Why designers can't understand their users
Verhoef, L.W.M.
Citation
Verhoef, L. W. M. (2007, September 19). Why designers can't understand their users.
Human Efficiency, Utrecht. Retrieved from https://hdl.handle.net/1887/12347
9. Visual distance
9.1 What is ‘distance’
The second property is distance. Knowing all element properties of an element of the interface one still does not know its value on the property distance. There are two interpretations of what is meant by the distance of an element:
Designers measure the distance of an element to some physical fixed reference point in the interface. The position for a menu or a window on a computer screen is specified from the frame of the screen, e.g. from the top left hand corner of the screen or window. To determine human performance one should alternatively measure the distance of an element from some psychological fixed reference point. When human perform- ance is relevant the focus of the user is a good point. When human function is the point of reference the position for the last button pressed or the position for the cursor is a practical choice for a visual and motor point zero. None of these is ever the top left hand corner of the screen except, possibly, when one starts using an interface for the very first time.
Using a position defined by the needs of the user as reference point is uncommon and when suggested surprisingly difficult to get accepted.
Having a fixed position referring to the (physical) interface gives a very
‘quiet’ impression for those who do not have to work with the interface.
The menu is always, for instance, at the top of the screen. This makes technic al specification easy and the progamming task is straightforward – all screen information can be presented relative to some predefined (0.0) point such as the top left corner. Having a fixed pos ition referring to point zero of the user rather than the screen, results in an interface presenting the information at the point of fixation of the user. The menu can now be anywhere on the screen. For those not actively working with the interface and consequently, usually unaware of the user’s point zero, this gives a rather chaotic impression as they look over the user’s shoulder. For the user, however, presenting information wherever on the screen but always
at point zero for the user, minimises the visual distance property, and consequently for the user is the most ‘quiet’ presentation possible, despite what a non-user might imagine.
9.2 ‘Distance’ in other function fields
For the performance of human function, distance is relevant because over small distances other physiological functions are used than for large distances. The property distance maps onto physiological structures of the whole range of human functions.
• With finger muscles a human can press a button that is immediately under a finger. When the distance between the current button and the next one to press is large, finger movements are not sufficient and wrist or arm movements are needed to travel to the next button.
Changing distance will have an effect on human motor performance.
• When information is presented within the visual fixation area, all information is present for the reader at the same time. When infor- mation is presented outside the fixation area eye movements are needed to identify that information. Changing visual distance will have an effect on human visual performance.
• A similar structure applies to human memory. Information in Short Term Memory is immediately available to the user whereas infor- mation in Long Term Memory is further away, is stored in another type of physiological structures and takes longer to retrieve. When Short Term Memory is overloaded and information is pushed further away into Long Term Memory, this will have a detrimental effect on human memory performance.
• For human thinking it is more difficult to identify physiological structures corresponding to the field variable distance. However, the concept of comprehension might suffice. Comprehension might be an example of cognitive distance. When there is a large distance there is no understanding and when there is full comprehension distance is low. Greater ‘conceptual’ distance requires more effort to strive after meaning (Bartlett, 1932). This too will have an effect on human per-
9.3 What is ‘visual distance’?
Visual distance can be interpreted in two ways. In the first interpretation visual distance is the length of the line between the perceiver and an object, commonly referred to as the line of sight. When the position of an object along this line is changed, the perceived size is also changed. Size has already been discussed in Chapter 8. The literature on visual percep- tion shows how this kind of distance is regularly inferred from size (cf.
the Ames room where it goes very wrong).
In the second interpretation visual distance is the length of the line between the visual focus of the perceiver and an object, how far away an object is from the point of focus. The term ‘far’ denotes that we also use a distance term to describe this visual eccentricity. This line can also be expressed as a visual angle . In this interpretation of visual distance, the perceived size remains equal when this distance towards the visual focus changes, as the eye shifts. When there is more distance between the focus of the perceiver and the object, the object is projected more eccentrically on the retina. In the periphery the retina is less sensitive than in the centre. Consequently, acuity decrease as visual distance increases. For a monitoring task the subjects will have more difficulty in noticing infor- mation and for an identification task the subjects will have more diffi- culty in identifying the visual information. In order to compensate they will have to shift focal attention to the position in question, assuming that they have noticed a significant event in the first place.
