Gerência e Qualidade de Software - Aula 10 - Métricas

Summary

Professor Fábio Siqueira's lecture delves into the realm of software quality metrics, emphasizing the importance of quantifying subjective attributes like user interface aesthetics. He discusses the necessity of objective measurements for assessing software quality, the challenges in evaluating certain characteristics, and the utility of metrics in project management and process improvement. The talk covers various types of metrics, including product metrics, and highlights the Function Points method as a way to measure software size and functionality, despite its empirical nature and potential for subjectivity.

Takeaways

• 😀 Software quality metrics are essential for quantitatively evaluating the aesthetic and user satisfaction of software interfaces.
• 📐 Metrics provide a numerical way to assess whether software has achieved the desired quality attributes, which is crucial for discussions and improvements.
• 🔍 The challenge with quality attributes like aesthetics is that they are not always easy to quantify, yet metrics are necessary for objective measurement.
• 📈 Metrics are used to measure the product, process, and project in software development, serving different purposes such as quality evaluation and defect prediction.
• 🛠️ A good metric should be easy to calculate, intuitive, unambiguous, consistent, and not dependent on the programming language used.
• 📊 Metrics are indicators that may not always have a direct correlation with the expected outcomes but are used to aid decision-making.
• 🔑 Traditionally, metrics are divided into three types: product metrics, project metrics, and process metrics, with the focus of this script being on product metrics.
• 🔄 Product metrics are used to measure the quality characteristics of the software and can be challenging to measure directly, often requiring indirect measurements.
• 🔢 The number of lines of code is a widely used but controversial metric due to its simplicity and potential misinterpretation of productivity.
• 📚 Function points are a method to measure the functional size of software, allowing for estimation of development time and cost without actual implementation.
• 🔄 Function points consider both data functions and transaction functions, with different levels of complexity affecting the calculation of points.
• 🔄 Adjusted function points take into account 14 characteristics of the software, allowing for a more nuanced measurement that can vary by up to 35%.

Q & A

What is the main topic discussed by Professor Fábio Siqueira in the video?

The main topic discussed is software metrics, specifically how they are used to evaluate and measure the quality characteristics of software.

Why is it important to quantify the quality characteristics of software?

Quantifying quality characteristics is important because it provides a measurable way to assess whether the software meets the expected quality standards, helps identify potential issues, and aids in project management and process control.

What are the problems associated with evaluating quality characteristics without metrics?

Without metrics, it's difficult to objectively evaluate quality characteristics, leading to potential disagreements and lack of a clear direction for improvements.

What are the types of metrics mentioned in the script?

The script mentions three types of metrics: product metrics, project metrics, and process metrics.

What is the purpose of product metrics as described in the video?

Product metrics are used to measure the quality characteristics of a software product, also known as prediction metrics, to foresee if the software will possess certain qualities.

What are the characteristics of a good metric according to the video?

A good metric should be easy to calculate, intuitive, unambiguous, consistent in measurement units, not dependent on programming language, and useful for decision-making.

Why are static metrics important in software development?

Static metrics are important because they can be analyzed without executing the software, providing insights into the software's structure and potential issues before the software is run.

Can you provide an example of a static metric mentioned in the script?

An example of a static metric mentioned is the number of lines of code, which is easy to measure and has been widely used in software engineering.

What is the Function Points method and why is it significant?

The Function Points method is a way to measure the functional size of software, allowing for the estimation of development time and cost before the implementation phase.

What are the two main components used to calculate Function Points?

The two main components are Data Functions, which represent data managed by the software, and Transactional Functions, which represent elementary processes within the software.

What are the criticisms of using the Function Points method?

Criticisms include its empirical nature, which can introduce subjectivity into the measurement process, the difficulty in establishing a real relationship with other metrics like lines of code, and its dependence on the company's historical data for accurate calculation.

What is the adjusted Function Points method and how does it differ from the basic Function Points method?

The adjusted Function Points method considers 14 general system characteristics, each rated on a scale from 0 to 5, and adjusts the basic Function Points calculation accordingly, accounting for factors such as distributed data processing, performance, and ease of operation.