The Fresenius Kabi Applix Smart food pump is a masterclass in bad design. There is nothing smart about this food pump and its accessories. It has been designed without taking into consideration the context in which it is to be used. Consequently, some of its accessories are not just safe enough to used around patients and the home and could lead to the compromise of patient safety.

When designing any form of interface, a usability consultant will push for the interface to be: effective and efficient; easy to learn and remember; useful and safe.

After usability comes user experience (in this medical context perceived user satisfaction is enough, as it is never going to be fun to use a food pump on a chronically ill person) which encourages users to feel supported and motivated by a helpful interface.

Fresenius Kabi has failed the user of the Applix Smart food pump on usability and user experience.

Typically, a food pump is used on patients who have medical conditions which prevent them from eating and drinking the amount of calories necessary to sustain them. So, the patient is equipped with a feeding tube and is fed using a feed pump throughout the day and overnight - in hospital or at home, looked after either by a health professional or a carer (usually a person who is only medically trained to do what the patient needs). Thus, the food pump must address the needs of a range of patients who need diverse but often very precise amounts of feed and the range of carers who have little to a lot of technological experience, particularly in medicine.

How not to design a food pump

The Fresenius Applix Smart food pump at first glance looks easy-to-use. It has 10 buttons:

• Two buttons at the top of the pump - one which increases or decreases the alarm volume as you toggle through the options, and one to set a key code.
• Four buttons - two up and two down arrows - to the right of the information display, to set the mls rate per hour and the current volume of mls to give.

Underneath the display from left to right there is:

• The on/off button.
• The information button, which when pressed tells you how many mls you have given and how many you have to give.
• The priming button which when pressed feeds the liquid down from the bag down the line to the end of the tube from the giving set, which is then attached to the patient’s feeding tube.
• The start/stop button.

The not-very-giving giving set on which you can choke

The pump comes with one type of giving set, which is where the feed is stored. The giving set line is fed through the side of the pump and hangs above it. This giving set has a medicine port, which comes off really easily and which a small child could choke on. Children who are chronically ill still experience teething and like to chew and put things in their mouth - particularly the line on the giving set. The medical port is just dangerous, detachable and could easily kill a child.

The giving set line also has a clamp on it. During training we were told never to use the medicine port and the clamp on the giving set and the key code button the pump. No reasons were giving, but it does beg the question why are these features on the giving set if they are never to be used?

There is no way that you can use any other type of set with this food pump, so if you have delivery problems (see customer care), which are quite common with Fresenius Kabi, then you cannot use this food pump at all.

Each feed in the giving set needs an extra 30 mls to prime the lines as they are really long, so you either waste a lot of feed or don’t give your patient enough. This should be mentioned in the literature.

Context is key in good design

Noise control: You may be able to choose the loudness level of alarm from the three preset levels, but you can’t switch it off altogether, so if you set the food pump to run all night and are lying next to the patient to sleep, if that food pump finishes at four in the morning, so does your sleep.

When the stop button is pressed, the alarm will sound every minute. Shouldn’t the user be able to decide how often they need to be told that the pump is on hold?

Backlit screen: A backlit screen is great for when you need to get information in a dim room when your patient is asleep. And an energy saving screen which only lights up when you press the buttons is not the worst idea in the world. But, when coupled with buttons which are small and difficult to see and close together is a terrible idea, especially when the pump has to be running for the screen to light up at all as backlighting doesn’t work in ’stop’ mode. Fluorescent buttons would have helped visibility and reduced errors (particularly switching the machine on and off which wipes out current totals), especially during the night.

Information: The information button is a great idea. But like the backlighting, the food pump has to be running otherwise you cannot get the amount of mls consumed and mls to give. What is the point of the information if you can’t access it at all times?

Increasing and decreasing volumes: The pump will only let you go up in groups of five mls once you are over 200mls. This is fine if you need to give vast amounts of feed, but what if you don’t? What if the patient is fluid-restricted and every ml counts? You can’t, for example, set the food pump to 499mls - it won’t allow you.

The lack of control to set alarms, food limits, and the inability to access important information as and when required leads to a frustrated user.

Grouping similar tasks together

The order of the buttons doesn’t make any sense at all, and seems to suggest that this pump wasn’t sufficiently tested before it was released.

Cognitively, buttons should be grouped together into similar tasks. So, the on/off button should stand alone - not next to the information button. It is easy, especially in a darkened room when interacting with these non-fluorescent buttons, to switch the pump off when you are looking for information. Ideally, the on/off button should be on the top of the machine where it cannot be touched by accident.

Once the on/off button is off the bottom of the machine the order of the buttons should following the order of how the user interacts with the machine: information, priming, and then stop/start.

The stop/start button is used more often than any other button - as patients vomit and need attending to often - so this button should be situated away from the information and priming buttons so that you don’t start priming more fluid into the patient when he/she is vomiting.

The information button should allow the user to access all information including the total mls consumed at all times whether the pump is running or not. So that the user can make an informed decision at each feedtime as to how many mls to give.

Buttons together which encourage the user to make mistakes, reset totals, and inadvertently feed a patient when the patient is vomiting create angry users and little faith in the food pump.