The Dutch saying: ‘Verder kijken dan je neus lang is.’1 illustrates that the concept visual distance is ill understood. ‘Verder kijken dan je neus breed is.’2 would be better. In the saying there is confusion on the property level and on the human function level. In most cases not ‘visual search’ is meant but ‘cognitive search’. When saying ‘Verder kijken dan je neus lang is’ the Dutch mean: ‘Verder denken dan je hersenen breed zijn.’3
1 In English: ‘Looking past the end of your nose.’
2 In English: ‘Looking past the width of your nose.’
3 In English: ‘Looking past the width of your brains.’
9.4 Why the term ‘visual distance’
In common language, distance is indicated using the English word ‘far’
or the Dutch equivalent ‘ver’. This terminology confuses property (dis- tance) and a value for the property, in this case a high value. This confu- sion obscures the ‘property - value’ structure of elements of interfaces and should be avoided in a professional context.
In the literature, several terms are used to denote the distance between the focus and visual object.
In the psychological literature the term ‘peripheral vision’ is used (Bartz, 1976, 1992). ‘Peripheral vision’ is a property of human beings and not of elements of systems. Using a term indicating a property for humans is rather confusing in an interface design context when meaning a property for the interface. The corresponding system property could be defined as
‘peripheral visual position’.
Sanders and McCormick (1992) state that a component should be located at an optimum location in order to serve its purpose. To arrive at that optimum location the following principles should be applied: importance principle, frequency-of-use principle, functional principle and sequence- of-use principle. For visual perception this means that a component should be in the ‘optimum viewing area’. ‘Optimum’ is not a singular concept. It includes a particular position and an evaluation of that posi- tion taking account of the task. Therefore this concept cannot be used.
The same applies to ‘area’. An ‘area’ can be specified in several ways whereas a distance is more fixed as there are only two points to define.
In an ergonomic context several authors do not discuss visual distance, suggesting this is not a relevant concept for interface design (Bastiaans, 1987; Bekking, Elburg and Weerdmeester, 1989; Odo, 1985).
• Wickens (1987) is more specific and used the term ‘spatial separa- tion’ and ‘(close) proximity’ when discussing the position for indic a- tors on a panel. The adjective ‘visual’ is to be preferred to ‘spatial’.
‘Spatial’ suggests a 2 or 3 physical dimensional space model which
• In a perceptual psychological context ‘proximity’ is used by Barnett
& Wickens, (1986, 1988) and Woodward (1972). Barnett & Wickens define proximity as ‘closeness in space or time’. In a study on infor- mation on patient package inserts van der Waarde (1999) uses ‘prox- imity’ for the distance between graphic components. There are two comments on the use of the term ‘proximity’. The first comment is that none of the authors mentioned include the position of the per- ceiver in their definition. Consequently, they are referring to a dif- ferent concept. Nevertheless, one could include the position of the perceiver in the definition. This brings us to the next comment.
‘Proximity’ implies low distance and, as such, is not appropriate to indicate the whole range of distances, which includes both low and high distances. Consequently, when using the term ‘proximity’ a large distance would be indicated as a large proximity. When defined as such it is correct of course, but it sounds like a paradox.
• More or less the same applies to ‘eccentricity’, a term used by Erik- sen (1953) and Cole and Hughes (1984). Eccentricity specifies visual distance from one end of the dimension and not the dimension itself.
In addition, eccentricity can be interpreted in a technological context as eccentrically in a physical field, e.g. on a computer screen.
Figure 1. Some positional variations for the presentation of numbers to be compared
123 124 123
124
1 2 3 1 2 4
1 1 2 2 3 4
A final consideration for using the term ‘visual distance’ is that it is more obvious that ‘visual’ can be replaced by other human functions. There is a motor distance, an auditory distance (stereo, determined by stereo cuing), a memory distance and a cognitive distance. In this way the terminology meets the requirements for cognitive structure set in Chapter 11.
9.5 Which ‘visual distances’
In the horizontal plane, visual distance has a range that starts at approxi- mately -57 degrees left, to approximately +57 degrees right. Burandt (1978) concluded that this figure should be 52 degrees to both sides.
In the centre of the range the number of receptors is highest and conse- quently, visual acuity is at its best at approximately -2.5 degrees to +2.5 degrees. When the task is reading, the identific ation of approximately 15 characters is possible in one eye fixation. Nes (1986) concluded that characters further than 10 positions from the centre are not readable and Schiepers (1976) concluded that this figure should be 6 to 7 characters.
Cole and Hughes (1984) studied the effect of visual distance on a moni- toring task. They found that when visual distance increases conspicuous- ness decreases. They also found that visual distance has more effect on conspicuousness than do refle ction or size. In analysing the causes for way-finding problems for patients suffering from dementia Passini et al.
