Helmut Schmidt has provided six programs (written in C++) for testing retroPK. They are available here by anonymous FTP. Our eventual aim is to convert these to Java applets which can be launched directly within the WWW environment (work is in progress) - in this way, we will be able to remotely monitor subjects' progress, as is explained in the project description. These programs are being made available now so that potential subjects can practice or at least get some idea as to what is involved. We are providing both DOS-executable binaries and the C++ source code from which they are compiled.
The binaries should be transferred into your home directory as swing.exe, move.exe, etc. and can then be run by simply typing "swing", "move", etc. at the DOS prompt. Unfortunately for those Mac-users amongst you, we have been unable to locate a Mac compiler. If anyone is able to compile the source code for Mac execution, please get in touch. Similarly, if anyone feels that they could convert the source code into Java with relatively little effort, this would be a tremendous help.
The program SWING involves a horizontal pendulum, the aim being to "mentally enforce" a wide (or narrow) swing. CRASH and SLALOM both involve a scrolling sine-wave of varying amplitude, somewhat like the path of a slalom-skier, with variations. MOVE involves a simple block, moving left and right. COLOR and SPARKS are generally intended for a less active, more "meditative" approach to RPK. The former involves a stationary disc of gradually shifting colour, and the latter a pyrotechnic simulation. In all of these programs, the movements/shifts which occur are guided by prerecorded random numbers. Therefore, the sceptic would say, you will simply be watching a predetermined animation, and no amount of "psychic effort" will influence it. The extensive research documented elsewhere at this site suggests otherwise. We recommend that you try all of the programs, and let us know which you prefer. It seems that SWING has been the most successful for Schmidt and his subjects, so we shall probably introduce this first into the formal experiments.
The instructions which appear at the beginning of each program refer to "explore-runs" and "record-runs". Ignore this, as it refers to an application which Schmidt initially wrote the programs for (something very similar to our project, but much slower, as it involved snail-mailing of discs, etc. rather than immediate WWW-based involvement and monitoring). There is also reference made to associated .txt files, and a file called README which explains each of the programs, how the scoring works, appropriate psychological attitudes, etc. in Schmidt's own words. For now, we'll just say that positive and negative scores should be equally likely, and a rescaling is used so that one standard deviation corresponds to a score of 100. Non-statisticians needn't worry about this - simply aim for high positive scores (or high negative scores if you find this easier). Very little can be deduced about your rPK abilities from a few runs (roughly one minute each, although speed parameters can be set by the user). In order to separate genuine anomalies from chance fluctuation, quite a few runs are needed. Schmidt has come to the conclusion that for best results, one shouldn't attempt more than five runs per day (obviously there are exceptional subjects). Our current experiment design will involve subjects "logging in" whenever they feel capable, and gradually completing a total of 40 runs over a period of perhaps 2-3 weeks.
To download executable programs or source code, click on the following targets. All files are delivered in "zip" compressed archives which may be extracted with PKZIP) or a compatible program. The source code requires the Borland C++ to compile, since the programs use the BGI graphics package.