Accessories to complicate your life

We all love accessories and the backpack which comes with the machine at a first glances looks great. It is compact and has a velcro sealed panel which you open to interact with the pump. However, looks aren’t everything. You have to be really careful how to put the giving set into the backpack otherwise creasing leads to air holes and bubbles in the lines and the pump alarms constantly. You cannot prime the pump and giving set in the backpack at anytime, the alarms go off.

The backpack hole which the feeding line comes out of is not big enough for the giving set line to go through without the medical port falling off and without a big struggle with the clamp.

The pump also comes with a table stand which is awkward. It is slightly too big for a table, so has to be set carefully so that it will remain stable. It is not tall enough to stand on the floor, particularly if your patient sleeps in a cot or a raised bed. It is really heavy too so you have to be very careful where you place it, because if it was to fall on a child’s head, it would kill the child, an adult’s head would probably just need some stitches and the removal of the naso-gastric tube in the case of the resulting hairline fracture.

The clamp holding the pump has to sit at bottom of the table stand which would be fine if the a/c socket where you plug in the main adaptor wasn’t tucked so far under the pump holder that you cannot plug it in without tipping the whole stand and pump onto its side. But a pump clamped at the bottom of a stand which is on the floor is just dreadful.

Safety is key in all design. Medicine ports which can be eaten and stands which need a specific height table so that they are stable and can’t topple over and damage the patient’s skull are dreadful design problems which should be addressed immediately.

Customer Care Fresenius Style

The helpline number issued with the foodpump is 24/7, but interaction with the helpline number is dreadful:

Ringing the number outside office hours to talk about missing deliveries, or deliveries that aren’t yours and which have turned up at 6am can lead to you hearing comments such as: “I don’t care about your deliveries. Get off the line I am a nurse and I am here to deal with dying people.”

Ringing up to complain about how the above nurse thinks she can speak to customers anyway she chooses, leads to comments such as: “We have no formal process for complaint. All that happens is that I print out this apology letter whilst I am on the phone to you and I put it in the post. Do you want one?”

You are never given a time for deliveries so they can arrive from 9am until 6pm on your delivery day and if you ring to ask whether they have any idea what time the delivery will arrive, the answer is always resoundingly no.

Users need to feel that they have some control over their deliveries, their day and that they and their time are respected.

Sometimes, the delivery person has absolutely no idea about where you live and will ring you and ask you for directions. And if a delivery is incorrect, you will have to go through the same dreadful delivery process the following week as Fresenius cannot seem to correct their mistakes the same day.

Accidents waiting to happen

The combination of:

• the ill-designed food pump which supports user errors
• its badly-thought out accessories which encourage accidents and could lead to seriously damaging your patient
• the giving sets which have features which fall off and could potentially choke a patient, and ones which you are advised not to use
• the terrible customer care, along with the lack of respect for users on the telephone and the lack of respect for the user’s time when waiting for deliveries

all add up to terrible usability and worse user experience. It is bad enough that a user needs a food pump, as that means they are looking after someone who is chronically ill. This is a very stressful situation to be in already without being provided with a terrible piece of equipment and being treated dreadfully by the company who provides it.

Fresenius Kabi and their Applix Smart food pump are extremely lucky that they haven’t done more damage; their insensitivity and terrible designs add to users’ stress levels everyday of the week.

Today, I updated both versions of wordpress to 6.2.2 and found that everything works as it should do. A quick google about confirmed that I was getting the error message because my last wordpress upgrade hadn’t worked properly - I should have just reinstalled everything instead of fiddling where I shouldn’t.

There are ways of running two wordpress blogs with one installation, but if you decide to run two completely separate installations of wordpress like I have then you will need to make sure you have two .htaccess files. One in each directory. So, in the home directory my .htaccess (or the relevant lines) looks like this:

RewriteEngine On
RewriteBase /
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule . /index.php [L]

And, in the /jasmine directory the .htaccess file looks like this:

RewriteEngine On
RewriteBase /jasmine/
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule . /jasmine/index.php [L]

We have set up a blog here to share our news:
http://www.ruthstalkerfirth.com/jasmine/

Thank you to everyone who has contacted us to say they care. It means so much.

Steve Mann, inventor of wearable computing, came to the Ecole Polytechnique Federale de Lausanne in the 1990s when I was PhD student there. He had some difficulty getting on the metro as his head-mounted aerial added several inches to his height.

Watching him struggle to get through the door, I was inspired and excited by a researcher who wore and lived his work. Related MIT websites, where Mann was based, showed me how I could augment my reality by turning a gameboy into a wearable computer. The instructions came with a warning that it would affect my vision, though I would soon adapt to the constant red line. After all, the wearable was a lot smaller than Mann’s.

Today Mann’s wearables are very nifty indeed. Unfortunately, they didn’t escape Canadian airport security officials who unplugged him in Newfoundland and left him unable to function properly. Twenty years of living in an augmented world and having streaming information which assists his memory and enriches his world view cannot be switched off without causing damage.

Are we losing more than we gain?