(1999) concluded that the spatial disposition for messages on signs was one of the main causes for way-finding problems. A very interesting experiment was done by Woodward (1972). In the literature, at that time, there was a debate on how to present figures to be compared. In an interface technological tradition, all kinds of positional presentations were investigated using positional positions as experimental conditions without providing conclusive results. Figure 1 presents some examples of positional presentations for comparing the numbers 123 and 124. In that case the correct answer is that the numbers are not the same. The next page presents how the debate was ended using an interface property approach.
Woodward (1972) supposed that it was not positional presentation that is the fundamental variable but visual distance (proximity). When proximity was low, comparing performance would be best. See the figure 2. An experiment confirmed his expectation and the debate ended. Table I summarizes the range for values for visual distance.
Figure 2. Visual distance and the presentation of numbers to be compared
123 124 123
124
1 2 3 1 2 4
1 1 2 2 3 4
Tabel. I. The range for values for visual distance Range for visual distance
Visual task
Physiological left threshold
Psycholo- gical left threshold
Psycholo- gical right threshold
Physiological right thresh- old
Identi- fication
<-2.5° left, outside fovea
>-2.5° left, inside fovea
<2.5° right, inside fovea
>2.5°right, outside fovea not readable 10 to 20 characters readable
in one fixation
not readable
Moni- toring
<-60°left, outside range
>-2.5° left, inside range
<2.5° right, inside range
>60° right, outside range not notable notability very likely not notable
9.6 Experiment 1: comparison ‘Out of order’
9.6.1 Introduction
Previous experimental observations indicated that too great visual dis-
tance might be the cause of several of the problems of passengers had in using the 80-destinations ticket vending machine of Netherlands Rail- ways (Verhoef, 1986). Some of these indications are the following.
• 11% of the passengers pressed destination instead of the white button on the right (See Figure 4). It was suggested that, because of visual distance between the destination and the button, passengers reading the destination did not notice there was a separate button.
• 0.3% of the passengers bought a ticket to the wrong destination. Half of these passengers selected a destination that was one line higher or one line lower than the destination they intended to press. While trav- elling the distance between destination text and button they might have ‘fallen of the line’.
Figure 3. Visual distance between label left and button right
Figure 4. Visual distance between class label and the button for evening return (avondretour)
• There were several types of observations that have shown that pas- sengers did not press the button of the class they intended to press.
For instance, 1% changed class before payment and at least 0.15% of the passengers bought a ticket of another class than they intended (n=426). The distance between class indication and button was be- tween 2 and 11 cm. See Figure 5.
• 50% of the passengers do not notice an ‘Out of order’ message and they proceeded to select their destination (n=426). The visual dis- tance between the first step: ‘selecting a destination’ and the second step ‘noticing the text ‘Out of order’ (in red in the price display)’, was approximately 40 cm (See Figure 6.).
In order to be able to evaluate the psychological benefits and the techni- cal costs of reducing visual distance, an experiment was carried out in which the position for the ‘Out of order’ information was varied.
From the theory above, it can be concluded that conspicuousness can be increased by reducing visual distance. Technically the visual distance of
‘Out of order’ can be reduced by positioning the text closer to the list with the destinations. It had been observed that most passengers start by selecting a destination.
This hypothesis was investigated. It was expected that ‘Out of order’, positioned immediately above the list with destination will be noticed more frequently than the usual position for ‘Out of order’ (Buiten dienst), which was to locate it just above the price display.
9.6.2 Method
‘Out of order’ was presented on the 80-destinations vending machine above the price display with black characters on a red background. The text was mounted on a triangular bar called a toblerone. The two other sides of the toblerone could be used to display the texts ‘Temporarily no change given’ or ‘Change given’.
Figure 5. Large visual distance between ‘Out of order’ (red label on the machine at the right) and the list with destinati- ons on the 80-destination vending machine
smaller visual distance, which was located above the list with destina- tions. In that case the visual distance between ‘Out of order’ and the nearest destination was 2.5 cm (first line) and the furthest destination 51 cm (bottom left and bottom right).
In the experiments it was observed whether passengers noticed that a machine was out of order. This was rather easy to observe with an un- aided eye from a few meters distance. Any passenger, who fails to notice that a machine is out of order, will proceed to press a button. A second indication is the time between pressing a button and noticing ‘Out of order’. This is easy to observe using a stopwatch too, because a passenger steps away to try another machine or gives up and goes to the ticket window.
On Utrecht Central Station 894 passengers were observed trying to buy a ticket in one of three vending machines. In accordance with a random scheme one of the three machines was chosen to be set ‘out of order’.
Figure 6. Small visual distance between ‘Out of order’ (in red, above list of destinations) and de list with destinations on the 80-destination vending machine
9.6.3 Results
When visual distance is large, 6.4% of users continues and presses a button, and when visual distance is small only 1.9% presses a button. A Mann-Whitney statistical analysis showed that this difference was sig- nificant (U=94849; p=0.001). When visual distance is larger, 4.14 sec- onds is required and when visual distance is small 0.85 seconds is re- quired to notice that the machine is not working.
