Cracking the Code: Unveiling the Traits of a Great Software Engineer - Paper Review

Last updated Apr 1, 2024 Published Jun 27, 2023

The content here is under the Attribution 4.0 International (CC BY 4.0) license

What Makes a Great Software Engineer? May 2015 - Conference: 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering (ICSE) [1] - available online version - last accessed: Jul 19, 2023.

Introduction

The literature is based on novices and students starting their careers rather than experienced software engineers
ACM curricula say software engineers need technical knowledge and project management.

The contribution of the paper is the understanding of software engineering expertise as viewed by expert software engineers.

---
title: Mindmap based on the content of the paper
---

mindmap
  root((Great software engineer))
      Internal attributes
        Personal
          Improving
          Passionate
          Open-minded
          Data-driven
        Decision making
          knowledge about people and organization
          Update mental model
          Handles complexity
      External attributes
          Team mates
            Creates shared context
            Creates shared success
            Creates safety
            Honest
          Software product
            Attentive to details
            Creative
            Antecipate needs

Related research is extensive, but a few aim at directly software engineers’ expertise.

Insights in software engineering came from OOPSLA 2003, which discussed the need for design analysis, embracing diversity, multidisciplinary project teaming, large-scale development and quality code.
Dijkastra: obvious and elegant code, constructed with provable correctness.
Experts are more productive, systematic, and well-prepared.
Research suggests qualitative and environmental differences.
Setting, communication and alignment of goals are important.
Effective conflict management.
Experts know code ownership, the technical domain, and argumentation skills

Empirical

ideally sample wide range of software engineers.
Face to face semi structured interview - about 1 hour.
Licensure and accreditation for software engineers is still uncommon.
Authors based their selection on social “people having achieved some degree of recognition as software engineers”.
At or above Level 2 - those are experts by other engineers via hiring promotion process.
Random sampled engineers (round-robin fashion).
Out of 152 contacted engineers, 59 was interviewed.

Findings

53 attributes (internal and external)
internal: personality and ability to make decisions
external: focus on emotional intelligence and decision making models to their software, their team mates, stakeholders and potentially millions of users

Model of attributes

Personal characteristics

Improving
- Not great when started but learned to continuous learning - technology change fast
- This is also connected to what is dicussed in the course “Learning how to learn”
Passionate
- Intrinsically interested in the area they are in, not just looking for rewards like money
Open-minded
- Let new information to change how they think
- It is rare to have one engineering that have a complete understanding of they system
Data-driven
- Creates behavioral feedback
- Decisions should be made using data

Decision making

Knowledgeable about people and the organization
Mental model: synthesizing the current context, decision alternatives, probabilistic outcomes and values of outcomes
Handles complexity

Teammates

Creates shared context
Creates shared success
- Managers have a broader view, engineers have a detailed view
Creates a safe heaven
A place to improve from mistakes without negative consequences
- pains from mistakes are the best teachers
Not appreciate wasting time shifiting the blame for problems

Software product

Elegant
- Simple and intuitive designs
- Avoid complexity
Creative
Anticipate needs

Treats to validity

Treats to internal validity
- informants could produce more attributes after the interview
Treats to external validity
- small sample
- results might not generalize to other cultures
- organization size might also affect the results

Discussion

Productivity is one criterion for excellence
Software engineering is a socio-technical undertaking, not just a technical one
Simply delivering code is insufficient
For researchers
- not aware of tools that help engineers be more well-mannered in emails
- evaluate tread-offs
For novice engineers
- can use the attributes to aspire to be great engineers
- better present them to their employees
For managers
- help senior engineers to improve mentoring, raising challenges and walking-the-walk
- avoid not well-mannered engineers
For educators
- results rise questions about curriculum
- decision making is not part of the ACM computing curricula
- explicitly discuss what students will not learn in school/university