ABSTRACT:The attitude of software learning by students relies on transforming the Learning Model consistently into Software Development Model. The effort and time taken for development of different phases of software, has few issues, in particular, the principles and practices to be followed in academia. To address these issues, a generic software teaching and learning model called eXtreme Software Teaching (XSOFT) is proposed by integrating the essence of the Extreme Programming (XP) method with COSMIC FFP (Common Software Measurement Integration Consortium Full Function Point) Standards. The effectiveness of this proposed model was tested on real time projects assigned to students as a realistic study by selecting them with equal skill and knowledge on programming. They were divided into two groups, such as Pair Programmers and Solo Programmers. The performances of these groups are measured with respect to the trade in person-days using XSOFT. The empirical results show that the assessment of person-days are less on pair programming than solo programming and effective programming skill is found with pair programmers. In addition, the robustness of this model are confirmed by the results of another real time project as case study. ß 2014 Wiley Periodicals, Inc. Comput Appl Eng Educ 23:432-442, 2015; View this article online at wileyonlinelibrary.com/journal/cae;

Learning software engineering in practical laboratory based on present curriculum is under the clutches of hard practices and not well defined. The effort and time taken for planning, designing and coding, have a few issues in student's learning aspect. Same is the case for evaluating the student's program. With the effect, both the teacher and students have leaned towards lightweight learning method using Extreme Programming (XP). Pair Programming, which is one among the twelve practices of XP has been widely used by the pedagogical community. However, this practice is found suitable for introductory level small-scale programs. Also, the researchers incorporating pair programming, have not given much importance for program assignments and program correction. To address these problems, we have developed a Generic software teaching and learning model called GSOFT with few software development practices, and pair programming.COSMIC FFP (Common Software Measurement Integration Consortium Full Function Point) standard was used for program assignments and program evaluation. This method was applied on real time project assigned to student as large scale programs and examined. For this study, the students were grouped into pair programmers (PP) and solo programmers (SP). The performance of the PP and SP groups were measured using our generic model and found the person-days taken to complete the program. The results show that PP used less person-day than the SP. This study also proves that the program developed by PP has better coding.

No abstract

Research in the realm of Pair Programming in computer science students has begun to explore the methods for accomplishing the benefits. However, valid measurements of the opinion on adopting a pair programming in the practical sessions are required to be drawn, before concluding. An Attitude Survey Test (AST) was conducted before implementing Pair Programming as a teaching and learning tool among computer science students in Pondicherry Engineering College. The purpose of this investigation is to examine the student's opinion on four vital issues viz. i) General Perception about pair programming; ii) Personality conflicts among pair; iii) Relevant examination system and iv) Female student's attitude. We have engaged 154 B. Tech and MCA students as participants from Pondicherry Engineering College, for the AST, each answered a set of questions relating to Pair Programming. By accurately assessing the students' opinions, our AST results indicated the adequate convergent validity of scores obtained, that there is a room for implementing pair programming as a teaching and learning tool in laboratory course works.