9.6.4 Discussion
The results were significant and as expected. The results are consistent with the findings of Connell (1998) who reported in an error observation study four occurrences of passengers not noticing that no change was given. The total number of errors observed was 122, including technical failures. The machine investigated was the British Rail version of the same vending machine used in this experiment. In the British machine the information ‘Out of order’ is presented at the same location as is ‘Out of order’ on the Dutch machine. While the result seems to be correct to the point of triviality, several European transport companies, including Netherlands Railways and British Rail, had managed to implement a design that clearly induced non-fatal errors and, presumably, frustration amongst their customers. It is possible that the extra cost consequences of the better design may have over-ruled the requirements for usability.
When visual distance increases, conspicuousness decreases. Theoreti- cally, however, the differences found could be caused by other factors.
The two experimental conditions not only differ in visual distance, but also in colour and luminance contrast. Experience levels of the passen- gers had an effect on performance too. Initially 50% of the passengers noticed ‘Out of order’ (Verhoef, 1987a). After three years this figure had decreased to 6.4%. The passengers had learned to look carefully at the price display to see if a machine was out of order, but the figure of 6.4%
still remains much higher than the 1.9% experienced when distance was deliberately made low for inexperienced users. Furthermore passengers had to learn an extra step to compensate for an ergonomically poor
9.7 Experiment 2: comparison ‘Insert this way’
9.7.1 Introduction
A banknote or a magnetic card can be inserted in a slot in four ways. The top can be up or down and the left side of the banknote or the card can be on the left of the slot or on the right of the slot. Usually the correct way to insert a banknote or a card is not indicated on machines. Of course, one solution is to accept all four ways of inserting a banknote or a card as Lewis (1987) suggested. However, this is sometimes technically impos- sible or too expensive. The most common solution is to print instructions about what to do, making the customer work to provide the correct solution. However, an even better solution is also possible, when visual distance is taken into account.
Bastiaans (1987) does not mention the ‘Insert this way’ problem in his overview of ergonomics on electronic funds transfer point of sales, nor does Okada (1985) in his report on the introduction of automatic ticket vending machines accepting magnetic cards in Japan. A personal com- munication with a Japanese train ticket vending machine expert Odo (1985) revealed that when correct insertion is a problem, this problem is solved by Japanese Railways using a written instruction; when this written instruction has no effect, the solution is to use more written instructions. Zwaga (1988), however, observed that when buying a ticket in a Washington metro vending machine, 20% of the pa ssengers who
Figure 7. Visual distance between ‘Instructions insert banknote’
and the slot on the 80-destination vending machine
used the machine had problems with the notes. Bekking, et al. (1989) observed that 35% of telephone card users had problems inserting a telephone card in the slot correctly. (This is close to chance behaviour if they start with an initially correct assumption about top=up) It is worse than the bald figures suggest when one realises how close users are to the chance base-rate of success in completing their transaction.
Figure 7 shows how the 80-destinations vending machine informed passengers about how to insert a banknote. Above the slot there was an empty field that easily could be used for instructions. As noted before, it had been observed that such instructions were insufficient. The results do not contradict those of Zwaga and Bekking et al. Of all the banknotes inserted, 42% (Verhoef, 1987a) were not accepted by the machine. In addition, it was observed that 70% of the time needed to buy a ticket was needed for insertion of the banknote. This is clearly a situation where improvements to any sort would have value for users, both in time and reduction in frustration, as well as reducing the number of machines required to process the demand.
Some of the problems passengers have with banknotes have a technical reason. The 80-destinations vending machine regularly refused undam- aged notes that were inserted in the correct way, presumably due to problems with the recognition system. All the remaining problems with the banknotes were due to the information presented on how to insert a note. To perform the insertion task correctly, the passengers should know that a banknote should be inserted in only one particular way. Secondly, they should be informed about where to insert the banknote. Finally, there will always be some distance between the instruction and the focus of the passenger when inserting a banknote.
Bekking, et al. (1989) suggest solving the problem using visual forms; i.e.
using arrows and symbols. There is also another option. When inserting a banknote, the visual focus of the passenger is on the slot. In accordance with the theory mentioned above, and taking account of the experiments of Cole and Hughes (1984), it is expected that conspicuousness increases
note has to be completely unfolded, flat, with no folded corners. It is impossible to perform this task without focussing in the slot.
