PRAXA: A Framework for What-If Analysis

TL;DR

Praxa addresses the lack of a unified framework for what-if analysis by synthesizing 141 visual analytics and HCI publications (2014–2024) into a cohesive vocabulary and structure. By defining why (motivations), what (dataset and model as building blocks), and how (user and system operations), Praxa yields four analysis types: SENSITIVITY, GOAL SEEK, IMPORTANCE, and SCENARIO COMPARISON. Two case studies demonstrate how counterfactuals can be interpreted as GOAL SEEK or SCENARIO COMPARISON and how dimensionality reduction and parameter-space exploration fit into the framework. The framework enables standardized implementation, reproducibility, and the design of flexible, domain-agnostic tools and interfaces, with open research directions in trust, explainability, provenance, and mixed-initiative interaction.

Abstract

Various analytical techniques-such as scenario modeling, sensitivity analysis, perturbation-based analysis, counterfactual analysis, and parameter space analysis-are used across domains to explore hypothetical scenarios, examine input-output relationships, and identify pathways to desired results. Although termed differently, these methods share common concepts and methods, suggesting unification under what-if analysis. Yet a unified framework to define motivations, core components, and its distinct types is lacking. To address this gap, we reviewed 141 publications from leading visual analytics and HCI venues (2014-2024). Our analysis (1) outlines the motivations for what-if analysis, (2) introduces Praxa, a structured framework that identifies its fundamental components and characterizes its distinct types, and (3) highlights challenges associated with the application and implementation. Together, our findings establish a standardized vocabulary and structural understanding, enabling more consistent use across domains and communicate with greater conceptual clarity. Finally, we identify open research problems and future directions to advance what-if analysis.

Figures (10)

• Figure 1: Methodology of the paper selection, codebook development, and coding and analysis followed for the literature review.
• Figure 2: Example of selected schema and codes of the Wexler et al. wexler2019if paper. The colors illustrate how various schema properties captured contribute to the different building blocks of Praxa: dataset, model, motivations, user_operations, and system_operations.
• Figure 3: The components of the Praxa framework for what-if analysis are organized around three key dimensions identified through our literature review: why, what, and how. The why dimension captures the underlying motivations for performing what-if analysis. The what dimension defines its core building blocks: the dataset and the model. The how dimension describes the operational mechanisms through which what-if analysis is performed, comprising both user_operations and system_operations that act upon the what components. The operations marked with * represent more general data analysis operations that, while not exclusive to what-if analysis, are frequently observed and play a key role in enabling it.
• Figure 4: Combining different components of what-if analysis to form its different TYPES.
• Figure 5: Examples of input interactions for SENSITIVITY analysis type from reviewed literature.