A critical look at the validity of studies that focus on humans versus machines teaching grammer

Abstract

Almost from birth, the procession to learning language begins from responding to noises in the environment to learning new information throughout the lifespan. Yet, not every child learns how to speak properly. These children can benefit from tutoring. Whether discussing concepts with the teacher or studying with a classmate, tutoring has taken many forms and has proven to be successful with many populations of students. Some areas where tutoring has been effective include teaching social skills and giving academic help in subjects like mathematics and reading. Tutoring has been conducted with many populations of students, including those with developmental disabilities, learning disabilities, low ability, and those who are disinterested. Likewise, students in a variety of settings such as the inner-city schools or colleges have benefited from tutoring. This study looks the construct, statistical conclusion, internal, and external validity of research conducted on teaching grammar to children using computers or humans. With these advantages to using computers and the increased use of technology in tutoring, more analysis of studies using computers needs to be performed. Specifically, few analyses have evaluated using computers to teach language, especially grammar. Therefore, a meta-analysis looking at the effectiveness of computer-assisted and human tutoring on teaching language needs to be done. One of the first steps for this to occur is to look at the quality of studies that focus on humans versus machines teaching language.