It is essential for proper implementation of computers in education to understand certain characteristics of computers, and to understand how children develop cognitively. Every person can imagine an appropriate age for beginning to learn how to drive cars. Knowing these tools (cars), and also the general characteristics of children, certainly nobody would say that they should learn to drive at age 7 or even 10. One expects that a driver have a certain degree of responsibility, maturity and motor coordination to drive in traffic.

In the case of computers, age is not an obvious factor, because their operation does not produce physical disasters and requires very limited physical coordination, and then only when entering data (Monke and Setzer, 2000). This indicates that an understanding of the tool (computer) is necessary and can not be fully grasped before an approximate ideal age not only to learn about how they work, but an ideal age at which to begin using them.

The third argument I present to support my aim is that, computers work with an extremely restricted class of children’s thoughts, thoughts that do not have the same meaning to the machine that they represent to a child (to the machine they have no meaning at all). Computers process data, which consist of specific kinds of thoughts introduced into them. Through structural association one may couple two pieces of data, but what both mean and their relation to the real world, cannot be inserted into the machine (Bower, 1988).

This symbolic manipulation of data characterizes computers as abstract machines, as mathematical machines. In fact, it is possible to describe with mathematics all data processing done by a computer. Programming a computer corresponds to elaborating purely mathematical thoughts. It is a process analogous to theorem proving. Although it is not so obvious, this is also the case when one uses any software, as for example a word processor.

To align a text vertically, one has to give the machine a command, punching some keys on the keyboard or selecting an icon with the ‘mouse’. This activity is formal, always causing the same reaction by the machine. To execute some task through such a series of commands, one has to exercise exactly the same type of reasoning used in algebraic mathematics. To solve an algebraic equation, for example, one must work formally, logically, step-by-step, through a set of operations predefined by the algebraic system (Monke & Setzer, 2000).

The only substantive things that set the computer user and the programmer apart from the mathematician are the type of symbolic language utilized in each case and the user’s ability to get immediate feedback as the machine responds to each command (Bronfenbrenner, 1995). The restrictive environment of computer use, whether exhibited through programming or the menu commands used in word processors, spreadsheets, and so on, are examples of what one calls formal language, a language with a strict syntax, which may be fully described in mathematical terms (Papert, 1980).

The type of thinking necessary to program a computer or to use any software through written or iconic commands is of the same nature as the one used when doing symbolic logic. The computer’s absolute demand that the student reduce decision by stringing together exact and unambiguous expressions requires that the child operate in a cognitive straightjacket. Software applications and programming that make this straightjacket more comfortable by varying degrees, does not mean that we can ignore their constricting effects on a child’s cognitive processes (Bowers, 1988).

In addition to the preceding argument, it is advanced that, how computers are used in education is detrimental to children’s development. The various uses of computers in education may be classified into three broad categories. One form of using computers in educating children is “programmed instruction” introduced conceptually by B. F. Skinner in the early 1950s (Skinner, 1986). The computer presents a subject, often using sound and animation “unconditioned stimulus”.

After this phase (sometimes in the midst of it), questions are posed to the student, and the answers lead to other topics of investigation or the repetition of previous ones that were not properly “conditioned”. This is a classic example of conditioning. In programmed instruction, the computer forces the same type of thinking as in any other application, because the commands given by the students also constitute a formal language, and the computer reacts always according to a rigid mathematical formula, based on nothing more personal than the students’ previous responses.

Learning is here reduced to memorization and the capacity for solving problems directly related to the covered material; the program cannot take into account the level of maturity, creativity, and intuitive abilities of different users (Bower, 1988). Another form of computer use in education is simulating experiments. Instead of observing and doing something real, either in a laboratory or in the field, students explore simulations on the computer screen. C. A. Bowers (1988) pointed out a number of cultural problems created by trying to reduce problem solving to mere data analysis.

One aspect of this tendency is that the simulation, which is based on sophisticated mathematical models hidden from the user’s view, gives the illusion of conforming to the real world, when in actuality it only conforms to the very limited contingencies anticipated by the programmer. It fosters a mechanical view of nature just as a political simulation fosters a mechanical, rational view of social relationships, also available to manipulation and control (Bower, 1988 and Talbot, 1995).