In this program coloured cards and stencils are to be placed on one another so as to reproduce a model. All but one of the stencils are symmetric. In the asymmetric stencil a triangle is cut out. This stencil can be turned through 90 degrees thus making the task of reproduction a greater challenge.
The reproduction is done by clicking on the stencils. The chosen one then moves into the assembly zone, where it is either superimposed on the other stencils or, when the "Behind" button is pressed, appears together with the other stencils but just below them. Incorrect reproductions can be removed by clicking into the assembly region, making the highest or the lowest stencil go back to its initial position.
The models are presented on nine levels of difficulty. On the first level only one card and one stencil are needed to reproduce the model. Our experiments have shown that mentally retarded children are able to reproduce correctly first level models. On the ninth level up to nine stencils are superimposed. Even adults can have great difficulty reproducing these models correctly. This broad range of difficulties is a strong point of this program and means that together with the large number of models (5'106) training with one child can last several months or even several years. The program automatically adapts to the learner's level of performance by putting him on a higher level, when he has no difficulty reproducing the models, and putting him on a lower level when he makes many mistakes.
The final product - the correct or incorrect answer - is not as important as the path chosen by the trainee to find the solution. To understand the thought processes all activities of the trainee and the reactions of the program are saved in a protocol. The program recognises 44 types of errors and is able to react to them in an adequate manner. A statistic of all errors is listed. This analysis is process oriented, noting the number of stencil movements enacted, rather than simply the number of solutions attempted.
The program normally gives its feedback in oral language. It is possible to switch to written answers.
It is not recommended to let a child work alone with this program: The triangle between the trainee, the mediator, and the program is important. The mediator must establish a dialogue with the trainee and discuss the characteristics of the stencils and workout the strategies to be employed.