Guidelines
Suggest edit

Use a single-view specification workspace for targeted follow-up questions

For targeted follow-up analysis of tabular data, prefer a single-view specification workspace over a multi-view recommendation gallery to improve focused question answering and address cumbersome scanning for analysts.

  • purpose:select
  • basis:empirical
  • structure:single-view:use
  • structure:multi-view:avoid
  • data:tabular
  • quality:insight:use
  • lever:layout-structure
  • audience:analyst

advice

Manual specification workspace

Choose a single-view specification workspace when the analyst already has a specific follow-up question. For example, let the analyst directly place variables onto encoding shelves and change transformations or mark type on one chosen view instead of scanning a gallery of suggestions.

reason

Why direct specification works for focused questions

Once the question is precise, direct control becomes more valuable than breadth. A single-view workspace lets the analyst act on the intended variables and encodings without the extra step of browsing many alternatives.

Mechanism: Direct manipulation reduces indirection when the analyst already knows what relationship, summary, or transformation to inspect.

Evidence: In the study, participants overwhelmingly preferred the manual specification tool for question answering, and participants described the recommendation browser as good for exploration but cumbersome for specific tasks (https://doi.org/10.1109/TVCG.2015.2467191">Wongsuphasawat et al., 2016).

context

Use when the follow-up question is already formed

  • User Goal: Answer a specific question about the data.
  • Task: Targeted follow-up analysis after an initial overview or after a promising view has been found.
  • Data: Tabular data where the relevant variables are already known or suspected.
  • Chart Setting: Interactive environment that can expose direct controls for encodings and transformations.
  • Audience: Analysts refining one view to answer a precise question.
  • Success Criterion: The analyst can quickly construct and revise one relevant chart.

exceptions

Do not use when the analyst still needs broad coverage

Break it when: The analyst does not yet know which variables to inspect and needs a broad first pass through the dataset. Why: The study found that the multi-view recommendation gallery produced much wider variable coverage and was preferred for exploration.

costs

Tradeoffs of a single-view workspace

Sacrifice: You give up automatic breadth across many variable combinations. Risk: Analysts can fixate early on a small subset of variables. Mitigation: Start with broad exploration or return to it when the current question is answered.

mistakes

Common failure mode: staying in browse mode too long

Mistake: Keep analysts in a recommendation gallery after they can already state the exact follow-up question. Why it fails: Browsing suggestions becomes extra work when the task is to refine one specific view.

check

Check whether the structure should switch

Failure Sign: Users can name the exact variables or relationship they need before they start acting. Quick Check: Compare a single-view specification workspace against a recommendation gallery on the same focused question and see which one reaches a satisfactory chart with less browsing. Stronger Test: Ask users which structure they prefer for the focused task and switch when the single-view workspace is preferred.

fix

Fix the follow-up workflow

  • Add direct controls for assigning variables to visual encodings on one active view.
  • Let analysts change transformations and mark type directly on that view.
  • Move from the gallery into the single-view workspace once a specific follow-up question is formed.

References

Wongsuphasawat, K., Moritz, D., Anand, A., Mackinlay, J., Howe, B., & Heer, J. (2016). Voyager: Exploratory Analysis via Faceted Browsing of Visualization Recommendations. IEEE Transactions on Visualization and Computer Graphics, 22(1), 649–658. https://doi.org/10.1109/TVCG.2015.2467191