Iceberg Sensemaking: A Process Model for Critical Data Analysis

Area of Science:

• Data Science
• Information Science
• Human-Computer Interaction

Background:

• Existing data sensemaking models often rely on positivist assumptions.
• A critical, humanistic perspective is needed for a more nuanced understanding of data analysis.
• Interpretivism offers a valuable lens for reframing data sensemaking.

Purpose of the Study:

• To propose a novel, interpretivist model for the data sensemaking process.
• To reframe data analysis and visualization through a critical, humanistic lens.
• To provide a structured, three-phase model for understanding how analysts make sense of data.

Main Methods:

• Developed a three-phase process model (Add, Check, Refine) using an iceberg analogy.
• Incorporated the concepts of tacit and explicit schemas throughout the sensemaking process.
• Validated the model through four distinct data analysis scenarios.

Main Results:

• The proposed model emphasizes schemas as fundamental to data interpretation, treating data as a schematic artifact.
• It highlights the importance of tacit knowledge, power dynamics, and schematic multiplicity in sensemaking.
• The model effectively addresses scenarios like identifying missing data, data wrangling, and handling inconvenient data.

Conclusions:

• The interpretivist approach fosters epistemic humility and promotes pluralism in data sensemaking.
• The new model offers a more comprehensive framework for understanding the complexities of data analysis.
• This approach enhances the descriptive and prescriptive power of sensemaking models.