1. Define what a display is, in human-machine system terms.
2. Describe the nature of information from a human information processing perspective and understand the implications of the Multiple Resource Theory for design.
3. Analyse the interaction of the user with a display in terms of the nature of the task demands it gives rise to.
4. List characteristics of visual, auditory and other modes of display.
5. List characteristics of multimodal displays.
Examination of Figure 1.1 in Sanders and McCormick (1993) will show that the two main features of the machine in the human-machine system are the displays and the controls. This unit will be concerned with the display of information for effective human use. As one of the most dominant areas of ergonomics, it is a good place to start. As well as specific issues to do with displays, general points about usability should also emerge within this unit. Examination of the figure may convey two impressions. Firstly, that a system only involves visual displays, and secondly that these displays are to be found on specialised hardware or machines.
Displays can take the form of any technology and can be other than visual. Having said this, it is the case that the visual world is not only predominant in everyday life, it is also predominant in terms of displayed information. However, this unit will also give some attention to information that is displayed in the auditory as well as in multiple modalities (e.g. visual and auditory). In examining displays, the full range of technology that has been used has to be considered. Taking the visual world as an example, the range will be from the paper and pencil message stuck on a wall, through to emerging technologies of virtual reality as used in advanced simulation facilities. In the auditory world, the range will be from the voice that we hear over the telephone speaker through to the synthesised speech messages that we will increasingly encounter in interaction with computers and other forms of advanced technology.
Another thing we need to bear in mind is that whenever we look at a particular aspect of ergonomics in isolation, in this case displays, we must never lose sight of the fact that this aspect is inextricably linked with other aspects of a system. Any attempt to optimise a display in isolation from how it is going to be used will yield less than ideal solutions. If we pin our hand written note in a dark corner of a corridor, or at a very low height, it may not be noticed by the person for whom it is intended. If we make virtual reality software part of a car simulator’s visual display that can only be seen by an individual wearing an uncomfortable headset, the use of which gives rise to feelings of nausea, then again we will have a shortcoming in our overall system. However, we will only make progress if we do consider things initially in isolation, develop the best principles for their design, and then see how they interact with the other system components.
It is also important to make clear that as well as providing information, we also have to bear in mind the likely consequences of responses to that information. For example, providing a notice on the outside of the building, telling people that they are entering a no smoking area may well produce a desired result inside the building. However, if we do not provide facilities in the vicinity of the sign for people to extinguish their cigarettes, then we may well find an unwelcome mess of cigarette ends at the entrance of our new prestigious building. The whole system has to be considered at some stage. To optimise the system, however, we must initially get compliance with the no smoking request, so a well-designed and well-positioned sign is of importance as the first stage.
What is a Display?
A display has been defined as follows:
Displays are devices, no matter how simple, which are used by the information sender to communicate with human receiver. Buck (1983, p. 196).
The point emphasised above about the range of technology used is reflected in Buck’s definition referring to “no matter how simple”. It also reveals that there is an information sender and an information receiver. The information sender may be an individual or an organisation. For example, an organisation may have a policy that has been implemented by the display, e.g., the no smoking sign. For individuals, it may be a more immediate sender who is generating information for the user. For example, a message over the telephone. The information sender may not be another human being, but may be an aspect of the machine or artefact that the user is interacting with. Our interactions with the cash machines use a scripted dialogue, with alternatives being presented at different points in the interaction. Another human being has designed and programmed this dialogue, but our interaction with human beings within this system will rarely happen, unless for example, we are trying to extract money from an overdrawn account when the machine may retain our card and we will have to ‘face the music’ with a bank employee. Buck goes on to clarify the communication role of displays;
Since the role of displays is to communicate, displays need to be chosen to meet the necessary conditions of communication, provide the types of information needed, be relevant to the situation, and fit the people who need or want the communication. Buck (1983, p. 197).
Buck also introduces the concept, central in ergonomics, of compatibility to help explain how we can make displays more effective in their communication role. A first principle is that the display should be compatible with the human senses. The human band-width is well understood in terms of visual and auditory sensitivity. What this translates into in terms of visual displays is that they should be visible, for example, either from their own sources of luminance or through the amount of illumination they receive from a lighting source. The next form of compatibility that Buck refers to is that displays must be compatible with a human language system. At first glance, this would appear to be compatibility with a familiar spoken or written language, but there are also coded and symbolic languages that we can learn and become familiar with. Symbols are usually visual representations, concrete or abstract, with which we have associated concepts. The no smoking sign, displaying a picture of a lit cigarette, usually with a red bar across it, is well understood whether the language in a particular country is understood or not by an individual. Similarly, musical notation is a universal symbol system for conveying the structure of Western music.