More and more, it seems that humans cannot move forward with technology without there being a loss. As Mann demonstrated at the airport, augmented humans or cyborgs, can no longer function after being dependent on technology for so long.

Ubiquitious computing, where computers are off the desktop and embedded into our infrastructure so that we interact with them without our knowledge, offers the same results. Yes, it is all sexy to have your phone talking to your keys to remind you where you have left them; or, to have your thermal coffee cup reheat your coffee when it senses that it is cooling; or, to have the lights switch on and play mood-setting music when you enter a room. When it works, it is all very sexy indeed.

Tagged and bound

When things break down, we all grind to halt. You can’t find your keys, your coffee is cold, there are no lights or music in the house, because you have gotten out of the habit of doing these simple tasks. And now you can’t do them without debugging what went wrong. You are useless without your computer yet you are dependent on technology.

And that is an innocuous example application. In the ubiquitious world you are tagged and monitored by the smart badges, invented at Xerox PARC so that others can see how long your toilet break takes, who you are chatting to, and where you are in the building. You are trapped in a more sinister way.

Is this what we want from our technological advances? I’m only asking because it is often what we get. The alternative to controlling invasive applications is to train your users to become absolute system ninjas. But does technology make your ninjas’ lives easier if they have to go on intensive learning curves?

Augmenting reality for good

Augmented Reality (AR) researcher Ronald Azuma introduced such a system with his ultrasound-guided biopsy of breasts. This was an AR application with which I was so impressed that I always use it as an example of technology for good during my human-computer interaction lectures, even if saying breast to a group of 18-year old computer scientists is going to make them titter for the rest of the lecture course.

Azuma says himself that ultrasound-guided biopsy is difficult to learn and perform and needs good hand-eye coordination and three-dimensional (3D) visualization skills to guide the biopsy needle to the target tissue area with the aid of ultrasound imagery. Ultrasound pictures can be difficult to interpret, ask any mum-to-be about the scary information she gets from scans during hospital visits.

Lots more research is needed to augment our worlds usefully, which is why a stand-alone virtual world can sometimes be more attractive.

Virtual Reality

Interpretation isn’t a problem in the virtual world, regardless of whether the computer-simulated environment is real or imagined. Users can normally understand what it going on because virtual reality (VR) simulates the real world. Users learn how to fly aeroplanes or perform medical operations in simulators. The main drawback is that high-fidelity experiences are difficult to create because of processing power, image resolution and communication bandwidth limitations.

The other main problem is interacting with a 3D world. Most VR applications are so specialised, there are no standard interaction techniques. Many ways of interaction have been introduced to give the impression of real world feedback. Datagloves, or wire gloves, react to user hand movements to give the impression of resistance when a user comes into contact with artefacts in the virtual world. The user wears a head-mounted display to see only the virtual world. It is a far cry from the desktop and keyboard and mouse.

VR equipment, like its AR counterpart, is difficult to use, and can leave the user with headaches and other symptoms as the user adapts back to the real world. It is also very expensive which is why it remains in specific domains such as the military. Everyone else uses the cheap and cheerful version: the desktop virtual world.

Desktop virtual worlds: combining old and new

Second Life is one example of a desktop virtual world. It sidesteps interaction problems by keeping the desktop, keyboard and mouse and using old familiars such as instant messaging as a way of communicating. Users don’t have to learn lots of new interaction techniques, they can concentrate on learning the world they are in. Admittedly because it is run over the Internet, it can be clunky and slow. Still, it works well enough for users to create transport systems, IBM training seminars, or psychology courses which demonstrate how real hallucinations can be.

Learning and training takes place in Second Life in a cheap and uninvasive manner without interfering with your users’ brains and leaving them with headaches, or worse still, unable to function in the real world without technological support. Desktop virtual worlds let users choose whether they want to be ninja users or just observers. This leads to more satisfactory user experiences and better interactions. We can lead users gently into different technologies and allow them to feel that they have some control over their interactions and experiences so that they want to use them again.

Remembering what we have learnt

Technology is advancing constantly. And virtual and augmented world applications have the potential to improve our lives in so many ways. But we must learn to apply them usefully and to retain the knowledge we have learnt along the way so that we are enriched by applications, not damaged by them.

Last summer I found myself exploring an early Iron Age home at The Crannog Centre on Loch Tay. The Crannog was cosy, as its focal point was the Iron Age hearth - a large open fire. During the day the inhabitants would peel back wicker shutters to let in fresh air whilst they tended to their animals, making food and clothing and ground spelt for bread.

Today, wearing a woolly jumper and eating spelt pasta, with my back to the radiator, it seems to me that our needs and motivations have changed little since the Iron Age.

The Crannog would not have only sheltered its owners and kept them warm, satisfying their physiological and safety needs, it would also have been viewed by others as a status symbol, demonstrating the owners’ power and self-esteem, whilst raising them in the esteem of others. It would have given them a community too, along with a sense of belonging and social satisfaction. The Crannog dwellers would have been quite satisfied with their lot according to Maslow’s Hierarchy of needs.

