💻 computer science

Corporate Training in Brazilian Software Engineering: A Quantitative Study of Professional Perceptions

This quantitative study of 282 Brazilian software engineers identifies cognitive engagement, activity variety, and instructor performance as the primary drivers of training effectiveness, while highlighting that mandatory participation negatively impacts motivation and perceived quality, thereby validating the applicability of general training frameworks like Salas and Cannon-Bowers' to the software engineering domain.

Original authors: Rodrigo Siqueira, Antonio Oliveira, Breno Alves Andrade, Lidiane C. S. Gomes, Danilo Monteiro Ribeiro

Published 2026-04-08

📖 5 min read🧠 Deep dive

CC0 1.0

Original authors: Rodrigo Siqueira, Antonio Oliveira, Breno Alves Andrade, Lidiane C. S. Gomes, Danilo Monteiro Ribeiro

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are the captain of a massive spaceship (a software company). Your crew consists of highly skilled engineers who keep the ship running. But the universe is changing fast; new stars appear, and old navigation maps become useless overnight. To keep the ship moving, you need to constantly train your crew on new tech.

This paper is like a report card from the crew members themselves, asking: "What actually makes a training session feel useful and good, and what makes it feel like a waste of time?"

The researchers surveyed 282 software engineers in Brazil to find the answer. Here is the breakdown in simple terms:

1. The "Secret Sauce" of Good Training

The study found that three specific ingredients make or break a training session. Think of these as the Three Pillars of a Great Class:

The Coach (Instructor Performance): It doesn't matter how fancy the textbook is if the teacher is boring or confusing. The crew needs a coach who explains things clearly, is enthusiastic, and actually helps them understand.
- Analogy: It's the difference between a tour guide who just reads a script and one who tells funny stories, points out hidden gems, and makes you feel excited to explore.
The Menu (Variety of Activities): If you only eat steak every day, you get bored. Training needs a mix of things: lectures, hands-on practice, group discussions, and maybe even some games.
- Analogy: A good training session is like a buffet, not a single plate of plain rice. You need different flavors to keep your brain engaged.
The "Aha!" Moment (Cognitive Engagement): The training needs to make the engineers feel like they are actually solving problems, not just memorizing facts. They need to feel their brains working.
- Analogy: It's the difference between watching someone else juggle (passive) and actually trying to juggle yourself (active). You learn more when you are the one doing the work.

The Big Takeaway: If you want your engineers to love training, don't just focus on the content. Focus on who is teaching, how they teach, and getting the students to think.

2. The "Mandatory" Trap

The study looked at a very common situation: Forced Training vs. Voluntary Training.

Voluntary: "Hey, check out this cool new course on AI! Want to sign up?"
Mandatory: "You must attend this meeting on Friday, or you're in trouble."

The Result: When training is forced, the crew's motivation tanks. They feel like they are wasting their personal time, and they think the training is less relevant, even if the content is actually good.

Analogy: Imagine you love pizza. If your mom says, "You can have pizza whenever you want," you enjoy it. If she says, "You must eat this pizza for breakfast every day for a month," you will start hating pizza. The pizza didn't change; the choice did.

The study found that forcing people to train makes them feel like their personal time is being stolen, which kills the learning experience.

3. The "Time" Myth

One interesting finding was about personal time.
Many managers think, "If we just schedule training during work hours so it doesn't eat into their evening, they will love it."

The study says: Not necessarily.
Even if the training happens during work hours, if the instructor is boring or the activities are dull, the engineers will still feel like it was a waste of time.

Analogy: If you are stuck in a boring meeting that lasts 2 hours, it feels like an eternity whether it happens at 9 AM or 5 PM. Fixing the schedule doesn't fix the boredom. You have to fix the quality of the meeting first.

4. Do We Need to Reinvent the Wheel?

Software engineering is often seen as a unique, high-tech field. You might think, "We need special, futuristic rules for training our coders."

The study says: No.
The factors that make training work for software engineers are the exact same factors that make training work for doctors, teachers, or salespeople.

Analogy: You don't need a special engine for a race car that runs on a different type of physics than a regular car. The principles of how an engine works (fuel, spark, movement) are the same. Software companies can use the same "training rulebook" that other industries have used for years.

Summary for the Boss

If you are a manager trying to improve your team's training:

Hire great teachers (or train your current ones to be better).
Mix up the activities (don't just lecture for 4 hours).
Make it about solving real problems.
Don't force it if you can help it. Let them choose, or at least explain why it matters so they feel like they are choosing it too.
Stop worrying about the clock. A boring 1-hour meeting is worse than a great 2-hour meeting. Focus on quality, not just scheduling.

The researchers built a survey tool to measure this, and while it's a bit of a "draft" (needs more testing), it's a great starting point for any company wanting to stop wasting money on bad training.

1. Problem Statement

Despite significant investments in corporate training within the software industry, there is a critical knowledge gap regarding the specific factors that drive the perceived quality and effectiveness of these programs from the perspective of Software Engineering (SE) professionals.

