Reliability (say painting a still-life by number) goes near the low end. Novelty (painting a still-life from life) is nearer to the high end. Differences in variability levels occur between individuals on the same task and between tasks for the same individual. The latter, intrapersonal differences led to the theory that habitual or learned variability levels, like skills, are learned. The term habitual refers to a variability level that someone usually displays in a particular task or domain. The idea is this, along with learning how to do something, you also learn how differently to continue doing it. We use computer games to test predictions based on this theory. Our past research has shown two things: first, that these levels are acquired early in exposure to new problem or skill areas, and second, how early constraints determine whether an acquired level will be high or low. More recent research involves transfer. We haveshown that variability facilitates skill transfer to novel tasks by sensitizing learners to changes in conditions/requirements. We are currently investigating how decomposing an initially very difficult problem (involving variability) affects transfer.

Variability Papers

Creativity appears on our variability continuum under novelty. This is because all creative responses are novel, but not all new ones are creative. To be called creative, the new thing must also be useful or generative or influential. Useful means that it solves a problem; generative, that it leads to other ideas or things; influential, that it changes the way other people do or think about things like it. Our creativity model is based on constraints. We contend that creativity depends on two things: one is expertise which means mastering the constraints that define an existing domain (like painting); two is devising new constraints that will expand the domain. For example, Impressionism is based on Monet’s precluding an existing constraint (differences in value) and promoting a new one (differences in hue).

Our work here is field based. We visit museums, galleries, and studios; read about past painters (architects, composers, etc.) and interview ones currently
at work.

Creativity Papers

High variability facilitates learning. Older children are more functionally variable in problem solving than younger ones. Children of the same age who use more strategies acquire new ones faster than those who use fewer. Compared to their peers, gifted children are highly variable. In the area of a gift (or talent), a gifted child will use more strategies (and better ones) than an average child. In art, a gifted child will acquire skills more rapidly and produce more elaborate drawings and paintings than an average child.

Findings like these led to our interest in developing an early test for giftedness based on high variability. Our hypothesis is this: ease in mastering novel nonverbal tasks with constraints that require high variability may be an early marker of giftedness. Our research in this area is field based. We are currently testing 1st, 3rd, and 5th graders at the Washington School in Lodi, N.J. on computer games that require low or high variability. We will compare between- grade scores to devise age-norms, and then compare an individual child’s score with those of her age-mates and those of older children. Finally, our predictions will be compared with results from current gifted evaluation procedures used in the school.

We are also investigating a puzzling age-related transition in problem solving. Here are the pieces. First-graders have no difficulty with maze games that require repeating a correct path through the maze. Fifth-graders have none with games that require repeating or varying a path to earn points. Third graders have no difficulty with varying to earn points, but do not perform well on mazes that reward repetition. What’s happening in third grade? That’s the puzzle.

Papers on Children