Humans are needy

Abraham Maslow proposed his hierarchy in 1943, presenting them as a pyramid with physiological needs at the bottom and self-actualization at the top. Maslow suggested that humans work through the pyramid from the bottom to the top - as one need is satisfied, the next need is raised.

• Physiological - food, water, and shelter. Strong needs which force humans to think of little else when they are in discomfort.
• Safety - personal or job security. These needs appear once physiological needs are satisfied.
• Social - the need to belong to a club or a family gives meaning, love and affection, and staves off loneliness and depression.
• Esteem - humans need to respect themselves and have others respect them. They do this through their achievements and position in communities.
• Self-actualization - humans want to do what they were born to do, realise their potential, and feel fulfilled.

Later, Maslow added a self-transcendance or spiritual needs category. He put it at the top of the pyramid but stressed that spiritual needs could go hand in hand with the lowest of needs such as food and water. Spirituality or belief is a great motivator.

Spiritually uplifting buildings

In his classic book Why Buildings Stand Up, architect Mario Salvadori agreed with Maslow’s spiritual addition. Salvadori pointed out that the spiritual needs of humans have always preoccupied people even when they are lacking basic physiological comforts. Once humans abandoned their nomadic lifestyles to live in fixed communities, the larger architectural dwellings were usually places of worship and communion.

Monuments too, expressed a culture’s conception of spiritual needs, and of life and death. The Pyramids demonstrate Ancient Egypt’s obsession with death. The Mausoleum of the First Qin Emperor, Qin Shi Huangdi, complete with terracotta army, illustrates his obsession with immortality.

Our cultures and the way we do architecture may have changed, due to the advances in science and technology, but our needs, from physiological to spiritual, have not. We live, eat, sleep, and procreate in the same way. Today though, the largest buildings we create are shopping malls, as fewer people are motivated by organised religion. We worship the material and these large structures express our concept of our greatest need: to consume as much as we can.

Motivating your user

Even new libraries look like shops. Traditionally, libraries fit with the two motivations Maslow’s identified alongside his pyramid:

• Cognitive : the search for knowledge and meaning.
• Aesthetics: the appreciation of beauty and balance.

Some older libraries look like temples from Antiquity. However, David Adajaye’s Idea Store in Whitechapel, Tower Hamlets’ newest library, looks like a department store. The London Borough is trying to motivate people who don’t want to read, but enjoy shopping, to go inside.

With most buildings, schools, prisons, hospitals, and police stations, we can tell what is going on inside by looking at the outside. Form follows function. When we enter them, we do so with certain expectations, and we are motivated by a specific goal which we would like to achieve, otherwise we wouldn’t go in.

This applies to most things with which humans interact: websites, graphical-user interfaces, other humans, and any artefact that will or won’t move a user nearer to the goal he is motivated to achieve. If the user does not find satisfaction, he continues his search elsewhere.

Human motivation and satisfaction has the greatest influence on individual behaviour which is why business sociologists spend so much time trying to come up with ways for organisations to motivate their employees for the greater good of the company, not the individual. Cognitive science research shows us that interfering with human perceptions and processes has a negative impact on human performance and goal achievement.

Satisfying users

So do the shopping users who go into the Idea Store, stand about wondering where to find ladies underwear? What about the reading users? Are they still provided for? Time will provide the answers to these questions. And if we see more department store libraries popping up across the UK, is this what library users want? Or what local authorities want?

Users are individuals with a common set of motivations, defined by needs, and shaped by cultural patterns. Users are conditioned from birth and according to Joseph Campbell certain patterns are hardwired in the human brain. As designers and architects, we need to tap into user patterns, needs and motivations to create a better infrastructure for a better society, not one dictated by organisations, whose idea of the greater good can mean creating structures which do not address local community needs but allow them to spend their budgets meeting meaningless government targets.

Like many usability consultants I have spent hours locked in rooms with strangers saying: “What do you think about this web page?” It is boring way to earn a living especially as you often know the answers and could tell clients without asking the questions.

Alas, most clients only believe opinions about their websites when it comes from random users - not you, the expert. Luckily the industry takes Jakob Nielson’s advice on testing: five users only to establish a pattern of responses (and because it’s cheap). Although, if we were really serious, we would need 30 users to talk about the statistical significance of our results.

Some of this boredom could be avoided (and client money saved) if everyone employed the patterns which already exist in users’ heads to create more intuitive webpages and GUIs. Cognitive science, the study of mind and intelligence, enables us to understand what makes our users tick.

The human as a factor of design

We need to understand what makes our users tick because they are a key factor in the design of any system. And when you need your new design to interact with another system, you begin by trying understand that system and design around it.

By adopting this approach, we assume that humans are rational. Their logic has developed over time in order to meet the challenges of the environment in which they live. From long ago when managing their meagre sources of food and shelter to today’s information overload in the frenetic world of business, humans have developed a range of cognitive mechanisms which allows them to interact with their environment and designers should capitalise upon these cognitive interactions.

Understanding humans, in particular how they perceive, reason, and act, is not a new idea. The Ancient Greeks spent many hours philosophising about the meaning of existence and how the mind and body are related. This century, artificial intelligence researchers have wondered whether we can model the mind and if so, can machines think?

