This content originally appeared on HackerNoon and was authored by Pair Programming AI Agent
Table of Links
3. Experiment Design and Conduct
3.2 Subjects, Tasks and Objects
4.2 Analysis of Variance (ANOVA)
4.4 Effect Size and Power Analysis
5. Experiment Limitations and 5.1 Threats to the Conclusion Validity
5.2 Threats to Internal Validity
5.3 Threats to Construct Validity
5.4 Threats to External Validity
6. Discussion and 6.1 Duration
7. Conclusions and Further Work, and References
3.2 Subjects, Tasks and Objects
Junior students enrolled in the DOE course participated as subjects in the experiment; in total, for this experiment there were 21 subjects. Most of the subjects were in their third year of the program's degree in SE; the rest of them (three subjects) were in their four year. According of Dreyfus and Dreyfus programming expertise classification [12], we categorized subjects as advanced beginners; subjects have working knowledge of key aspects of Java programming practice.
\ Subjects were randomized and allocated into two groups: Pair and solo programmers. The experiment was split into two sessions, where in each session subjects wrote a different program. In both sessions we employed the same subjects, so we collected 14 measures with respect to solo programmers (7 solos per session) and 14 measures regarding pair programmers (7 pairs per session). In the first session, subjects that worked individually used NetBeans IDE (as tool support) to write the first program, whereas subjects that worked in pairs used only a text editor. In the second session the tool support was changed, so subjects that before worked individually with the NetBeans IDE, in the second session they worked with a text editor and conversely (See Latin square design arrangement in Table 2).
\ Before the experiment was conducted, we gave a talk to the students about pair programing. In this talk we explained the main concepts of this programming practice and how it can be used in practice. We also explained how to compile a Java program using only a text editor. Finally, we explained to students how to collect the measures during the experiment sessions. The collection procedure consisted in writing down the time duration that students spent writing a program. They recorded the start and finish time and computed the difference (in minutes).
\ We selected to small programs that subjects could write, compile, run and test in each session. In the first program (identified as calculator) we asked the subjects to write a calculator that evaluates expressions with decimal numbers, and the operators: Plus (+), minus (-), times (!), divide (/), and prints the result on the screen. In the second program (identified as encoder) we asked the subjects to write a simple encoding-decoding program. Given a specified letter switch the program must be able to encode or decode a line of text.
\
:::info Authors:
(1) Omar S. Gómez, full time professor of Software Engineering at Mathematics Faculty of the Autonomous University of Yucatan (UADY);
(2) José L. Batún, full time professor of Statistics at Mathematics Faculty of the Autonomous University of Yucatan (UADY);
(3) Raúl A. Aguilar, Faculty of Mathematics, Autonomous University of Yucatan Merida, Yucatan 97119, Mexico.
:::
:::info This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license.
:::
\
This content originally appeared on HackerNoon and was authored by Pair Programming AI Agent
Pair Programming AI Agent | Sciencx (2025-08-19T01:50:19+00:00) The Role of Tool Support in a Pair Programming Study. Retrieved from https://www.scien.cx/2025/08/19/the-role-of-tool-support-in-a-pair-programming-study/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.