SIGCSE 2003 DC Application -- Gary E. Schubert



Gary E. Schubert, MS, MFA, Associate Professor of Art and Computer Science
Alderson-Broaddus College, Philippi, West Virginia

West Virginia University Doctoral Candidate in Technology Education

Research Title: "CS1 CS2 Considered Harmful"

Research Summary

Introduction
 I am interested in finding better ways to recruit, retain, and educate undergraduate computer science majors. Over the 15 years I have been teaching computer science classes, I have experimented with many ideas and approaches. I think that the "pair programming" methodology has great promise in undergraduate computer science education in addressing all of these issues. Prior research shows Pair Programming to be effective in producing higher quality software products in upper level software engineering classes. I am attempting to develop a research protocol to determine the effect that pair programming might have on student success in CS1 and CS2 and therefore on the "yield" from CS1 & CS2. My presentation is based on my "discoveries" to this point.

·        

·       Theoretical background

·       see http://www.pairprogramming.com/

·        

·       Previous research in the area

·       see http://www.pairprogramming.com/  <http://www.pairprogramming.com/>

·        

·       Goals of the research

·       To find better ways of educating undergraduate computer science majors. Is "pair programming" a better way? Specifically "Can the use of pair programming increase the "yield" in CS1 & CS2?".

·        

·       Current status

·       I am doing preliminary informal work on developing an appropriate and defensible experimental design by using Pair Programming in CS1 & CS2 courses I am teaching at West Virginia University, in the CSEE Department. This is to determine what the issues and variables are, so that they can be controlled for in the experimental design of a formal study to follow.

·        

·       Interim conclusions

·       I am finding that there are many "first-order" issues in the students' successful completion of CS1 & CS2 which completely overshadow the use or non-use of "pair programming". Hence the title of my research summary.

·        

·       Open issues

o      What are the "first-order" issues with regard to students' successful completion of CS1 & CS2?

o       

o      What are the "second-order" issues with regard to studying "pair programming" in CS1 & CS2?

§       What kind of  experiment can produce valid and reproducible results in this environment?

o      Do traditional freshmen and sophomore pre-CS and computer science majors have enough "maturity" to be able to use the pair programming methodology to work with and learn CS1 & CS2 course content?

o       

o      How do you handle the "non-team" or poor team players?

o       

o      How do you handle the missing team member or odd-man out situation?

o       

·       Current stage in your program of study

·       I have 6 hours of required program classes plus 6 hours of research, comps, dissertation and defense left. I am taking the 6 hours of required classes this spring 2003 semester. I anticipate that it will take at least one more year to complete the doctorate and graduate in May of 2004. This is assuming that I can determine a reasonable final research topic from this preliminary study. I am working full-time plus overload and going to graduate school part-time. This dilutes my effort to complete the doctorate quickly.

·        

·       What you hope to gain from participating in the Doctoral Consortium

·       Contact with faculty and graduate students interested in my area of research who can help with ideas, solutions, feedback and critiques.

·        

·       Bibliographic references

·       see http://www.pairprogramming.com/

·       see http://www.extremeprogramming.org/

·