Does Outside-In Teaching Improve the Learning of Object-Oriented Programming?

Author(s): M. Oruc, F. Akal, H. Sever
Venue: 2016 4th International Conference in Software Engineering Research and Innovation (CONISOFT)
Date: 2016

Type of Experiement: Case Study
Class/Experience Level: Undergraduate Student
Participant Selection: undergraduate course enrollment
Data Collection Method: Survey


Teaching Object-Oriented principles in an academic setting has historically followed the same path. Beginners are first taught the basics of sequential programming and then are brought up to speed with object-oriented principles such as classes and objects. Due to the prevalence of object-oriented solutions in the field of software engineering, it is proposed that a more object-oriented-centric curriculum would benefit beginners. The idea of teaching classes and object-oriented design principles first (called Outside-in) in a beginning programming course was thusly proposed and implemented by the researchers in this study.

At two universities in germany, three courses were involved in determining the impact of an Outside-in approach. One course was taught in the traditional manner whereas the other two courses were taught outside-in. The study was built around answering the following three questions: 1. How does the self-assessment and motivation of students taught OOP with the Outside-In method differ from those taught with the “traditional” method? 2. Which learning outcomes does the Outside-In method generate compared to the “traditional” method? And 3. : How and to which extent does Outside-In teaching contribute to an improvement of teaching OOP from the lecturers’ perspective?

The first question was determined by giving two surveys, one at the beginning of the course and one in the middle of the course. The surveys focused on the student’s confidence levels in the subject matter and the course by having them rate their feelings on a scale from 1-5. The most significant difference between the control group and the Outside-in groups was that the latter felt that the course was more structured and manageable after taking the course for several weeks whereas the control group felt more negatively about the course’s manageability.

The second research question was evaluated with a quantitative test that was given at the same time as the second survey. However, due to a large degree in variability in the different universities’ technical levels, the results of the quantitative test are inconclusive.

For the third research question, the four instructors involved in teaching the courses stated that they believed the new approach definitely improved the student’s comprehension of object-oriented principles. However, due to the relatively unorthodox ordering of concepts being taught, constructing lectures was difficult. Gaps in students’ understandings of algorithmic approaches to problems made it difficult to fully convey the benefit of object-oriented design.