Scientific terms in context of software dev: Empirical, Hypothesis, Objective etc.

While my interest is in practical software development and not computer science per se, in technical discussions sometimes scientific terms come in. Note that I am a software technologist but did study Physics and am a Physics graduate.

This post captures an exchange with Google Search AI (GSAI) on these terms.

Me: Give me crisp and clear definitions of empirical, hypothesis and (scientific) theory in the context of software development.

GSAI: Here are the definitions of empirical, hypothesis, and theory tailored for the software development context. 

1. Empirical (Evidence-Based Software Engineering) 

Definition: Decisions, processes, or validation methods based on direct, measurable observation and real-world experience rather than theory, opinion, or intuition. 

• Context: Using actual project data (e.g., git logs, JIRA metrics, user engagement, A/B test results) to validate if a tool, methodology, or code change actually improves productivity or quality.
• Example: Running a controlled experiment comparing team velocity when using Pair Programming versus Individual Development. 

2. Hypothesis (Product/Technical Hypothesis) 

Definition: A testable, tentative prediction about the outcome of a change, usually formulated as "If we change

$X$

(Independent Variable), then

$Y$

will occur (Dependent Variable/Metric)". 

• Context: Used in hypothesis-driven development or A/B testing to validate assumptions about a new feature or technical refactoring before full commitment.
• Example: "If we implement lazy loading on the image gallery, then average page load time will decrease by 20%, resulting in a 5% higher conversion rate". 

3. (Scientific) Theory in Software Development 

Definition: A well-tested, widely accepted, and rigorously verified framework that explains why a particular software engineering practice or phenomenon works across different contexts. 

• Context: Unlike a simple rule of thumb, a theory is built upon numerous empirical studies and provides predictive power about software behavior.
• Example: "The theory of balancing human, social, and organizational capital" for software development, which explains how to manage team capabilities based on project demands. 

---

Summary Comparison Table 

Term	Context in Software	Focus	Source
Empirical	Evidence-driven	What did we see?	Observation & Metrics
Hypothesis	Assumption validation	What do we expect?	Prediction & Testing
Theory	Model explanation	Why does it happen?	Tested Generalization

Me: Give me crisp and clear definitions of subjective and objective in the context of software development.

GSAI: 1. Objective (Data-Driven / Factual)

• Definition: Requirements, metrics, or decisions that are based on hard facts, quantifiable data, or technical limitations. They can be verified independently of personal opinion.
• Context: What the software must do, or measurable performance metrics (e.g., response time).
• Examples:
  • "The system shall display the total price including a 5% tax."
  • "API response time must be under 200ms."
  • "The application must support PostgreSQL 14".
• Key Traits: Verifiable, Measurable, Consistent, Non-biased. 

2. Subjective (Perspective-Driven / Opinionated)

• Definition: Requirements or feedback based on personal feelings, aesthetic tastes, user experience interpretations, or preferences.
• Context: How the user feels when using the software, or qualitative improvements (e.g., usability, design).
• Examples:
  • "The interface feels slow." (Subjective) vs. "The page takes 5 seconds to load." (Objective).
  • "The button should be more appealing."
  • "The new dashboard is confusing."
• Key Traits: Interpretable, Variable, Qualitative, Emotional. 

Key Differences Summary

Feature	Objective	Subjective
Foundation	Facts, Evidence, Technical Spec	Opinions, Experience, Feeling
Measurement	Quantifiable (metrics)	Qualitative (user sentiment)
Verification	Pass/Fail (Binary)	Subject to Interpretation
Usage	Requirements, Testing, Backend	User Interface, UX, Design

Key Takeaway: Good software development often requires taking subjective user desires (e.g., "I want it to be fast") and converting them into objective engineering requirements (e.g., "Load time must be <1s").

In software development, distinguishing between objective and subjective elements is critical for scoping, user experience, and testing.