Preparing the Optimization Study
Simulation complements empirical study methods.
-----------------------------------------------------------------------------------
-------------------------------------------------
Simulation vs. Empirical Studies
Simulation complements empirical studies -- like cinematography and electromyography. Simulation uses theoretical models rather than empirical techniques. Instead of physical measurements, simulation involves at most only virtually real representations, like computer displays.
Based on models, simulation is only as good as the assumptions underlying the models, and is limited in that it doesn’t itself obtain direct empirical data about skills. But it does augment empirical studies.
The importance of simulation lies in the fact that it investigates the effectiveness of designs prior to their implementation, and it does so without risk to persons who might be involved in the real environment of the simulation.
Observational techniques such as cinematography and electromyography involve live, active subjects performing specific events under study, and are intended to yield factual information. They quantify movement and thus yield factual information about skills. Cinematography does the job by using cameras to obtain motion pictures of the skilled events, and electromyography, by embedding electrodes into muscle tissue to determine how muscles and muscle groups are involved in the events. For good results you must know your equipment.
Cinematography
measures events on a frame-by-frame basis. This is difficult and tedious work, but sharpens estimates obtained visually. To yield accurate information, you have to minimize picture distortion. Aspects like camera speed, running time, projection angles, distances, camera rotation and translation rates, have to be known to a high degree of accuracy and precision, but with practice can produce valuable empirical data.Electromyography
aims to discover the internal workings of muscles. By measuring electro-chemical potentials of the various muscles -- the neuronal innervation patterns from brain to muscles -- during performance of the events, it is possible to estimate the degree of activity of muscles and the sequence in which the activation occurs.Computer simulation
, on the other hand, is purely a model-building technique. Studies involving simulation draw out the consequences of the models. The models incorporate the mechanisms of the subject’s movement and thus characterize its living dynamics. The model is then subjected to What if...? conditions that might otherwise be imposed on the real subject. Through repeated trials -- each time by altering the starting values of the variables and/or by changing the values of the parameters that bound the dynamics or circumscribe the action -- simulation replaces the real performances. By observing how the models behave, you can deduce how their real counterparts would act under similar operating conditions.
------------------------------------------------
In simulation, physical, physiological, and psychological aspects of the movements might be considered. Both movement and perception might be modeled. The person you wish to study might have a physical impediment, for instance, or could be seeing- or hearing-impaired. Otherwise the person might be suffering from psychological problems. Either of these personal conditions could significantly affect the person's perception or movement. Using a dynamic model of the subject, you can learn which kinds of motions are worth developing for the subject and which are not. Also, the models can be altered at will, though not necessarily successfully, new ideas implemented and old ones eliminated, and the models can readily incorporate empirical data generated from other sources. Simulation thus provides a flexible and productive way to learn about skills.
Unfortunately, what is needed for full-fledged visualization is the processing power of a supercomputer. Even with limited technology, though, selected studies (which are essentially part-task studies) can be conducted, so long as you simplify the model requirements and make the best use of available equipment and program development tools. For one thing, it isn’t necessary to use fully realistic figures to conduct meaningful investigations.
---------------------------------------------
The ultimate goal of any study of skills is to obtain information about the skill. The aim is to learn to make an intelligent choice among possible action alternatives -- like playing the stock market, taking a course in statistics, shooting jumpshots, driving a truck, running a company, or keeping track of a terrorist. You would apply the programmed version of the situation model and run many trials to get the information.
A complete study of an individual's behavior could involve the perception component of the action in addition to the motion component, as well as their interaction. For instance, the tennis server might be given the ability to "see" the motion of the tennis ball as he tosses it in the air. Or a gymnast might be able to observe his progress on the high bar. Similarly, a player of the stock market would be able to "see" the price action of individual stocks as she takes the appropriate buy or sell action.
The simulation could represent both how the person went about doing the job (motor) and how the person went about getting appropriate information to do it (perception).
The model would spell out the structure and dynamics of the objective content of the person's activity and how the person actually perceives that activity. This is what I did for my thesis study of the tennis player tracking and intercepting a ball. Using such a model, the investigator would set parameter values to define the situation and then vary the starting values of the action variables to look for best or better end conditions on the actions.
----------------------------------------------