Donald Norman- The Design of Everyday Things…

Donald Norman- The Design of Everyday Things

I read chapters from this classic book many years ago, and I remembered Norman’s indictment of the VCR for its hard to read controls and widgets.  The VCR’s design doesn’t make its uses and functions “visible”, its “affordances” are not clear.  One of the most important principles of design, according to Norman, is “visibility”- visibility refers to the mapping between the design of an object and its operation, so that the design communicates cues and feedback about how to use and operate the object.

Mapping is another important idea. The user will have a mental model of how to use the object, and the object also yields a conceptual model (usually a product of the designer’s intentions) for how it can be used. When the two models coincide, then there is a close “mapping”. Norman writes that the designer usually expects the user’s model to follow the designer’s mental model- but this is obviously not always the case. The designer needs to better understand the user, the design itself must be user-centered.

Norman made the notion of “affordances” famous through this book, and though he didn’t invent the term, it seems he popularized it so that it entered the stream of standard design thinking.

The difference from Norman’s discussions in this book and his later book on ‘emotional design’ is in his de-emphasis here on aesthetics. Norman’s user-centered approach here is really more about functionality where aesthetics doesn’t really contribute a benefit. But clearly, Norman does move on to acknowledge the importance of aesthetics in the user’s relationship to the everyday things that support his or her functioning in the world, and also primarily to the processing of information about the world- that emotions play an undeniable role in the cognition and the forming of mental models.

Schnotz & Bannert: Structure Mapping Hypothesis

Schnotz, W., & Bannert, M. (2003). Construction and interference in learning from multiple representation. Learning and Instruction, 13, 141–156.

Summary:

The authors present the “structure mapping hypothesis” which makes a subtler connection (compared to the dual coding hypothesis) between the relationship between visual and verbal representations and their processing through internal mental model constructions. Essentially, the authors are saying that the dual coding hypothesis assumes that there is a univalent relationship between kind of presentation and kind of tasks. But the structure mapping hypothesis is saying that the effects of different pictures will be different depending on the type of tasks. This hypothesis thus makes a distinction between how texts and pictures function. Texts and pictures are 2 different sign systems. The authors state that texts are “descriptive representations”- they describe their objects through relational information. Pictures or physical models are “depictive representations” and they function as iconic signs. Furthermore, descriptions are more powerful in representing forms of subject matter, but depictive are better for drawing inferences. The research that the authors set up seems to back up the authors’ hypothesis. Their results showed that presenting task-appropriate graphics helped learner comprehension, and task-inappropriate graphics interfered with comprehension.

 

My comments:

The authors make a good point for splicing the model more finely. And I’m especially interested in what they have to say about the construction of internal mental representations because it is on the one hand, all speculation- because mental processes lack material form though maybe neural research is getting better at mapping neuronal activity as a material trace. And also because I’m interested in how mental models can be formed through non-visual forms of perception- the authors mention auditive and kinaesthetic:

“Mental models, on the contrary, are not sensorically specific. A mental model of a spatial configuration, for example, can be constructed not only by visual perception, but also by auditive or by kinaesthetic or by haptic perception.” (143).

Schneiderman & Plaisant: Designing the User Interface, chapt. 14

Shneiderman, B., & Plaisant, C. (2005). Designing the User Interface. Chapter 14.5: Information Visualization (pp. 580–603). Boston: Pearson.

Schneiderman & Plaisant (SP) write that it was the success of “direct-manipulation interfaces” that enabled users to work with the computer in a more visual way. And pictures can convey a multitude of meaning and may be the representational mode (rather than text) that is preferred by many for processing information. “Information visualization” thus emerged through computational technology as a term for how we use “interactive visual representations of abstract data to amplify cognition” (SP citing Card, Mackinlay and Schnedierman, 1999).  Interestingly, SP mention that information visualization is an abstraction of reality (590) but, like pure mathematics, the intention is to find the deeper patterns in phenomena, or the mess of relationships that is the real world.

