Social Data Science Scorecard Deck

The Social Data Science Scorecard Deck is an agile project management tool addressing core individual aspects of a social data science project.

The Deck helps social data science teams to:

Detect bottlenecks and reasons for project delays or other problems.
Ensure balanced development of a social data science project.
Track decisions and their provenance.
Structure project documentation and progress reporting.
Effectively distribute responsibilities.
Plan the next steps to take within a project.
Organise organisational practices and ways of working.
In the long term, reuse the emerged practices and streamline social data science.

The Social Data Science Scorecard Deck is inspired by the SEMAT Kernel model. The SEMAT Kernel model captures the Essence of Software Engineering. It identifies the three main areas of concern for a software engineering project: Customer, Solution, and Endeavour. Within each area, it specifies the core aspects (alphas) that are inherent to every software engineering project (see Figure 1).

^{The Figure 1. SEMAT Essence of Software Engineering Diagram (source: Essence Standard
v1.1).}

Underpinning of the Social Data Science Scorecard Deck is the observation that the areas of concern of a data science project overlap with those in software engineering. The Deck captures the core aspects of a social data science project by extending the original Essence model with new and updated areas of concern and alphas. The Essence of Social Data Science Diagram (see Figure 2) represents their relationships. Each Alpha is further elaborated using decks of alpha scorecards.

^{Figure 2. The Essence of Social Data Science Diagram.}

- Area of Concern 1: Demand

The Demand area of concern captures the stakeholders’ need for running a social data science project.

The Research Goal alpha represents what, in principle, needs to be achieved by the research to satisfy the stakeholders and why it is relevant and feasible. The goal informs the questions that the research aims to answer.

The Stakeholders alpha represents the potential beneficiaries of the research, the users of the project artefacts and other parties affected by the project. Achievement of the research goal is in their interest as is the proper and ethical conduct of the research. Stakeholders may support the project team through funding, guidance, participation in research activities and other means. The team needs the stakeholder involvement to make sure that the research actually progresses towards its goal.

- Area of Concern 2: Solution

The Solution area of concern captures the response to the demand in a social data science project.

The Research Questions alpha represents the questions that the research aims to answer. The bulk of the work on a social data science project is done to answer these questions, so they guide the work from start to finish.

The Artefacts alpha represents the outputs of the project that embed the answers to the research questions. Their types are numerous: from reports and publications to software systems and reusable components. The artefacts are designed to satisfy the research goals.

- Area of Concern 3: Analytics

The Analytics area of concern captures the research required to answer the questions and to inform the artefacts.

The Data alpha represents all the data that are available for the project. Data may come from both external and internal data sources and may be produced by the project interim work.

The Research Methods alpha represents the approaches that are applied to the data to answer the research questions. They may vary from statistical methods and data visualisation techniques to complex machine learning methods. By answering the research questions, the methods inform the artefacts.

- Area of Concern 4: Resources

The Resources area of concern captures the resources necessary for the research. Due to their scarcity, the resources put a natural constraint on the work done for the project.

The Data Sources alpha represents the platforms that provide data for the project. If a project does not use the external data sources, many of the technical concerns related to this alpha are automatically resolved. Yet it is not advised to exclude this alpha from consideration as methodological concerns may still prevail. comment: We may need to discuss this one. It is not as clearly formulated as it should be.

The Infrastructure alpha represents the software, hardware and data structures employed in a social data science project. The infrastructure is used to access the data sources, to acquire and store the research data and to execute the research methods.

The Compliance alpha represents the available ethical and legal clearances and permissions. Without compliance in place, it may be deemed unethical and even illegal to access certain data sources (for example, due to licensing restrictions) and store and analyse data from them (for example, due to privacy concerns).

- Area of Concern 5: Endeavour

The Endeavour area of concern captures the actual process of conducting the research.

The Work alpha represents all the activities that need to be carried out for the project. The work aims to satisfy the research goal.

The Team alpha represents the people who are actively engaged in running the project and executing its individual tasks. They plan the work and execute it through their usual ways of working. The team aims to produce the project artefacts.

The Ways of Working alpha represents the principles and practices of organising the work that the team adheres to. They guide the work and ultimately allow it to get done.

The Deck

Further following the steps of the SEMAT Kernel model, the Data Science Scorecard Deck helps to effectively ensure balanced development of a social data science projects. It contains a scorecard for each core aspect (alpha) of a data science project. A card lists the possible states of an alpha. The alphas should go through their consecutive states through the course of the project.

The Deck also has separate cards for each individual state of an alpha. Such a card lists the conditions that should be met by an alpha to achieve this state. Figures 3a and 3b present example scorecards for the Data alpha and for its Operationalised state.

^{Figures 3a and 3b. The “Data” Alpha Scorecard (front and back) and the “Data: Operationalised” State
Scorecard (front and back).}

- Using the Deck

The Social Data Science Scorecard Deck can be used in multiple environments.

During the project team meetings, the Deck can be used to track the progress of all the critical aspects of a social data science project and to identify the aspects that fall behind the others. It can also be used to plan further activities and their timespans. Furthermore, the scorecards can serve as a map to distribute responsibilities among the individual team members.

The Deck can be used in the routine work of individual team member. A team member can use the alphas, states and conditions that they are responsible for as a convenient structure for documentation of their work and for their progress reporting. The project managers / team leaders can use the scorecards for continuous assessment of the project state.

The Deck can be used outside of the scope of a particular project as a way to classify and organise best practices and ways of working that a team / an organisation has developed and to store pointers to relevant guidance.

Depending on the complexity of a project, the Deck can be used in the context of the project as a whole, of a particular branch of the project and of a particular iteration.