No is the answer. Machines only do what is asked of them. However, since it can be difficult to understand what the original question was, humans tend to anthropomorphise machine ‘behaviour’ and attribute human logic to them. This is know as superstitious learning, especially when the connections between actions and outcomes are misunderstood.

The use of heuristics

We make decisions all day long without gathering and analysing information. Because of our memory and the limitations of our nervous systems, there is a limit to the amount of environmental information and complexity that we can absorb and cope with. Even when we make a conscious effort to make decisions rationally, we simplify using assumptions and often work with incomplete information. So we use heuristics as a way of reducing the complexity of decision making: We associate a particular brand with quality rather than debating the pros and cons of different sports wear.

We create shortcuts and rules so that we can continue to absorb new information and reason ‘logically’ in the world.

Reasoning: deduction, induction, abduction

Philosopher Charles Pierce developed a classification system for reasoning known as deduction, abduction, and induction. Given the propositions A, B and rule R:

• A: It is raining
• B: I am wet
• R: If A then B

Then:

• deduction: A, R -> B : It is raining therefore I am wet
• induction: A, B -> R: It is raining, I am wet, the rain made me wet
• abduction: B, R ->A: I am wet therefore it is raining

Only deduction is logically sound unless we use a closed world assumption - if we don’t know about something, it is not true. This happens a lot in computing. Humans, however, know instinctively to add extra information such as problem specific details (I took a shower and that is why I am wet) and context dependency (I am indoors). However, humans do learn inductively. After 30 instances of A and B, they learn R: A -> B and think everytime it rains I get wet, I had best take a brolly. However, there are times when their rules are faulty which leads to supersititious learning.

Getting rules right often depends on how something is represented and it is no coincidence that Pierce himself defined a model for describing signs. He was interested in semiotics (the study of signs and symbols) as well as logic and reasoning.

If you torture the data it will tell you anything

The way we represent or ‘frame’ a problem directly influences the decisions people make. This cognitive bias or the psychology of choice is illustrated in the following example (taken from Universal Principles of Design, Lidwell et al).

In 2002 Russian Special Forces gassed Chechen rebels in order to rescue their 750 hostages in the Moscow Theatre. The ploy worked but the gas killed over 100 hostages. Newspapers across the globe reported the event in two ways:

• Gas Kills Over 100 Hostages
• Gas Saves 500 Hostages

The negative framing of the tragedy makes the Russians look like they handled the crisis badly. However, the positive framing suggests a satisfactory solution to an intractable problem.

This framing technique is used in marketing when vendors talk about all mobile calls free when you actually have to pay a £35 per month contract, or food as 95% fat free whilst ignoring the 5% fat or unhealthy trans fats. Politicians use the same sort of spin to hide unsuccessful outcomes of ill-judged policies.

Framing can, however, be a powerful tool. Designers frame problems differently in order to get creative and ‘think outside the box’.

Visualisation: A picture speaks a thousand words

Pictures and graphs help us more easily assimilate information and there are many different forms of visualisation. Graphical representations are powerful and can give you new insights when they are meaningful.

Statistician Edward Tufte has written several books about designing more meaningful graphics, especially when representing multiple sets of statistics. His main message is that viewers are sophisticated individuals:

…give them the greatest number of ideas, in the shortest time, with the least ink, in the smallest space…

Giving humans good graphical representations allows them to find and understand patterns on many levels. We look for patterns to make sense of the world. We interpret everything from the weather to the stories mankind has told since time began.

Joseph Campbell stated that myth and legend provided us with good patterns which guide us to live successful lives. Dropping the classics from education has led to younger generations missing out on these patterns, but they recognise them inherently from Hollywood films as there are only a finite number of storytelling possibilities.

Perception

Gestalt theory is the study of how humans recognise and interpret patterns and is the result of research carried out by German psychologists in the 1920s. Their works shows that we tend to:

• Proximity: interpret items close together.
• Similarity: group items which look similar together.
• Continuity: interpret lines as continous if they don’t bend sharply
• Closure: prefer regular shapes and complete them (inferring occlusion or interposition) if necessary.

These ideas underpin graphic design and the design of 2D graphical-user interfaces.

Vision uses context: Colour, Illusions, Adaptations

Our eyes are sophisticated pieces of machinery and apparently, 90% of our brain is devoted to vision but we still do not understand exactly how the brain works. Because of the interaction between the brain and eyes we can interpret different levels of brightness and contrast as coherent and meaningful displays. We also use other cues to interpret the world around us. We use depth perception and take our cue the 3D information to estimate distances and constancy:

• Depth from motion: when something is getting smaller or larger it is moving from or to us.
• Colour interpretation: objects are interpreted as distant or close because their colour is at the end or beginning of the spectrum.
• Perspective: converging parallel lines give us a sense of distance.
• Constancies:
  • Shape: if a door is closed or ajar, we still know that it is a door.
  • Occlusion: if a shape is partially covered up, we complete the shape in our minds due to Gestalt law.
  • Location: if I move the rest of the room doesn’t.

There are many other examples of our ability to use extra information to interpret the world around us. Graphic artists, like Escher, have played with these cues to create illusions in their work.

