A cultural anger at being bamboozled by numbers is expressed in the classification of degrees of untruth: Lies, Big Lies and Statistics. Increasingly our lives and our political decisions are meshed with big numbers. But we have to rely on “experts” whom we don’t always trust or even understand to tell us what they mean. Frustrating. Anti-democratic.

A mathematical issue that hit headlines all over the state last week provides an excellent example of how seventh-graders with laptops could escape from the dilemma and help parents do so as well.

A New Hampshire firm (MKV) aroused the ire of our own Maine experts on polls by reporting that one of our gubernatorial candidates had a 40 percent lead over his closest rival. The fuss was not simply a matter of distrusting pollsters from away: our experts certainly gave lots of technical-sounding reasons for their criticisms. But whether or not these reasons make sense for experts in the polling business, they are gobbledy-gook for most citizens. Even I, although I was trained as a mathematician and have even taught advanced courses on statistics, had to do some work before I made up my mind about the arguments around MKV. What I want to talk about today is how our youngsters could undertake this kind of work in ways that are dramatically different from what I (or anyone) could have done 50 years ago. Provided, of course, that they have the tools to do so – tools like computers and fluency in using them.

Fifty years ago (when I was a junior faculty member at a university) my tool would have been algebra peppered with stuff like “sigma equals the square root of n times p times q.” Ten years earlier (when I was a seventh-grader) I couldn’t have done much at all. But today I know how to bring to the problem methods that could be perfectly accessible to seventh-graders … at least to our seventh-graders starting next year. I spent the last half-hour composing a little computer program using a programming language (“Logo”) that I have taught in elementary schools. To give you an idea in a few sentences of what this might mean, I’ll describe how programming could be used to clarify a similar situation that is simpler but still bothers a lot of people.

Some of the smartest people I know (for example my own wife “They never asked me! Or anyone I know”) find it hard to accept the idea that asking a few hundred people can tell you anything reliable about how tens of millions of people will behave. And very few people have any idea how to judge for themselves how many hundreds of people the pollsters would have to ask. To see how a kid with a computer and an easily accessible level of technological fluency could do so without much trouble lets first imagine a much more difficult experiment with old-fashioned materials. Imagine that I have put 6 million blue balls and 4 million green balls in a large urn. You are given the job of guessing the color of the majority of balls by taking out some balls (in statistician-talk “a sample”) and judging as best you can from what you see.

How many would you have to pull out to make a reliable decision? If you pull one ball the best you could do is choose its color, but if you made a habit of this you’d be wrong four times out of 10. You’d do a little better if you drew a sample of five and judged by the majority color. You would do a bit better. With a sample size of 50 you’d be right 95 percent of the time. If you went up to 400 (the number used by MKV) you’d be wrong only once in many thousands of times. But you should be protesting that I am asking you to take all these numbers on trust. Not a Mainerly thing to do! How could you know this independently? You obviously can’t do the experiment with 10 million real balls. But it is (quite literally) child’s play to program a computer to do the experiment with 10 million “virtual balls” and to repeat it thousands of times to see how often any given strategy for deciding would lead to make a wrong decision.

Of course deciding what kind of sample you need to estimate how big a lead John Baldacci has among Maine voters is more complicated. But it is a problem of the same kind. Our kids will all have the tools they need to gain experience in such things and so learn to rely on their own judgment, as every Mainer wants to do.

The only obstacle I can see would be if our schools insist on keeping the kids so busy being tested on the knowledge we old geezers learned in the past century that they don’t have time to master what would best prepare them for citizenship in the 21st.

Seymour Papert is professor emeritus at the Massachusetts Institute of Technology, a Distinguished Computer Scientist at the University of Maine and a member of Maine Learning Technology Task Force. He may be contacted

at Papert@midmaine.com.