The next level of compatibility is with the user’s intelligence. An alternative view of this is that it means compatibility with an understanding of the task in hand. A visual display consisting of a few bars of music notation will not be of any assistance to an individual who cannot read music. If, however, the bars of music have been sent by a friend who knew that the individual was trying to recall a particular song, then that particular display of information, given musical knowledge, will immediately have relevance for the task in hand.
This later point is an important aspect of information communication. Information can be seen as communication that removes uncertainty. This somewhat abstract view has an important meaning. We are constantly bombarded with stimulation from the outside world. For a stimulus to have an information value for us requires that we are in a state of uncertainty. In a highly structured task environment, this is easy to see. If we need a particular piece of information to move forward, then we will be very much dependent on the arrival of that information. When it does arrive, it removes all the uncertainties that we have, and the situation is resolved.
Other stimuli will have a more general value for many of us, but a specific value for others. The non-smoker arriving at the new building will possibly notice the sign, but it will only have value for them if they have a particular concern about being in the presence of smokers. Therefore, the state of uncertainty is whether they may be in close proximity to a smoker. Seeing the sign resolves or removes that uncertainty for them. For others, without any particular concern with these matters, it would convey no information other than to classify the building in a particular way for them. For the smoker, however, it conveys a lot of information. Their uncertainty will centre on whether they would be allowed to smoke in the building or not. The presence of the sign eliminates a possible situation they would desire, therefore it carries information for them. The difference between the stimulus and its information value is hopefully conveyed by the following simple example.
A lecturer confronted a class with a new situation every week when they attended their lecture. The lecturer threw a six-sided dice and the number which appeared face upwards on the dice determined the next course of action in the lecture.
• If it was a ‘1’ it meant that the lecture proceeded as normal.
• A ‘2’ meant that a two-hour lecture, rather than a one-hour lecture took place.
• A ‘3’ meant that there would be a half hour test followed by a half hour lecture.
• A ‘4’ meant that there would be a one-hour essay exam instead of a lecture.
• A ‘5’ meant that the lecturer would produce a detailed handout instead of a lecture.
• A ‘6’ meant that there would be no lecture that week.
It is clear that the students in the lecture would have a great interest in the outcome of the dice throw. In particular, they would have six alternatives, some much more attractive than others. If the lecturer wanted to be unpleasant on a particular day, they might say, having thrown the dice (which is out of the view of the students), that the outcome was “an even number”. Having made this particular verbal statement, the state of uncertainty in the minds of the students, should now have changed. There are no longer six alternative outcomes for the day’s session, but only three. If the lecturer then says “it’s a 4” then, all their uncertainty is removed and they know the outcome. Thus the value of the statement “it’s a 4” in this situation removes half of the uncertainty that they had before the dice was thrown. It therefore carries a certain informational value. If the lecturer did not want to delay things then they would say at the beginning “it’s a 4”. The statement at that time has a very different informational value, it removes all the uncertainty that the students had in the situation. The important point here is that the two statements consist of the same stimuli, “it’s a 4”. However, the information value of these two identical statements differs depending upon the state of uncertainty of the students. This approach forms a basis of a very important body of theory in relation to communication of information to humans, Information Theory (Wickens, 1992).
Any stimulation that we receive that has no informational value we can call ‘noise’. This word has an everyday meaning of ‘annoying sounds’, but in its technical sense refers to stimulation that carries with it no information. Any display can have ‘noise’ within it if this is not information that is currently being sought. Searching for a particular directional sign in a building can be made more or less difficult by the presence of other signs and other information in the same vicinity. This other material constitutes noise. In listening for a particular auditory message, there may be interference from real world noise, in the everyday sense, or it may be that there are a number of competing auditory messages that can be heard at the same time. The power of the attention mechanism enables the selection of information from a noisy background. The ergonomist’s task is to improve what is known, in the technical sense, as the signal-to-noise ratio. Sometimes this is fairly easy to do, for example, making a telephone bell noise much louder than most of the surrounding information. On other occasions, it is more difficult to do because what is signal and what is noise will depend upon the state of uncertainty, or the informational needs, of the individual. A set of directional signs in a building will need to be clearly presented, and the users need to have their search task enhanced by the way that the signs are laid out and they should find meaning in the messages that the signs convey.
Problem of Display Design
Another way of typifying the problem of display design is to look at it from the perspective of the tasks that an individual will need to carry out in order to meet the demands of the system within which they are working. Easterby (1984) listed the tasks involved in working with displayed information as follows:
Determining the presence of an object, target or symbol.
Determining that differences exist; discriminating between target objects and non-target objects is determining differences on the basis of which identifications can be made.
Attributing a name or meaning to some object target or signal. Discrimination and identification are often parallel processes, but in psychological terms, they make different demands of the presented information.