Attention: limitations and allocation

Illusions may be easy to create in part because humans are constantly bombarded by signals to their senses and have limits on what they are able to interpret at any given time. German physician and physicist Helman von Helmholtz, inventor of the ophthalmoscope, discovered that the nervous system is a single communication with a limited capacity. To get people’s attention designers highlight specific information - by underlining, or using white space, so that people actively attend to the information the designers want them to see. Allocating attention works well in the theatre with spot lighting and in films with zoom lenses.

Another way is to give humans a specific task to perform. This allocates their attention and can lead to inattentional blindness - that is they are so fixated on one thing, they don’t see other things that are going on in the room. Daniel Simon’s basketball video is a great example of inattentional blindness.

We tend to fix our attention early on, so getting humans to change their attention elsewhere can be done only by interfering with their perceptions and colour and their memories. However, humans are accustomed to carry out numerous tasks at the same time - driving the car, and listening to the radio. Sometimes we don’t even remember doing certain tasks that we do routinely - like locking the door on our way to work. So in a demanding environment like the cockpit of a plane, where multitasking is essential, humans need cognitive aids to get reminded of certain crucial moments. Pilots have checklists to tick off and they also have little tags tied to various knobs and levers - to remind them to perform tasks.

Memory

Visual cues help pilots more easily because humans remember procedures better when they are described in pictures and words. This concept is known as dual encoding. It works more easily with words which are not abstract such as chair, dog, window. Abstract representations are more difficult to reforce using dual encoding because beauty, truth, freedom mean different things to different people.

Chunking down information into manageable sizes for humans helps them to retain information. Five to seven is the optimum range as discovered by George Miller. Mnemonics (from Greek times) are another way of enabling our recall.

German psychologist Hermann Ebbinghaus was the first person to make a scientific study of memory in a laboratory so that it became standard. He identified three stages in memory:

1. Encoding (original acquisition of memory).
2. Storage (retaining information over time).
3. Retrieval (gaining information when required).

The standard experiment is to encode the information, particularly in education, and then test it later. However, we have no way of of telling if the information was encoded correctly so the test breaks down.

We have many types of memories:

• Procedural: tying shoelaces, riding a bike, walking , talking.
• Declarative: knowing facts about the world and your life and past.
• Episodic: remembering episodes of our lives that are contextually bound.
• Semantic: NaCl symbol for salt what a platypus is.

And these move from our short-term memory or working memory to our long-term memory. No one knows how long this takes, or how, or when exactly, but this information does move. We do know that when we learn, we have levels of processing to make information ’stick’:

1. Graphemically - visually.
2. Phonemically - aurally.
3. Semantically - meaning.

As artists like Escher tried to interfere with our perceptions, so it is easy to interfere with memory so that we can have memory illusions and false memories. We can have retroactive interference where we are eyewitnesses to an accident and then later as we discuss our memories with others we remember their information too - which can be false. And over time our memories degrade - no one knows why, it could be age, and the greatest time when we forget the most occurs just after learning.

As humans we are motivated by all sorts of factors so forgetting and remembering can be influenced by our motivations and by others’ motivations, but then that is another blog.

After my first year at university I spent the summer working in a delicatessen in Putney. One morning during my first week, whilst in the middle of carefully carving six slices of Parma ham for some lady’s dinner party, we were told to evacuate the building as security had been warned that there was a bomb. I dropped everything and ran for my life. We stood around in the car park until we got the all clear and I arrived back at the counter to find the same woman ready to berate me for abandoning her dinner party plans.

Every week after that there was on average three or four scares per week and I became like the dinner party woman: nonchalant. Instead of leaving the building immediately, I would go upstairs to the locker room and get my purse - just in case we went to the pub. Sometimes people would shout up, “Bring my fags/jacket/bag down will you?” Bomb scares became part of my daily routine, so much so that they were no longer life-threatening traumatic events but welcome tea-breaks.

Humans are adaptable, which is why we have survived so long. However, this flexibility means that the security systems specialists work hard to put into place are undermined by users’ perceptions of risk. As the old adage says familiarity breeds contempt. As I did in the delicatessen, users get used to a certain level of risk, become blase about warnings, until security and safety awareness procedures get ignored. Even the simplest routines can be forgotten. How many of us carry out weekly back ups? Hands up if you check that your automated back ups work? Most people don’t even think about back ups until they lose an important document. The same goes for security systems. When there is a security breach 80-90% of the time the human or organisational element is responsible not the algorithms on which the system is built.

Users are not the enemy

Sometimes this leads managers to view users as the enemy and they spend all the their time and energy hating stupid users. By all means track their behaviour, collect data on their log-ins and outs, check what they are doing. Information is power. But rather than drowning in data and hating everyone, use the information to learn about your user group and visualise their needs. Redesigning log-in procedures and reducing human error can go a long way towards reducing security breaches. Your users are not the enemy, they need to be supported and viewed as part of the system.

