In recent years, researchers have tried to expand Astin's I-E-O model and theory of Involvement. In this study, a comprehensive conceptual framework for the JCIRP survey is presented, using the concepts of Involvement theory from Astin and a synthesis of the two models advanced by Hurtado and Terenzini. This comprehensive conceptual framework has been customized for use in relation to the Japanese higher education system, while at the same time maintaining its comprehensiveness on cognitive and emotional aspects and its utility for empirical research. A feature of the framework is its ability to distinguish 'Student Engagement' from 'Student Involvement', and place engagement at the center of the framework. Student Involvement covers all the elements of involvement, such as Involvement with Faculty and Staff, Involvement with Student Peers, Student Engagement, and Academic Involvement. In the comprehensive conceptual framework, Student Engagement includes the cognitive and emotional aspects of involvement, such as acquiring academic skills, adapting to the uncertainty of student life, and students' feeling of well-being. On the other hand, in relation to the behavioral side of Student Involvement, aspects such as the frequency of class attendance, time allocation and frequency of learning activities, job searching, and so on, are defined as Academic Involvement. Furthermore, Involvement with Faculty and Staff, and Involvement with Student Peers, were defined as the relationships with faculty and staff, and the relationships with student peers, respectively.
An analysis was made of the data from the JJCSS 2009 data set on junior college students who majored in the education field. This data set was obtained from the JCIRP surveys conducted in November 2009 by the Japanese Association for College Accreditation. The sample size was 2,426: 1,242 first-year students; 1,177 second-year students; and 7 students who did not provide answers. Female students comprised 95% of the total. The data set provided no institutional information, such as size, establishment, and location. The procedure and research findings are as follows.
First, as the indicators of learning outcomes, the growth of competencies and knowledge since admission were used. These indicators are retrospective self-reports from students.
Second, using factor analysis, the learning outcomes were divided into three categories: 1) basic special knowledge, 2) contemporary general knowledge, and 3) classical general knowledge.
Third, the determinants of basic special knowledge were examined through multiple regression analysis. The analysis proved that the determinants of Student Involvement were effective, especially that the effects of Student Engagement and Involvement with Peer Group were very effective. Because of the particular research interests, three variables were added to the model. These were 'Sex', 'Reason for college enrollment: I did not want to work after graduating from high school', and 'Time allocation for study and homework'. These were not statistically significant or only a small effect was found.
Finally, to identify the model and assessment of effects on learning outcomes, the SEM (structural equation modeling) was used. As a result, the total effect was very effective for Student Engagement (.338) and involvement with Peer Group (.396). The Involvement with Faculty and Staff was also effective (.299). It is clear that engagement and peer-group relationships are important for the improvement of learning outcomes but that support from Faculty and Staff are also important. As for the variables that are not statistically significant to learning outcomes, the effect of Student Engagement to the time allocated to study and home work, was effective (.155) and statistically significant. It is said that students will spend more time on study and homework if they are fully engaged in academic college life. The 'Reason for college enrollment: I did not want to work after graduating from high school' was not statistically significant, meaning that the lack of willingness is not a serious problem for Junior college students majoring in the education field.
From the analysis of four-year college students surveyed by the JCIRP, 'Time allocation for study and homework' was the second largest factor, although Student Engagement was the largest factor in relation to learning outcomes. A feature of Junior college students was that human relationships with peers, faculty, and staff were very effective on the Student Engagement, Academic Involvement, and learning outcomes. This is not to say that it is not important to enhance the learning environment by the use of various approaches, such as the provision of scholarships to ensure there is adequate time for study without the need to engage in paid employment. The research findings also suggest that strategies to promote relationships with peers, and also support from faculty and staff, are effective for Junior college students majoring in the field of education.