Developing a computer program to facilitate language training

The present research demonstrates how an instructional program developer can be used as a subject in a series of experiments to determine optimum parameters for a computer program. The research included developing and testing a new training procedure for presentation of English to Spanish and Spanish to English word pairs. The experiments demonstrated that number of word pairs used per lesson (5, 10, or 15 pairs) did not influence how much was learned when lessons were repeated on 3 consecutive days. Constant as opposed to varied orders of pair presentation did not affect the rate at which pairs were learned. Tests with feedback on word pairs 5 and 15 days after training prevented an increase in errors 25 days after training. Repeating lessons on 3 consecutive days as opposed to just 1 day led to fewer errors during tests that occurred 5 or 15 days after training. It did not matter if Spanish to English pairs were presented before or after English to Spanish pairs during a daily training session. In addition, intermixing Spanish to English and English to Spanish pairs did not lead to better performance as compared to presenting English to Spanish or Spanish to English pairs first. The present experiments show how a software developer can be used as the subject in a series of experiments to determine parameters for an educational computer program.