Psychological responsiblity

Another step is to give users the tools and information necessary so that they can become part of the solution and not the weakest link. Since we first started interacting with machines, researchers have studied what makes us tick. Cognitive science in particular concentrates on what humans are thinking when they are presented with a computer. And humans have been practising cryptography - hiding and transmitting sensitive information - since time began.

The best way to get users to not violate the security built into systems is to encourage psychological acceptability so that users learn to apply security procedures correctly. Psychological acceptability involves good usability, feedback, system transparency, and a sense that users are responsible for their actions.

Give users transparent interfaces and good feedback

Usability is the effectiveness, efficiency and satisfaction with which people can carry out a given task using a computer. A system should be easy to use. Normally the graphical user interface (GUI) or just a user interface is responsible for the way users interpret what is going on and is the main factor in influencing how users feel. Good GUIs should be transparent - familiar or somehow intuitively apprehensible and the internal workings of the system should match the users’ mental models. And if it can’t then users need to be given the appropriate guidance and level of feedback so that they can learn how the system works and what it does.

By giving good feedback and creating a transparency of interface, so it is easy to understand how a system came to a specific conclusion, users begin to have confidence in their computers and will trust what computers say. Loss of trust means that users start ignoring warnings and overriding defaults.

Allows users to accurately estimate risk

With good feedback and a sense of trust users get a more accurate perception of the risks to security. And if the GUI is a good one - easy to use, with a clear uncluttered design - users can more easily concentrate on the information they are being given.

Some researchers have suggested user profiling for improving users’ estimation of risk. Target users are split into groups and system responses are designed to respond to a particular user profile. Each profile ideally contains an individual’s cognitive style based on users’ behaviour that has been tracked and analysed. If a given user’s behaviour changes and a security breach is anticipated then this user would be assigned a new profile so that the level of information security increases.

Avoid the erosion of security awareness

Jose J Gonzalez, a professor of systems dynamics based in Norway, argues that the erosion of security awareness or users’ inability to follow safety standards is comparable to the erosion of safe sex practices that we are seeing in teenagers across the globe. We all want to interact safely but when the moment arrives, circumstances or time pressures, and lack of cooperation may cause users to deviate from safe practices.

When dealing with users whose compliance with security measures declines as perceived risk declines, Gonzalez recommends that users should know about the ever present dangers. He suggests using refresher training courses during risk-free periods to remind users and managers to carry out safety procedures. In this way users avoid behaviour built on superstitious learning, which is when they learn how systems work in an almost cross your fingers way instead of viewing their interactions as a logical process with a finite number of steps.

It is also good to tell users when a security breach has been averted. In this way users know that although they are protected, the systems they use are not invincible and are constantly under attack. Reminding uses that they are part of this system, that they are key to the smooth running of the place, and that they need to follow safety standards is just good practice.

Make your users responsible so that they don’t become the ones who inadvertedly let your system down.

Another Apple marketing frenzy has led to the UK bracing itself for the launch of the iPhone tomorrow. The Carphone Warehouse is expecting large queues and Scotland Yard are warning customers to hide their new handsets so that they don’t get mugged.

Aside from the excitment there are criticisms. The main ones centre on the iPhone’s choice of network: O2. O2’s coverage isn’t great, apparently even in the Apple store on Regent Street. And unlocked iPhones that early adopters are already using, thanks to Ebay, won’t be able to download new software without damaging them. Vendor lock-in experts Apple are as bad as Microsoft with their need to dictate to customers how their products should be used, which ultimately is a big problem when you talk about the iPhone’s user experience and usability.

Multi-touch interfaces

And user experience is what most reviewers, thus far, focus on. The multi-touch interface of the iPhone is offering something new on a mobile phone. Multi-touch and direct manipulation research has been undertaken in the human-computer interaction (HCI) community for over 20 years, even if Steve Jobs is telling everyone that multi-touch is new. New interaction or not, if the rest of your technology isn’t delivering - and all over the web there are disappointed users grumbling about EDGE, streaming difficulties, speed issues, capped page number downloads, and pricing - then a good interface alone isn’t going to save your user experience. It is rather like raving about the interface on an ATM that doesn’t deliver the money you want.

Increased functionality = more complex interaction

It is a shame for Apple that the multi-touch interface is not adequately supported as it is a good interface solution. When you have a device with many functions - phone, camera, web and email, mp3 player, etc., - you potentially increase the complexity of interacting with it. One way to deal with many functions is to include multimodal interaction which offers several ways of interacting with a system such as keyboard, mouse, light pen, voice recognition and others. It can cover a larger target audience because those users who have difficulties with a mouse may find a light pen or voice recognition easier to use. The downside is that more devices or peripherals lead to more things to carry around. The multi-touch enables you to provide one way of interacting and in the case of the iPhone, on a small screen. The disadvantages are that you need to learn how to gesture properly and flex those fingers.

Gestures

Gestures have been used by humans since we first starting communicating and they have long been of interest to the computer community. Facial gestures are researched in virtual reality systems. Hand gestures are common in games and animation. So it is worrying that Apple have seen the need to patent a multi-touch gesture dictionary in the US. I only hope that they are giving veteran multi-touch researcher Bill Buxton a cut of the revenues this dictionary will generate.