Another great point that SP make is when they state that humans have fine-tuned perceptual abilities that aren’t accessed in most interface designs- and the examples SP give are visual perception related, but I believe we can extend the range of untapped abilities to the other sensory perceptions- touch, proprioception…. 

Anyway, the overview of the language and practice of information-visualization is laid out by SP.  They state the basic principle, or “mantra” of information-visualization to be:

“Overview first, zoom and filter, then details on demand…”

This mantra of information visualization invokes the basic activities or tasks involved. And SP provide a useful TAXONOMY that sorts information visualization then into a set of 7 tasks (overview, zoom, filter, details-on-demand, relate, history and extract) with 7 data types:

1. Linear data: 1-dimensional, including texts, dictionaries, alphabetical lists…

2. 2D map data: maps, plans and layouts with domain and interface attributes…

3. 3D World: include real world objects which have volume and complex relationships to each other, sometimes in the form of dimensional representations such as 3D maps- related to 3D computer graphical imaging & design, virtual reality design… “information-visualization in three dimensions is still controversial” (585)

4. Multidimensional data: Statistical and relational database contents that can be manipulated as multidimensional data…

5. Temporal data: Data that needs to be viewed temporally- as a time series…

6. Tree data:  Hierarchical or tree structured data

7. Network data: When a tree structure is not enough, relationships conveyed through linking as a network…

SP go on to describe the 7 tasks of information visualization in greater detail with examples, and then conclude the chapter with notes about “challenges” to information visualization. The most important one for me is: 1) the challenge of collaboration, that “support for social processes is  critical to information visualization”… a software design challenge and also theoretical challenge

Richard Mayer’s Multimedia Learning Theory

My comments:
While I don’t know if I totally am convinced by the “dual channel assumption”, I agree with the attention paid to sensory modalities in cognition, processing information. But one point about the division between the presentation modes into words and images, I think rather of images as having a heavily visual component, that reading requires visual processing, not just auditory and while Mayer acknowledges this, I think it should receive more attention. Multimedia that takes place on two dimensional surfaces (screens, monitors) is already based on a visual interface. Text that is read on a screen may be accompanied by some external audio and the internal sounding out in one’s mind, but the sensory modality that gets taxed the most, I think, is the visual…This leads to my general argument for incorporating other sensory modalities into the presentation.

Summary of Mayer’s Multimedia Learning:
Mayer delineates three views of “multimedia” for learning. 1) in terms of the medium of delivery, 2) in terms of its presentation modes, its re-presentational formats, and 3) in terms of the sensory modalities used to process the material.

The focus on the first view, the delivery media, is obviously “technology-centered” rather than human-centered (Mayer also brings in reference to Donald Norman’s concern for human-centered design of technology aids to ‘make us smarter’).

The focus of the latter two views is human- or “learner-centered”. Concern for the presentation modes and sensory modalities is based in cognitive theory on how people learn and this is what drives the design. Also, these two views take a constructivist learning ideology, i.e. that knowledge is actively constructed rather than transmitted and passively acquired. Optimal learning is not just knowledge acquisition then, or just a matter of retention, but it is retention with knowledge transfer- the ability to apply the knowledge.

Mayer’s cognitive theory specifically ascribes to a “Dual Channel Assumption” which basically assumes that humans have separate information processing channels for verbal and pictorial information, or (to put it in terms of the sensory modalities) for auditory and visual stimulus.

Mayer also brings up theory of cognitive load- that understanding how humans learn also includes understanding our cognitive limitations or the optimal load for processing information, whether images or text or combinations of both.

With Multimedia Learning theory, Mayer is proposing that multimedia can be used to design material that optimizes learning, and that an optimal condition for learning (information processing) is a combination of auditory and visual, words and images, presented to minimize extraneous cognitive load and maximize learning in terms of retention and transfer of knowledge.