21 days, 3%, and other self-help myths

Evaluating the evidence behind common self-improvement truisms

Focus: Achieving Your Highest Priorities, FranklinCovey's flagship time-management seminar, asks seminar-goers to make several commitments in a "21-day experiment"--to plan their day each of the next 21 days, for instance.

When I took the seminar in the fall of 2005, our facilitator explained: "Why 21 days? Because that's how long it takes to form a new habit." I can’t find this in the abridged audio of Focus, but a quick Google search confirmed others also remember this statement. FranklinCovey co-founder Stephen R. Covey and FC blogger Stephanie Vozza both repeat the 21-day story on the web.

Unfortunately, it isn’t true. Jeremy Dean wrote a great summary of recent findings on habit formation by Phillipa Lally and colleagues (Lally et al., 2009):

Although the average was 66 days, there was marked variation in how long habits took to form, anywhere from 18 days up to 254 days in the habits examined in this study. As you'd imagine, drinking a daily glass of water became automatic very quickly but doing 50 sit-ups before breakfast required more dedication. . . .

What this study reveals is that when we want to develop a relatively simple habit like eating a piece of fruit each day or taking a 10 minute walk, it could take us over two months of daily repetitions before the behaviour becomes a habit. And, while this research suggests that skipping single days isn't detrimental in the long-term, it's those early repetitions that give us the greatest boost in automaticity.

Dean blames Maxwell Maltz for the 21-day myth. According to Maltz, amputees take 21 days on average to adjust to their condition; he argues that the pattern applied broadly. But as Dean says, "Unless you're in the habit of sawing off your own arm, this is not particularly relevant."

This is also a problem for other common self-help tropes. Take goal-setting: Zig Ziglar, Brian Tracy, and Tony Robbins have each said that only 3% of Americans set written goals, and that goal-setters outperform others. The original study behind the 21-day story was interpreted too broadly, but the study behind this 3% stat never happened, according to a 2007 Fast Company column:

The story, as told by consultants, goes like this: In 1953, researchers surveyed Yale's graduating seniors to determine how many of them had specific, written goals for their future. The answer: 3%. Twenty years later, researchers polled the surviving members of the Class of 1953 -- and found that the 3% with goals had accumulated more personal financial wealth than the other 97% of the class combined!

It's a consultant's dream anecdote: a vivid Ivy League success story that documents the cause-and-effect relationship between goals and personal success. It's powerful! It's compelling! It's also completely untrue . . .

Research Associate Beverly Waters reports that a recent outbreak of articles citing the study in publications as diverse as Dental Economics and Success magazines prompted her to undertake an exhaustive search of Yale alumni archives -- where she found no evidence that such a study had ever been conducted. Says Waters, "We are quite confident that the 'study' did not take place. We suspect it is a myth."

Not that hard evidence like that has ever stopped a consultant. From his Solana Beach, California office, consultant Brian Tracy responded to Waters's findings: "Heard this story originally from Zig Ziglar. If it's not true it should be."

A motivational speaker or author trying to validate a technique faces a challenge: the literature of the social sciences is vast, dense, and not focused on personal achievement. Far from providing pat advice for success, the results of most research (in my experience) are complex, probabilistic, and heavily qualified.

In First Things First, Covey writes, "In the field of personal development, one of the few things that can be empirically validated is that individuals and organizations that set goals accomplish more." I do not know if he was thinking of the Yale study, but he has a point: few things can be empirically validated, so evidence is valuable. The finger-pointing (more extensive in the full article) about the Yale study illustrates how these "stats" spread: once one guru cites it, others assume its veracity and cite it themselves.

I don't believe this invalidates goal-setting, or the entire Focus seminar (obviously). But a self-improvement "truism" should be, you know, true. The background matters. So does the application. Some principles generalize from one situation to another. Maltz's amputee observations do not.

In the face of any advice or principle, consider the underlying evidence. Is there any? Is it accessible? Does it actually apply to you? When developing a habit, the difference between 21 and 66 days is not a month and change; it is failure versus success. Accepting the 21-day myth means trying to live up to a standard which never applied.

UPDATE (2:30 PM, Feb 12, 2010): FranklinCovey's retail spinoff, FranklinCovey Products, recently invoked the 21-day myth in a February 4 post on their community blog, in which they vaguely assert, "Studies have shown it takes 21 days to develop a habit..."

Monday, October 12, 2009