Determining whether objects in the display have been seen and perceived before. Identification often accompanies recognition.
Understanding the meaning of a given display so that an associated consequent course of action is both apparent and possible. Comprehension involves recognition as a necessary but not sufficient condition. (Easterby, 1984, p. 21)
Depending upon the nature of the task and the nature of the information displayed, different features of human performance will be given more or less emphasis. Whatever the purpose of the display, all of these tasks will be involved to some extent, the ease of which they can be completed will be strongly influenced by the way in which the information is designed. This in turn relies on knowledge of human perceptual and attentional processing as well as the specific skills, expectations or knowledge of user. If the potential user is any member of the general public, then certain assumptions have to be made. In the case of locating a room in a building, an assumption might be that for the majority of the users, they will speak the language of the country in which the building is located. Given that some visitors may be partially sighted or not possess the necessary language skills, then a fall-back position is needed with a receptionist to provide back-up. However, in a highly coded situation, a simple amount of information such as “it’s an alpha 1, 2, 3” may provide all the information that the experienced person needs to meet their current task demands. If we return to the definition of usability, “effectiveness, efficiency and satisfaction with which specified users can achieve specified goals in particular environments”, then we can see that understanding the task demands, in particular knowing about the system goals, the users and environment in which the information is being displayed, are all parameters that we must explore in order to enhance the usability of the situation that any users encounter.
Classification of types of Information
Buck (1983) provides a useful classification of types of information that can be displayed to users.
Each of these types of information can, in theory, be provided on most types of displays. However, some lend themselves better to one form of display rather than another. The characteristics of each of these types can now be briefly discussed. The particular forms of technology that can be used to implement them will be discussed in more detail in a later section.
1. Instructions refer to information that guides behaviour in a particular way. In other words, it supports performance to carry out a task by prompting on what to do and when to do it. A simple sign telling people to enter or not enter a door would be one example. Other simple cases include the dialogue messages that are provided on automated cash machines (ACM). More complex instructions will appear in printed form on the packaging or the instructional manuals for pieces of equipment.
2. Command messages give a very straightforward statement on what is or what is not permitted. ‘Do not enter’, ‘do not smoke’, ‘do not eat or drink’, are examples of command messages. Sometimes they are similar to instructions, but are much more focused on simple statements that refer to high priority items.
3. Advisory messages are somewhat watered down versions of command messages. In some cases, these will be recommendations to avoid a situation, at other times they would be information allowing for the preparation or planning of particular activities. For example, we might be advised that our train is late by a spoken message and we might, possibly, be given an accurate time estimate for when the train will be available.
4. Answers information may be provided in response to a particular enquiry that has been made. This is typical of an interactive information-handling situation, where we have a particular question in mind or degree of uncertainty and we seek information from a source with regard to removing that uncertainty. It turns out that most of the information that is sought from displays is of the answer kind. If we want to know what the time of day is, we look at our watches and clocks to find the answer. If we want to know what speed we are doing in our cars, or what level of fuel we have, we look at the gauges.
5. Historical displays are used to look back at the state of a variable over a period of minutes, hours, days or even years. A graphical representation of road accidents over the last century would be a historical display of information. If we want to know what the temperature fluctuation has been in an office on a daily basis, then specialist devices can be brought in and placed in the office that will give a pen recording over a fixed period of time. It is much easier to see if there is a trend in information if it is displayed in this way; the alternative is to hold in memory a general impression of what the temperature readings have been at a number of points during the day or record them manually on a chart. Gauging the temperature in an office concerns a relatively low risk situation. However, if the concern is with the temperature in a critical vessel in a chemical process, then the temperature trends exhibited over the time are quite important. If the current value is near a safety value, it may well be that it has been near that value for several weeks and is not a critical event. On the other hand, it may have reached that value in the last few hours; looking back at the trend in the information will indicate the rate of change of that variable and whether it constitutes a particular risk to the system.
6. Predictive displays are much more specialised, but increasingly found in complex processes. In the same way that historical data support performance in making a judgment based on the current value, predictive information enables examination of the current value and indicates any likely change in the future. Predictor displays enable better control over vehicles, typically at sea or airborne, and enable smoother transitions from one state to another. They are used in slow response systems where it is difficult to see the immediate effect of an action that has been carried out. Predictive displays will enable a variable to be plotted into the future. The same graphs that are used as historical displays can also be used as predictive displays. If a steady decline in road accidents over time is seen, then the best prediction of the future would be a continued decline. However, it may be that this does not turn out to be the case because of some other factor that can enter into the situation. A predictive value is based on the best evidence available. But in the case of control of dynamic situations, such displays have much to offer in extending the human skill.
Some of these points can now be illustrated by reference to particular areas of display design and the discussion begins with visual displays.