Chapter 3: Task Abstraction Flashcards
What is the trade-off between having your user ‘design’ a visualization themselves versus having a visualization designer pre-design visualizations for the user?
A visualization tool can be narrow, where the designer has built many choices into the design of the tool itself in a way that the user cannot override. These tools are limited in the kinds of data tasks that they can address, but their strength is that users are not faced with an overwhelming array of design choices.
Tools can also be more general, where users have many choices to make. The breadth of choices is both a strength and a limitation: users have a lot of power, but they also may make ineffective choices if they do not have a deep understanding of many visualization design issues.
Why analyze task abstractly?
Transforming task descriptions from domain-specific language into abstract form allows you to reason about similarities and differences between them. Otherwise, it’s hard to make useful comparisons between domain situations. The same visualization tool might be usable for many different goals. Another reason to analyze the task is to understand whether and how to transform the user’s original data into different forms by deriving new data. The task abstraction should guide the data abstraction.
Which three levels of user actions are there that define user goals?
The high level choices describe how the vis is being used to analyze, where information is either consumed or new information is produced.
The mid-level choices cover what kind of search is involved.
The low-level choices pertain to the kind of query: does the user need to identify one target, compare some targets, or summarize all of the targets? The choices at each of these three levels are independent from each other.
Which three goals are associated with the analyze user action in terms of consuming data?
- Discover: find new knowledge that was previously unknown. The goal can either be to generate a new hypothesis based on the new information or to verify/disconfirm an existing hypothesis.
- Present: refers to the use of visualization for succinct communication of information. A crucial aspect of presentation is that the knowledge communicated is specific, in addition to that is already known to the presenter in advance.
- Enjoy: This goal refers to casual encounters with vis. In these contexts, the user is not driven by previously pressing need to very or generate a hypothesis but by curiosity that might be stimulated and satisfied by the vis.
Analyze can mean consuming data, as well as producing data. Which three user goals are associated with Producing data?
- Annotate: refers to the addition of graphical or textual annotations associated with one or more preexisting visualization elements.
- Record: refers to saving or capturing visualization elements as persistent artifacts, like screenshots. In contrast to annotate, an annotation does not have to persist, and annotations can be recorded.
- Derive: here the goal is to produce new data elements based on existing data elements.
Which mid-level goals are associated with the search user action?
All of the high-level analyze cases require the user to search for elements of interests within the vis as a mid-level goal. Four classification can be defined for search:
Lookup: If users already know both what they’re looking for and where it is.
Locate: to find a known target at an unknown location.
Browse: when users don’t know exactly what they’re looking for, but they do have a location in mind of where to look for it.
Explore: When users are not sure of the exact target or location.
Which low-level goals are associated with the query user action?
Once a target or set of targets for a search has been found, a low level user goal is to query these targets at one of three scopes:
Identify: The scope is a single target. If a search returns known targets, then identify returns their characteristics.
Compare: scope is multiple targets. The difference in characteristics is returned.
Summarize: the scope is all possible targets. Providing a comprehensive view of the characteristics of all targets.
User action refer to some target, meaning some aspect of the data that is of interest to the user. Which high level targets are relevant for many kinds of data?
- A trend is a high-level characterization of a pattern in the data.
- Outliers are data points that do not fit well with the backdrop of the trend.
- A feature is a structure of interest in the data, which is task dependent.
What type of relation can multiple attributes have with each other?
- A first attribute can have a dependency on a second
- There is a correlation between one attribute and another if there is a tendency for the values of the second to be tied to those of the first.
- The similarity between two attributes can be defined as a quantitative measurement calculated on all of their values, allowing comparison between them.
