stickiest idea in presenting is just plain wrong

The secondary misinterpretation of Mehrabian‘s research

Related Posts with Thumbnails
If you enjoyed this post, make sure you subscribe to my RSS feed!


  1. But Mehrabian is right! How else can we explain the marginal success of the telephone?

    “The issue is far more complicated than Mehrabian’s figures lead us to believe.”

    Allow me to use a decidedly non-academic term here: Duh. Once the conditions of Mehrabian’s original experiments are explained, anyone with a lick of common sense can easily see/hear/comprehend that those numbers have so little value as to be nearly meaningless.

    Thank you for doing the digging to find additional research. As you point out, the issue of verbal/non-verbal communication, much like other highly debated issues (like global climate change and economic recovery) is indeed quite complicated.

    I am glad to have other sources to cite. Thank you sincerely.

    (Or am I kidding? Did you read the real intent of my post that my fingers transmitted through the keyboard?) ;-)

    • Hi Todd

      Part of the problem is that it’s actually quite difficult to explain the original experiments in a short and simple sentence (I know ‘cos I tried in my previous posts on the topic).

      Yes, it’s good having other sources.

      And of course, although we may be fascinated by it this issue is not nearly as important as climate change or economics!


  2. Please note that the (sarcasm) and (/sarcasm) tags that surrounded the first paragraph were trimmed by the editor in my previous comment.

  3. In addition to your references, Trimboli & Walker (1987) note that the closer to natural conversation, the less dominant the non-verbal cues Support for this can be found in: Friedman, 1978; Furnham, Trevethen & Gaskell, 1981; Strahan & Zytowski, 1976; Trimboli, 1984. Therefore, spontaneous conversations, such as a live debate, a courtroom testimony, a lover’s quarrel, an interactive discussion etc., may indicate that words and actions have equal weight.

    To be fair, there are also supporting studies that indicate the non-verbal cues may dominate (to varying extents) when the communication is more artificial (planned, organized, rehearsed, and manipulated to meet an objective). In addition to Mehrabian-1967,1971, these include: Onwuegbuzie-1997, Crumbley-2001, Hulsmeyer-1986, Caso-2006, Ambady-1993, Rosip-2004, Coulson-2004, Smith-1979, Babad-2003, Meijer-1989, Khuwaileh-1999, De Gelder-2006, Gorham-1988, and Argyle-1965.Argyle, AIkema & Gilmour, 1972; Argyle, Salter, Nicholson, Williams & Burgess, 1970; Burns & Beier, 1973; Graves & Robinson, 1976; Haase & Tepper, 1972; Scherer, London & Wolf, 1973; Walker, 1977.

    Thus, presentations, speeches, classroom lectures, webinars, and even plays, movies TV shows, news editorials, and other preplanned public communications are likely to carry a persuasive intent that is less camouflaged by nature (conversation) and more streamlined to meet a particular need (presentation). It is in these settings that the non-verbal cues appear to dominate.

    Finally, Mehrabian does mention “feelings” as part of the equation leading to likeability (approachability). In my experience, the impact of a presentation can be highly emotional and does involve “feelings” on the part of the sender and the receiver. Therefore, the Mehrabian “myth” is quite plausible if one ascribes to the notion that content is preplanned, and meant to be persuasive in some way.

    I tend to use the Mehrabian figures as a wakeup call to let presenters know that if only 7% is the content, then the margin for error is tiny, so it has to be the RIGHT content (a single typo will be noticed). If 93% is delivery, there is some margin for error (an accidental turn away, a filler of “um”, a poorly spoken phrase, etc.).

    Regardless of the percentages, even a well-crafted speech, well-written play, or well-constructed story is subject to visual performance and vocal tones.

    I hope that this adds some more light to the topic.


  4. Hi Tom
    Thank you for adding all those references. I would need to read them to see what they actually say. My experience with reading Mehrabian’s original research papers is that they do not say what many communications authors and other experts say they do. Mehrabian himself says:

    “I am obviously uncomfortable about misquotes of my work. From the very beginning I have tried to give people the correct limitations of my findings. Unfortunately the field of self-styled ‘corporate image consultants’ or ‘leadership consultants’ has numerous practitioners with very little psychological expertise.” (31 October 2002, email to Max Atkinson, reproduced in Max’s book Lend me your Ears)

    On your points about Mehrabian’s studies, I’d note that Mehrabian was not researching the likeability of the speaker, but rather a third party listener’s judgement of whether the speaker liked the person they were talking to. Mehrabian labelled this “liking” in his equation:

    Total Liking = 7% Verbal Liking + 38% Vocal Liking + 55% Facial Liking

    I explain this in more detail in this post: Nor does his equation mention feelings.

    I tend to agree with your point that mistakes in content are more important than mistakes in delivery, but I don’t think we can use Mehrabian’s research to back that up. Mehrabian’s studies did not look at the impact of mistakes in any of the channels he studied.


  5. So that’s why “Tele communication/ Video Conference” system can not to accept ordinary people.
    Becase, display monitor showing the person’s face clearly but, he(or she) dosen’t looking at you. (Camera view is different from Eye point.)


  1. Public speaking myths - [...] in detail here: Mehrabian and Non-verbal communication. In the 43 years since, there has been no research that replicates …
  2. Govor tela – istine i zablude « Mind Readings - [...] Detaljne analize eksperimentalnog dizajna primenjivanog na istraživanje neverbalne komunikacije i njenog zna?aja u komunikaciji kao takvog, otkrile su i …