Formative and summative evaluations are both used to assess software systems, but they occur at different stages and serve different purposes. Formative evaluation happens during development. It is used to provide ongoing feedback that helps improve the system before it is completed. This might involve early user feedback on prototypes, informal testing sessions, or feedback loops in agile sprints. It is typically iterative, helping developers refine features, fix usability issues, and adjust design decisions based on user needs. In contrast, summative evaluation takes place after development is finished. Its goal is to determine whether the system meets all objectives and is ready for deployment. This includes assessing the system against formal evaluation criteria such as correctness, efficiency, usability, maintainability, robustness, and scalability. Summative evaluation relies on final testing results, structured user feedback, and documented performance metrics. Both types of evaluation are essential: formative improves the system during creation, and summative validates it at the end.

Evaluation provides a clear understanding of how well a system performs and where improvements are needed. By using defined evaluation criteria, developers can pinpoint which aspects of the system fall short and categorise issues based on severity and impact. For example, if evaluation reveals that usability is poor due to unclear navigation, this might be prioritised above less urgent updates like design enhancements. Evaluation can also uncover non-critical bugs, performance bottlenecks, or maintainability concerns that don’t require immediate fixes but should be addressed in future releases. Additionally, if scalability tests show the system struggles with increasing data loads, developers might prioritise architectural changes to prepare for growth. Feedback from users often identifies high-impact problems or frequently requested features, which can be used to guide update roadmaps. Ultimately, evaluation transforms subjective opinions into actionable data, helping teams make evidence-based decisions on what to fix, improve, or add next.

Automated tools play a significant role in streamlining and enhancing the evaluation process. They provide consistent, repeatable testing that reduces human error and saves time. Performance profilers can analyse a program’s execution, identifying functions that consume excessive time or memory, aiding in the assessment of efficiency. Static analysis tools check code quality, enforcing style guidelines and detecting maintainability issues like unused variables, complex methods, or poor naming conventions. Automated test frameworks (such as JUnit or pytest) enable developers to run unit, integration, and system tests regularly, quickly identifying regressions or logic errors. Load testing tools like JMeter or Locust simulate high user traffic to assess scalability under stress. Additionally, usability testing software can track mouse movement, clicks, and session length to analyse how users interact with the system. These tools complement manual methods and ensure that evaluation is comprehensive, precise, and grounded in data.

Documentation plays a crucial role during the evaluation phase by providing a clear record of what the system is supposed to do, how it was built, and how it has been tested. Firstly, documentation such as the requirements specification allows evaluators to determine whether the system’s functionality matches the original objectives. Without this reference, it’s difficult to judge correctness or completeness. Secondly, technical documentation—including system architecture diagrams, API references, and data flow models—supports the assessment of maintainability and scalability. Well-documented code is easier to understand, modify, and extend, which is essential for future development. Additionally, test documentation, such as test plans and result logs, proves that evaluation has been thorough and systematic. User manuals and interface guides are also reviewed for clarity and accessibility, contributing to the system’s overall usability score. In professional settings, thorough documentation is often a legal or contractual requirement, reinforcing the importance of detailed, accurate records during evaluation.

While technical evaluation focuses on functionality and performance, a thorough evaluation should also consider ethical and social implications. These include issues such as user privacy, data security, accessibility, and potential misuse. For example, a system that collects personal data must be evaluated for how it handles and stores that data, ensuring it complies with regulations like the UK GDPR. Accessibility evaluation ensures that the system can be used by people with disabilities, following standards such as the Web Content Accessibility Guidelines (WCAG). Social impact considerations might include assessing whether the system inadvertently reinforces bias, excludes certain user groups, or enables harmful behaviours. For example, a recommendation algorithm should be evaluated for fairness and transparency, especially if it affects users’ choices or opportunities. Ethical evaluation often involves stakeholder consultation, risk analysis, and reviewing usage scenarios to identify unintended consequences. Including these concerns in the evaluation process promotes responsible software development and long-term user trust.