It’s finally starting to get cooler and October is upon us. I think instantly of Keats’s poem, Ode to Autumn, in which he lovingly describes how the season itself conspires with the sun to produce fruit, and imagines the season, personified, sitting patiently by a cider-press to watch “the last oozings hours by hours.” I’m not nearly so winsome about the season. I am conspiring with my stomach to consume as many new pumpkin-spice flavored sweets as I can. My level of commitment to this annual task is quite real. In fact, last autumn, I found myself involved in a deep and meaningful conversation with one of Liz Lippy’s mock trialers over the wonders of pumpkin-spice flavored Frosted Mini-Wheats. By Christmas, I was oozing myself into jeans that weren’t nearly as loose as they were on October 1st. A few weeks after Christmas I asked myself a New Year’s question: “Will you exercise?” Of course I will. “I will exercise!” I shouted into the void. I was wrong—about a great many things, actually.

In voir dire we might, often because of the pressure of a ticking clock, ask our jurors questions such as “Will you fairly consider the defenses we will raise against the claims here?” or “Will you presume throughout the trial that Dr. Jonathan Crane is innocent of the charges?” As it turns out, clock or no clock, science suggests we could be wrong about trying to secure commitments that way and there might be a better way—a brain way—to try to produce the fairness or presumptions we want.

Way back in 1980, a clever researcher by the name of Steven Sherman decided to run some interesting experiments on the citizens of Bloomington, Indiana In one experiment, he had a researcher call up random numbers from the phone book. The people surveyed were asked one of two things. The first group was told that a survey was being conducted by the university because the researcher heard the American Cancer Society (ACS) was calling people asking for help. The researcher then asked, “If they called you, would you volunteer 3 hours to help them with their cancer research funding drive?” Three days later, the same people were called by a researcher posing as a solicitor from the ACS and asked if they actually would commit the three hours. None of those surveyed knew the two calls were related in any way. The second group, however, was called just by the presumed ACS solicitor and asked, “Will you help us by giving three available hours to our cancer research funding drive?” 2

Of those in the latter group—asked immediately to volunteer—only 4% volunteered. However, those in the first group who were first asked to predict their later behavior…well, nearly half of everyone in that group predicted they would volunteer if asked. When they were ultimately asked to volunteer, 31% of everyone in this “make-a-prediction” group actually did volunteer. More importantly, 92% of those people who actually did volunteer on the second call were people who predicted they would volunteer! 3 A rather startling number when one considers that only 4% of test subjects volunteered immediately without first making a prediction about their future behavior. The author attributes the high-percentage prediction of one’s later volunteerism to human beings following a kind of internal, moral script or “stereotyped response sequence” in which we humans will tend to over-predict our own moral goodness.  4 If you’re saying to yourself, “Well, shoot, I recognize all this behavioral stuff as the self-fulfilling prophecy!” then you’re doing much better than Steven Sherman who decided 41 years ago to call it “the self-erasing nature of errors of prediction.”

Now here is the fun part. Sherman is perhaps more whimsical than his hyper-technical word choice suggests. He conducted another experiment similar to that above. Here, one test group was asked to predict whether they would sing the Star-Spangled Banner over the telephone if they were later asked to do so on a different phone call. The second group was just asked to belt it out on the spot—no prediction. In the first experimental group, 44% predicted they would sing it and 40% of the group ultimately sang it when asked a few days later. However, a full 72% of those directly asked to sing it gave proof through the call that they were up to the musical task asked of them! I am sure you spotted a bit of a difference in singing as opposed to volunteering for cancer research: Those predicting whether they would volunteer for cancer fundraising were predicting their own moral behavior. Those predicting their willingness to sing were not following any similar moral “script.” By now, we can see the potential applications to the formatting of questions on voir dire, so I need not spell it out for the trial lawyers in the room.

I must, however, and for the sake of my future autumn-involved, pumpkin-spice- consuming self, mention one related study. As you might imagine, marketers have seized on this self-fulfilling prophecy phenomena. A marketing professor at Washington State University conducted an experiment on people who had a gym membership but hadn’t attended their gym in at least a month.   The first subject group in the experiment was asked if they expected (read: “predicted”) they would use the gym in the next 7 days. Like Sherman’s experiments above, the second experimental group was simply asked if they were members of a health club and not asked to predict anything. Researchers then tracked the gym attendance for both groups over the following week and then for 6 months afterward. 60% of those asked to predict their attendance said they thought they would attend the gym during that week. However, only 7% of those not asked to predict their attendance actually did attend the gym that week. Sherman’s study above helps us understand that over-estimated prediction of future good behavior. But here’s the whip cream on that slice of pumpkin pie: Over the next six months, those who were asked to predict their usage ended up using the gym twice as much as those who made no prediction at all!  6

Yes, there is something wonderful and fascinating about autumn. The intermittent, trundling parade of Monarchs flying south, guided by some mysterious and magnetic pull. The brassy chimes of dry leaves stirred by northern winds. The last, sweet oozings from the cider press. Pumpkins. Spices for pumpkins. I am just going to say that I predict this fall will be no different from the last—and I will be full-filled.


1Steven J. Sherman, On the Self-Erasing Nature of Errors of Prediction, 39 J. PERSONALITY & SOC. PSYCHOL. 211 (1980).
2See Id. at 217. I am really summarizing the spirit of the questions asked and not directly quoting the questions posed. Even Sherman doesn’t provide the words used for the questions posed to each group. I use “would you” and “will you” here to draw the distinction between being asked to predict and being asked to commit.
3 Id.

4 Id. at 218.
5 Eric Spangenberg, Increasing Health Club Attendance Through Self-Prophecy, 8 MKTG. LETTERS 23 (1997).
6 Id. at 27.