Wednesday, January 16, 2013

Designing NexGen Products: Bending the Users' Learning Curve

Where did my Parking Brake Pedal Go?

Say, you seat yourself in your rental car, settle down, fire the engine, and lift your leg to press & release the parking brake. Oops! Your left foot instead pedals "air"! Where did the parking brake pedal go? This shouldn't come as a surprise to many of us, who may drive a car with a parking brake that is foot-operated (Figure 1) on a regular basis and suddenly come across a car where it is hand-operated (Figure 2).

Figure 1: Foot-Operated Parking Brake

Figure 2: Hand-Operated Parking Brake
The muscle & mental memory of the sequence and procedure to drive a car is so strongly encoded, this stereotypical action-patterns -- "foot searching for the pedal" -- just gets executed without a thought.

The above example highlights the inconsistency in the control (foot pedal vs. hand lever) and user-response(foot vs. hand) that is provided in different car models to perform the same function (apply/release parking brake).  If one has never driven or come across a car with a hand brake, that person may spend the next hour in the parking lot figuring out how to release the parking brake. On a more serious note, if the parking brake which doubles-up as an emergency brake can't be operated in an emergency due to the above described inconsistency, it could make the difference between life and death.

Creative Disruptions: NexGen Technologies

The above type of inconsistency in design is not limited to various car manufacturers, but also arise when a product manufacturer chooses to upgrade a technology or platform from one generation to another (a.k.a., "NexGen"). The product design team has the best of intentions to provide the most value for their customers but they risk alienating them when the end-users loath the idea of having to re-learn the user-interface all over again. In a professional setting, such as healthcare, command & control, aviation, etc., this also incurs a large cost in training. Some cases are illustrated in Table 1.

TABLE 1: Previous vs. "NexGen" Technology (for that particular epoch)
Horse-drawn Buggies to Automobiles

Sail to Steam to Diesel to Nuclear

Needle-Dial-Gauge (Conventional Instruments) to Glass Cockpits (digital)
Manual flight control to fly by wire

Command-line Interface (DOS) to Graphical User Interface (Point/Click/Manipulate; Touch Screens/select/manipulate) and their various versions. (e.g., most recently Windows 8 for PC's)

Invasive surgery to minimally-invasive surgery (e.g., Laparascopic surgery)
Also to be noted is the impact on the users in terms of the learning effort and overall benefits that might be derived from "NexGen" technologies. 

The creative disruptions brought about by technology have produced confusions, errors, and catastrophic accidents from nuclear power plant control rooms to ships. In the recent past, it has been conjectured that the sinking of the Titanic was caused by the difference in the way the steering operated in sailing versus steam ships (Tiller order vs. Rudder order):

Tiller [sailing ships] and Rudder orders [steam ships] were the complete opposite of one another: the command to turn "hard a-starboard," for example, meant to turn the wheel right under one system and left under the other. 
As First Officer William Murdoch spotted the iceberg, his "hard a-starboard" order was wrongly interpreted by the Quartermaster Robert Hitchins, who was trained under the more archaic Tiller orders. 
He turned the ship right instead of left, exactly toward the iceberg.

Most recently hackles were raised when Microsoft introduced Windows 8, with its "split-personality," which attempted to incorporate the best of both worlds (layering traditional desktop for mouse point & click with "tiles" optimized for touch screens.). It has been reported that the steep learning curve it imposed took a toll  (Video 1) on typical computer users and has adversely affected its sales.

Video 1: Windows 8 Fails the Dad Test


  1. Will knowledge of English help someone learn Chinese?
  2. Can a skilled table tennis player make a good court tennis player?
  3. Can a talented mathematician become a good programmer?
  4. How quickly can a Windows 7 user master the newly released Window 8?

The above questions raise the issue of transfer of skills. That is, to what extent do existing knowledge, skills and abilities (KSA; Table 2 below) as it applies to a language, sport, abstract vocation or product use-experience transfer to something similar or entirely different?

TABLE 2: Knowledge, Skills & Abilities (KSA)
Knowledge is what a person knows that is relevant to the job (e.g., knowledge of how menus and information are organized in a computer application ).  
Skill is what a person is able to do on the job. This includes both mental tasks (e.g., skill in doing algebra) and physical tasks (e.g., skill in driving an automobile).  
Ability is the capacity to learn a skill (e.g., cognitive ability is the capacity to learn mental skills). Abilities include mental abilities, physical abilities, and psychomotor abilities.

The extent to which KSAs are transferred from one system to the other is measured by the Transfer Effectiveness Ratio (TER). The best case would be to have a positive transfer or TER, where existing KSAs help and speed-up learning the new sport, product upgrade or new software version; the worst case is to have negative transfer, where existing KSAs actually hinder and slowdown the learning of a new system; the neutral case is where we have zero transfer, where it neither helps nor hinders.

Designing NexGen: "Creative Disruptions" with Minimal Impact to the End-User

So how does one facilitate and promote positive transfer when designing the NexGen product, system or software? Here are some means to that end. (This is not an exhaustive list, but are just high level pointers. You can contact HVHF Sciences to obtain solutions for your specific design problem.)

