SBIR-STTR Award

Web-based training is ready to come of age and is in need of controlled, scientific studies to ascertain which design elements are crucial to optimal learning. We intend to build web-based modules that train trainers/tutors how to teach beginning reading skills. The modules will apply to all readers, but they will be especially helpful for those dealing with struggling readers. We propose a 2X2 between-subjects study with four web-based training conditions containig varying degrees of inter activity and dynamic assessment. The conditions are: 1) typical static web-site, 2) web-site with automatic scoring of constructed answers, 3) web-site with hands-on demonstrations, and 4) a fully dynamic web-site with both scoring of constructed answers and hands on demonstrations. Anticipated Result We anticipate that the fully dynamic web-based condition that includes both automatic scoring of constructed answers and hands-on demonstrations will result in the greatest learning. The innovation of using an automatic essay grader will require trainers to put more cognitive effort into their answers. In addition, because we believe that knowledge is embodied, the hands-on demonstrations and simulations in the module will utilize multimodal memory systems which will increase the trainers' comprehension and retention ofmateriaIs from the web-site. We will publish these results in peer-reviewed journals and presenthem at relevant conferences. We will apply for Phase II funding to build a full set of seven reading-related training modules to be used by many different parties interested in training reading, these include (but are not limited to): * trainers who fall under the aegis of governmental agencies like Head Start and America Reads, teachers and trainers affiliated with Reading First. * teachers in public and private school systems, as well as parents who are home schooling. * corporate and private volunteers who wish to tutor readers. (If the organization is a non-profit volunteer organization, NeuronFarm will allow them use of the modules for free, otherwise modules will rent on a monthly or yearly basis.) Two major benefits should result from this study. First, in1portant design elements for the burgeoning web-based training industry will have been empirically tested. Second, access to quality training modules may positively impact the outcomes of thousands of struggling readers

The purpose of this project is to enhance the quality of literacy tutor training and formulate best practices in e-learning design. Research indicates that in the year 2000, 37 percent of fourth graders were below the “basic” level of skill in reading. Some studies have shown that tutoring can be an effective means for improving children’s learning, and that students with well-trained tutors learn more than students with untrained tutors. The researchers in this project are designing and implementing eight tutor-training e-courses based on empirically-tested principles of learning and current knowledge in reading education. They are also embedding experiments in three of the eight e-courses to study specific features of student learning. In one experiment, they are comparing the learning of students who have to manipulate colored icons on the screen in a practice task to that of students who simply have to choose an answer. In a second experiment, the researchers are studying the effects of having students construct answers to pretest questions on how well they do in a later test. In the third experiment, the researchers are testing the hypothesis that students who are warned about a pop quiz learn more than students who are not warned. The research team is putting the web-based learning courses into a digital library called CyberSilo to make the training more widely available. The goal of this project is to develop effective methods of learning in distance education, particularly in the area of cost-effective training of tutors