User Experience, Usability and The Theory of Affordance in Modelling Technology-enabled Processes

Download the full article in pdf here

At the supermarket today, it was almost deserted, except for a few customers using the self-service checkout machines, all seemed to be in a rush to get out as quickly as possible to avoid prolonged exposure to human contact due to the coronavirus scare. They were all taken, the checkout machines, except for one at the far end, hidden around the bend. The young shop assistant told me to go around and use it. As I tried—in autopilot, and probably in a hurry to get out of there fast, like the rest of us—to place the bag in the shopping loading area, it wouldn’t let me. I quickly realized that the loading area was on the left side of the machine and the basket stand on the right, unlike the rest of the machines in that checkout area at the supermarket that I have frequented for many years as a regular customer. A glitch in the process flow, where fast checkout is one of the reasons for the introduction of these machines—besides cutting back on staff. Whatever the reasons, it is not the point of this article. 

Process and human performance

A process is essentially about human performance. Whether technology-enabled or technology-agnostic, without a human operator or actor to actuate the process, the process is no more than static sets of tasks waiting to be performed most likely by a human actor, or by a machine or both. Without considering the individual, physical (time and space), psychological, and environmental factors affecting performance, a process is bound to fail to enable its users to complete or achieve the required tasks. Consequently, it is human-process interactions that need to be focused on in designing and modelling processes, and not just how process activities are sequenced and distributed. Overlooking this critical aspect of process performance, as often seems to be the case, is one of the primary reasons for process operators to deviate from the official version of the process and work around the constraints and obstacles a process design might present. When the process and people’s ways of working are at variance, dissonance between the process logicality and people’s habits of mind is bound to occur. While adaptation to new processes is expected, and is a part of training, skilling and socialization, it does not always succeed if the process design ignores the human factor and focuses on the mechanical side of the process. Process modellers and designers often make assumptions about the intended users and the tasks these users are supposed to perform. These assumptions, which are sometimes captured under assumptions and dependencies, are rarely based on proper and thorough task analysis and situational user information needs analysis. Unfortunately, in quasi agile and pseudo agile development environments, process design is guided by a set of heuristics, assumptions and rules of thumb. In draconian project management contexts, these are imposed on the process designers by the manager or someone assuming superior knowledge based on their flimsy certifications. It is a well-known side effect of this approach is that it leads to cognitive biases about the process users.    

Going back to the self-service checkout example, from a process design, usability and user experience perspective, swapping the orientation of one single machine disturbs the pattern and disrupts the subconscious operations of users. Note that the customer traffic going into the self-service checkout area is from left to right. The customers enter from the left, head to the next available self-service checkout machine and exit from the right. The process flow of the machine should pragmatically match the customer traffic process flow. Now, imagine mapping both processes (the traffic flow and checkout).

As the diagram shows, the last checkout machine goes backward (perhaps to complete a clockwise flow continuing from the last machine on the other side of the checkout bay).

The Self-service Checkout Bay

Human performance is defined as “the accomplishment of a task or series of task by a person or team of people.” (Dolgov, Kaltenbach, Khalaf S., & Toup, 2017). Such tasks may be performed as standalone or in a sequence of some sort that may or may not form an end-to-end process.

A process is composed of a series of interconnected activities and sub-activities and tasks. To complete a process, these activities must be completed. How these activities are performed and what the factors facilitating or prohibiting the accomplishment of these activities are will determine whether a process has achieved its goal and purpose. These factors are summed up in Bailey’s (1982) model of human performance, which consists of four parameters: (1) accuracy, (2) efficiency (time), (3) proficiency (skill mastery and acquisition), and (4) (self) satisfaction.  

Adapted from Bailey’s (1982) Human Performance Model

Frame of reference and frame of expectation

Two things that I learned a long time ago in my work in usability and design are still valid today: a consistent frame of reference and a consistent frame of expectation. The first frame acts as an anchor of information, process activities, screens, buttons, and so on and defines the cognitive map that determines the user’s interaction with the design. For example, similar screens have similar positions and shapes of buttons. In consecutive steps that require user intervention, changing the positions of buttons arbitrarily from one form to the other within the same application and across upgrades weakens the frame of reference. Within the immediate frame of reference, presenting screen components inconsistently not only confuses the user, but it also renders the procedure inefficient because it disturbs the regularity of the pattern that makes it possible for us to understand the information and knowledge it represents.

The second frame, frame of expectation, which builds on the first one, builds up the expectations of the user as to what comes next, and develops the subconscious, autopilot cognitive processes. A frame of expectation enables the users to anticipate what will happen next (without a dramatic or plot twist, and without guesswork).