Be it a physical user-interface (knobs, buttons, dials) or a GUI or a mix of both, it is important to understand how the current version has been learnt, comprehended, and interacted with by users.

The following aspects in the NexGen product require special attention so that they have sufficient visibility, transparency and congruence with the old system.

Critical & Primary Interactions
This class of interactions, when they must undergo a change due to a giant leap in technology (e.g., glass cockpits in aircraft), should retain their unique traits that make them stand apart from the rest. Furthermore, it may be well advised, that the user-interaction design change with these features carefully identify, with research and testing, the ideal location on the change spectrum (incremental to radical).
E.g. of critical & primary interactions: emergency inputs/feedback, start/shut-off, escape, recovery, connect to Talk, steer, etc.

Anchoring & Landmarks
Users become highly dependent on well-known landmarks on an user-interface and anchor on it to perform oft repeated tasks. These landmarks should be treated with care and their design -- or even elimination (when necessitated) -- should be carefully considered for the new system. If substitutes are provided, it maybe worthwhile to ensure that their stimulus properties (affordance, visual design, location on physical/virtual real estate, etc.) have some correspondence with the old system. A poor execution of this was seen with Windows 8 due to the missing "Start" button, which hindered user-interaction.

Information Density & Cognitive Load
The new system should not appear dense, crowded and intimidating. The first few interaction experiences with the new system can be guided with a tour of critical & primary interactions. Complexity is bad at all times. But more so, when the user just wants to get started on those 20% of the features that he uses 80% of the time. There is no need to "wow" the user with 100 different new features ("featuritis") at once or out of the box. This may actually backfire and turn-off the users' interest in the product. Finally, visually aesthetic and culturally tempered design will make the experience less intimidating and even fun!

Instructive Errors
Errors that may occur should be instructive. That is, use errors to provide helpful hints and learning opportunities for the user -- instead of a cryptic bad bonk or error code on the screen. There should be no room for destructive errors, where the user is left in the dark after having hit a dead-end, taken a garden path, or crashed the system.

User Cognition

Mental Models
Users develop a deep rich, understanding of a system or product over the years. It typically concerns what the system contains, its sub-components, how they interact, information architecture, quirks, among others. A formal knowledge elicitation maybe needed to map this understanding and best utilize it when designing the new system so that the "plasticity" that maybe present in the mental model can be exploited and migrated to the new system.

Information Processing
Humans process incoming information using codes consisting of verbal-phonetic loop (speech, text), visual-spatial loop (pictures, locations) in various perceptual modalities (visual, sound, haptic); and they respond via certain modes (speech, manual control).  The last one becomes the input to the user-interface or machine. Disrupting this relationship (location, affordance/visual appearance of the control), as seen with the parking brake example (hand instead of the foot; lever instead of the pedal) can stall learning and adversely effect user perceptions of the product.

A human factors expert is required to analyze the code-modalities-mode framework and provide a robust design solution to transfer skills positively, or not be negative in the worst case.

Identical Elements
The perceptual elements (icons, tones, affordances, etc.) don't have to be needlessly changed unless required. One can use old furniture in a new house. Why not? Specially, if the old furniture is something that is preferred, evokes familiarity and sentimentality, and is in good working condition.
Furthermore, it is better to emphasize "Recognition" over "Recall." In other words, these identical or perceptual elements should not be hidden, requiring recall from memory, and imposing higher cognitive load; they should be promptly made visible to emphasize recognition and utilization.

General Principles
The inherent properties of a system -- for example, pointing and clicking on an app icon opens it or pressing the brake pedal stops a car -- fall under this category.The utilization of General Principles maybe challenged severely, when the NexGen change is akin to the difference between oranges and apples. Say, moving from command line interface (DOS) to graphical user-interface (Windows). Yet, it is must be done for obvious reasons.

Here the strategy should be to provide an easy entry using the general properties from the old system (as applicable) and immediately start demonstrating the ease with which tasks can be completed. Speaking of GUIs, what is of interest here is, it promotes "recognition" of UI widgets (icons, symbols, menu items) over "recall" from memory of complex commands used for DOS.

User Test, Test, Test
Finally, user test the new design as it goes through several iterations. It should be done by an unbiased, third party. Based on the metrics obtained from testing, calculate the Transfer Effectiveness Ratio (TER). Every effort should be made to maximize TER. The business and marketing team should be well aware about these metrics and lay the necessary foundations for the smooth introduction of the product through communiques and commercials.

I also recommend that you read my previous post on hyper-intuitive user-interaction design, which is closely related to this topic.

You can contact HVHF Sciences to obtain user-centered design strategies to bend the learning curve for your next version product and software upgrades or when designing user-interfaces for NexGen platforms. And user-testing to measure the type of transfer (positive, negative, zero) that would occur when the product is released in the field.  This applies to a range of commercial and consumer products, appliances, vehicles, medical and other instruments, software, etc.

Moin Rahman
Founder/Principal Scientist
HVHF Sciences, LLC
"Designing systems and solutions for human interactions when stakes are high, moments are fleeting and actions are critical."