These Apple gestures include flick and drag - for browsing photographs and emails, squeeze and stretch - for zooming in and out when looking at photographs. They are great and snazzy but not without drawbacks. Depending on your finger size a flick can be interpreted by the iPhone as a tap, which is used to stop scrolling. And it remains to be seen how satisfying gestures are once you are navigating a large library of photographs and emails. Like most input devices, your finger can be great for many applications and rubbish for others.

Phone fingers needed

Typing out emails on a small virtual keyboard where the letters pop up as you press them may be exciting at first but will it go the way of many typing experiences? The ‘it would be quicker just to phone the person up’. Who can tell? User experience is subjective and difficult to measure. One interface aspect may drive some users mad whilst others love it. So when you rush out tomorrow night to buy your iPhone complete with finger condom to stop the smudges and a Scotland Yard coat with big pockets, don’t expect the ultimate user experience that the marketing hype is promising you, just enjoy owning a new flashy toy. All technology is still a work in progress and even intuitive interfaces have a learning curve.

Like every self-respecting human-computer interaction (HCI) lecturer, I introduce task analysis or the technique of analysing how people perform a task or job, to my students a couple of weeks into a given course. Each time I am aware that I fail to get excited about task analysis and so give it a bad press.

This is because I was taught to program as an undergraduate at Liverpool John Moore’s University (LJMU, then Liverpool Polytechnic) by a group of very good lecturers and researchers, in particular John Pardoe, Melvin King and Stu Wade. They used literate programming as the basis of their teaching methods.

Literate programming was first suggested by Donald Knuth in 1984 with the aim of changing attitudes towards programming. He proposed that instead of giving a computer a list of instructions, it was better to explain to human beings what we want a computer to do. This meant generating good designs and good documentation. At LJMU they combined the idea of literate programming with Niklaus Wirth’s Stepwise refinement design method and wrote a fantastic programming Pascal-based support environment which taught students to perform good design and appreciate self-documenting programs.

As an undergraduate I was taught to think about program input and output, code presentation and documentation: Could someone else pick up what I had done and understand it? And I appreciated the self-generating diagrams (flow charts) and documents. I hope the software is still in use, not least of all because Pascal is a good educational tool to introduce students to programming and to learn proper programming skills (Java isn’t - but that is another blog). The downside for me is that with such a good grounding in programming and design methods when I first looked at task analysis and the hierarchical task analysis diagrams (flow charts which are a bit more complex, hence the hierarchy) it seemed a derivative method which would only serve to confuse the person using it.

Theoretically there is nothing wrong with a derivative method because HCI often borrows from everywhere else, after all why reinvent the wheel? However, task analysis diagrams are useless when compared to literate programming/stepwise refinement diagrams because task analysis charts are trying to describe how a user goes about performing a specific task which may include using a computer.

The purpose of task analysis, in the main, is to understand how a user does something. However, if you show the flow chart back to the user or your client you can guarantee it will be a turn off because humans who are not computer scientists don’t want to look at a diagram that seems to be computer domain specific. A flow chart traditionally models information flow - hence its name. Placing human activities, often goal orientated tasks, into boxes meant for modelling information flow doesn’t seem the best way of trying to capture and describe human behaviour. There is no clear indicator of input or output, and there is no clear division of tasks which occur computationally and in the real world. If you interview five to ten users and you may well get five to ten unrelated diagrams back with no commonalities either. There is no indicator either of what would need to exist for the user to be better supported during this task.

In a nutshell, task analysis encourages a great mish-mash of ideas and provides very little help to the designer. This is supported by many articles on task analysis saying: ‘There is no wrong or right answer’, ‘There is no one way of performing task analysis’, and my favourite ‘It may not be suitable for all contexts’. The examples most textbooks use to illustrate task analysis don’t help either: making the tea or hoovering the house are not about to be automated, don’t need a computer, and are very different from inputting data into a program and getting data out of it. During lectures, I have seen great rows break out about tea-making: cup or pot, milk in first or last (often students think is a useful way of distracting the lecturer from making them think anymore about task analysis. If only they knew).

The results of task analysis theoretically help developers to write manuals and tutorials, detailed interface designs - particularly ordering subtasks in interface design, and scenarios for designing interfaces. However, these three areas are wide ranging and need many inputs not just one flow chart that isn’t a natural description of how a human behaves. During consultancy I have never used a task analysis diagram.

This leads me to conclude that task analysis is really just useful for exam questions. You can get students to read a text and then get them to draw some diagrams - which isn’t what you would do in the real world, as you would get the students to talk to users (I got them to do that once and they didn’t like it and wanted me to provide them a summary of everything the user had said, *sigh*)- and it becomes an ‘easy’ question to ask and mark.

If the students have been paying attention and a) know what task analysis is, and b) can show that they can draw a flowchart moreoreless, then as a lecturer you have asked them to provide some book learning in a) a definition of task analysis, and b) application of learning, and you get a pat on the back for constructing good exam questions from the external examiner.
That just leaves us struggling through the lecture where students are first introduced to task analysis and have to force all sorts of unsuitable information into a flowchart. Apologies to all students past and future!