Context: SE is characterized by rapid technological obsolescence and a structural gap between academic curricula and industry demands, making continuous training a prerequisite for productivity.
The Gap: Existing literature lacks validated instruments to assess training effectiveness specifically within the SE domain. Most research focuses on organizational policies rather than the participant's experience, and there is limited empirical evidence on how the nature of participation (mandatory vs. voluntary) impacts engagement in technical fields.
Objective: To quantitatively analyze which factors influence SE professionals' perceptions of training quality and effectiveness and to determine if general training frameworks apply to the software sector.

2. Methodology

The study employed a quantitative, cross-sectional survey design targeting Brazilian software engineering professionals.

Participants:
- Sample Size: 282 valid responses.
- Demographics: Primarily male (76.6%), average age ~34, with 49.7% holding undergraduate degrees and 38.7% post-graduate degrees. Most worked in companies with 100+ employees.
- Inclusion Criteria: Professionals working in SE with higher education in ICT who had participated in at least one corporate training in the last 12 months.
Instrument Design:
- Framework: The study operationalized Salas and Cannon-Bowers' training framework, covering four dimensions: (1) Training Needs Analysis, (2) Antecedent Conditions, (3) Training Methods/Strategies, and (4) Post-Training Conditions.
- Development Process: A four-stage iterative process reduced an initial pool of 106 items (from 20 literature sources) to a final instrument of 27 Likert-scale items (5-point scale) and 9 sociodemographic questions.
- Pilot Testing: Conducted with 10 professionals to refine clarity (e.g., replacing "training" with "learning experience").
Data Analysis:
- Statistical Approach: Polychoric correlation was used to analyze associations between ordinal scale items, as it is the recommended method for estimating relationships between latent variables underlying ordinal responses.
- Reference Variable: Item Q21 ("Overall satisfaction") was treated as the proxy for perceived training effectiveness to identify principal predictors.

3. Key Contributions

Identification of Core Predictors: The study identifies a tightly correlated "core" of three factors that drive perceived training quality in SE: Cognitive Engagement (Problem-solving reasoning), Variety of Activities, and Instructor Performance.
Evidence on Mandatory Training: Provides quantitative evidence that mandatory participation significantly reduces motivation, perceived relevance, and perceived quality, while amplifying the perception of time burden.
Cross-Domain Validation: Demonstrates that established training design principles from general organizational psychology (specifically Salas' framework) are applicable to the software industry, suggesting organizations do not need to "reinvent" training principles for SE.
Operationalized Instrument: Offers a new, open-access survey instrument based on Salas' framework, tailored for the SE context, serving as a candidate for future psychometric validation.

4. Key Results

Primary Predictors of Quality:
- Problem-solving reasoning (Q22): Strongest correlation with overall satisfaction ( $\rho = 0.806$ ).
- Variety of activities (Q19): Second strongest ( $\rho = 0.804$ ).
- Instructor performance (Q18): Third strongest ( $\rho = 0.785$ ).
- Interpretation: These factors align with Bandura's Social Learning Theory (attention, retention, reproduction, motivation), suggesting that effective SE training relies on clear instruction, cognitive engagement, and diverse learning stimuli.
Impact of Mandatory Participation:
- Participation driven by obligation (Q11) showed strong negative correlations with motivation ( $\rho = -0.546$ ), overall satisfaction ( $\rho = -0.523$ ), and perceived relevance ( $\rho = -0.508$ ).
- Mandatory training also correlated with a higher perceived burden on personal time.
Independence of Time Burden:
- The impact on personal time (Q6) showed no significant correlation with overall training quality ( $\rho = -0.094$ ). This implies that simply reducing the time cost of training without improving instructional quality will not increase perceived effectiveness.
Framework Applicability:
- The Salas framework produced coherent, interpretable results across all four dimensions in the SE context, validating its use as an analytical lens for software engineering training.

5. Significance and Implications

For Software Organizations:
- Prioritization: When resources are limited, organizations should prioritize instructor quality, cognitive engagement (problem-solving focus), and activity variety over other factors.
- Policy Adjustment: Organizations should minimize mandatory training where possible or frame it strategically to preserve autonomy, as coercion significantly degrades learning outcomes.
- Design Strategy: Improving scheduling (time management) alone is insufficient; the focus must remain on instructional design and content relevance.
For Research:
- The study bridges the gap between general HR/organizational psychology and Software Engineering, confirming that human learning mechanisms are consistent across domains.
- It highlights the need for future psychometric validation (reliability analysis, factor analysis) of the proposed instrument to establish it as a standard metric for the industry.

6. Limitations

Sampling Bias: Reliance on convenience sampling (LinkedIn, WhatsApp) may over-represent professionals already engaged with training.
Geographic Scope: Results are specific to Brazil and may not generalize to other cultural or economic contexts.
Self-Reported Data: The study relies on subjective perceptions rather than objective metrics (e.g., code quality, productivity).
Instrument Status: The survey is currently exploratory; it has not yet undergone full reliability or confirmatory factor analysis.

Conclusion

The study concludes that software engineering training effectiveness is not unique to the domain but follows established psychological principles. The most critical levers for success are instructor performance, cognitive engagement, and activity variety, while mandatory participation acts as a significant deterrent to learning motivation and perceived value.