Contribution: We re-think the ‘Statistical Analysis’ curriculum building upon system engineering tools where assumptions (e.g., ABET criteria and student profiles) are carefully assessed, a learn-by-mistake approach ensures that several of the main statistical mistakes are learned, and advanced topics are proposed to make a strong connection with forthcoming courses in data science.
Background: Today’s data science requires students and prospective data scientists to have a strong foundational background in statistical analysis methods and decision making. Given the diversity of students' profiles, and a multitude of statistical analysis curricula across the USA, we seek to provide guidelines on a curriculum that is in line with today’s data demanding era.
Intended outcomes: The target audience comprises students in engineering courses that deal with data and seek to obtain a domain-specific technological or societal solution. Using a learn-by-mistake approach, we try to mend some of the most common mistakes in statistical analysis for the new generations of data professionals. The proposed curriculum equips students with multiple statistical methodologies that enable them to understand, process, extract, visualize, and communicate statistical evidence.
Application design: We propose a systems engineering approach to design the curriculum that leverages tools and methodologies from operations research and statistics.
Findings: Our approach ensures that the designed ‘Statistical Analysis’ course satisfies some of the intended constraints and goals by design. In particular, we designed an overarching hands-on example that integrates the topics covered in the curriculum into a transversal example and can be further customized to the different students’ profiles.