When something like this design at the supermarket is encountered, it disrupts these cognitive processes and renders the process inefficient. The arbitrariness of the design diminishes the effectiveness of the design and frustrates the user. A frame of reference is defined as “a set of ideas or facts that a person accepts and that influences the person’s behaviour, opinions, or decisions”. Beyond this dictionary definition, a frame of reference projects a cognitive structure onto the design, thus facilitating the user’s interaction with the design and defining the relationships of objects of the design.  

In user-centred designs, the frame of reference relies on the user’s intuition and subconscious procedural knowledge of the world. For example, a house door has three possibilities to open it: you push it in or pull it out or up, or slide it left or right, or right to left. These possibilities (push, pull, slide) form the frame of reference to which the user refers. When the user has to bring the subconscious procedural knowledge to the conscious to figure out how to open an unlabelled, handless door, the design has failed the test of efficiency of minimum effort (cognitive and physical) and maximum return (fulfilling the objective. In this case, opening the door). This faulty design is known as the Norman Door after Don Norman who observed this phenomenon.[i]     

Affordance

From the perspective of the theory of affordance, a design that does not provide the user with the function it is intended for has already failed. For example, a table design with four legs cannot remain stable with three legs, unless you redesign the table to achieve balance and weight distribution on three legs. A car design to run with four wheels will not function with three wheels. You may change the design and end up with a moped. But the original design itself, with a missing wheel, will fail to afford the user the intended function or use.[i]  

In the past, standardization was a key factor in the design of appliances, user interfaces, processes and procedures, etc. Today, we are more likely to encounter disruptive designs, but not in the sense of innovation and breaking established patterns of thinking and doing. These disruptive designs are often designer-focused, ignoring the user, as we have seen it in the design of mobile phones. A consistent frame of reference formalizes and standardizes activities into easily recognized patterns. In process design, very rarely is task and activity analysis performed as part of designing a process. Often activities are identified in a string or sequence towards an outcome without much thought about how the user is going to perform them. In most situations, the designer has a preconceived idea of how the process operator is going to perform each activity and task.  

Where the eye drops the hand follows

A design that does not consider the frame of reference and the frame of expectation is potentially a faulty design. Designs rely on the largely subconscious procedural knowledge that enables people to function. Procedural knowledge appeals to the ontic properties of the objects of reality. A principle of effective design that obeys the laws of nature is summed up in the statement “where the eye drops, the hand follows”. This is exemplified in martial arts, in boxing, and in human-machine interactions.

Checkout User Journey

The 2-minute rule

Going back to process design, for a user to be able to use a process, the process must be obvious and simple, and the user must be able to learn how to use it in just a few minutes. In process mapping and modelling, the different states of the process must be visible to the user, limited in number and visibly distinguishable as the user navigates through the process. When a process doesn’t allow the user to fulfil the required tasks, the process must be re-examined, and the states re-evaluated to improve the user experience. If the user cannot follow the process with ease, it is the process and not the user that is at fault. When the user complains, the user is always right. As a golden rule, a process map that takes longer than two minutes to navigate through is a faulty map. One is willing to go as far as saying modelling a process is in effect modelling human performance. Submitting an application, shopping online, paying a bill, and so on, are tasks or activities that human beings perform. A process organizes how these activities are performed based on a clear understanding of how human beings behave and the logicality of the sequence of tasks within a spatial-temporal context.

? 2020 Ali Darwish. All rights reserved.

References

Cuevas, H. M., Velazquez, J., & Dattel, A. (. (2017). Human Factors in Practice: Concepts and Applications. London: Taylor & Francis Ltd.

Darwish, A. (2011). Business Process Mapping: A Guide to Best Practice. Melbourne: Writescope Publishers.

Darwish, A. (2011). Knowledge Management: Myth and Reality. Melbourne: Writescope Publishers.

Dolgov, I., Kaltenbach, E., Khalaf S., A., & Toup, Z. O. (2017). Measuring Human Performance in the Field. In H. M. Cuevas, J. Velazquez, & A. (. Dattel, Human Factors in Practice:. London: Taylor & Francis Ltd.

Grenander, U. (1996). Elements of Pattern Theory. Baltimore: Johns Hopkins University Press.

Norman, D. (2013). The Design of Everyday Things. New York: Basic Books.

Rubin, J. (1994). Handbook of Usability Testing. New York: John Wiley and Sons Inc.

Notes

[i] See Norman, Don (2013). The Design of Everyday Things.

[ii] See Darwish, Ali (2019). Simplification https://www.youtube.com/watch?time_continue=24&v=foHIgmdpEWA&feature=emb_logo&fbclid=IwAR394yxSL8SeygnO1zDhBGk2iVaUzuhTzhabvdNyE_bKCx8MAf9MVpCEEx0