In addition, to the information about ‘how to insert a banknote’ the passengers should first have been informed about the fact that the bank- note must be inserted in a particular way. The conspicuousness of that information is also dependent on visual distance. That is why the text
‘Insert this way’ is positioned on the banknote. The position for this text is the same in all conditions.
We tested the hypothesis that reduction of visual distance increases the conspicuousness of the information on how to insert a banknote. It was expected that more passengers would insert a banknote in the correct way when the information is pos itioned in the slot than when that information is presented above or to the side of the slot.
9.7.2 Method
The experimental conditions are as follows. In one condition information is presented as is usual on such devices i.e., no information at all on either where to insert, how to insert, or that how to insert is critical. Secondly , two conditions that both inform the passenger about the orientation of the banknote. The difference between the conditions is visual distance. In the high visual distance condition information was presented above the acceptor and visual distance was approximately 12 cm. In the low visual distance condition information was presented in the slot and visual distance was 0 cm. Figure 8 gives an overview of the conditions.
In 1987 it was observed how 822 rush hour passengers and 529 non-rush hour passengers inserted a DFL 10 note. It was noted which side of the note was up and which side was inserted first. This observation can easily be performed with an unaided eye when standing a few meters aside from the passenger.
Figure 8. Experimental conditions
Large distance Large distance Small distance Conventional way Experimental way I Experimental way II
Distant informa- tion
Proximal informa- tion
No information on where to insert
Information on where to insert
Information on where to insert No ‘Insert this way’ ‘Insert this way’
presented
‘Insert this way’
presented No information on
how to insert
How to insert is presented
How to insert is presented
9.7.3 Results
Table II shows that information in the ‘small distance’ condition is noticed in 78% of the cases, whereas in the experimental condition large distance this figure is 69%. A Kruskal-Wallis analysis showed that this difference is significant (Chi²=4.29; p<0.01). The percentage of correct insertions for the conventional way of presenting this information is only 51% - in fact people insert notes at chance levels under these conditions.
However, as mentioned above, there were more differences between this condition and the small distance condition than just visual distance.
9.7.4 Discussion
This second experiment on visual distance confirms the hypothesis that decreasing visual distance increases conspicuousness. It should be noted that there were some differences between the high and the low visual distance conditions that will also have had an effect on conspicuousness.
• Luminance contrast for the high visual distance condition was higher, a white drawing on a white background. In the low visual distance there was less contrast as the white drawing was positioned on an aluminium (light) background.
• Physical dimensions of the information in the acceptor are smaller than the information above the acceptor. Perceived size for the low visual distance information was also smaller. The information above the acceptor was posit ioned vertically and therefore more or less ver- tical in the line of sight. The information in the acceptor was posi- tioned horizontally, thus resulting in a line of sight less than 90 de- grees. The effect of this can be seen clearly in the pictures in Figure 9 Table II. Correct insertions
Large distance Large distance Small distance Conventional
51%
Experimental - Above 69%
Experimental – In the slot 78%
where the information is in the slot. Because of the angle the infor- mation is hard to read, in the figure as well.
• The information above the acceptor was always visible. The informa- tion in the slot was, from a certain moment, obscured by the bank- note.
• The white arrow, close to the information above the acceptor, also has a positive effect on noticing the high visual distance. When inser- tion of a banknote was possible this arrow flashed.
All these differences will have improved the conspicuousness of the large distance condition, so the difference between 78% and 69% is larger than the descriptive statistics might suggest. Conspicuousness is not a simple function of contrast or flashing but is determined in the context of the task and the ability to support focal attention. Figure 9 shows the design as it was finally implemented by Netherlands Railways.
Figure 9. Information on how to insert bank note, as it was imple- mented by Netherlands Railways
The first and the second experiment suggest that visual distance is a relevant concept when designing vending machines.
9.8 Generalization of knowledge
9.8.1 Interface technology independence: the 360- destinations vending machine
When the 80-destinations vending machines were taken out of service, Netherlands Railways introduced a machine that sold tickets to approxi- mately 360 destinations. Knowledge obtained earlier on visual distance was applied in several ways.
Selecting a destination by a user was no longer carried out by pressing the button located right next to the destination but by entering a code for that destination. Although this is a quite different way of entering the destina- tion, visual distance again plays a role in determining human perform- ance. As Figure 10 shows, the visual distance between a destination and its code is kept as low as possible. The code is not right aligned but is positioned immediately after the destination. Interface technology had changed substantially from using a button to denote a destination to selecting and entering a code for a destination but psychological know- ledge about interfaces that was obtained by investigating quite different interface technology still remained applicable.
Figure 10 Minimal visual distance between destination and his code on the 360-destinations ticket vending machine
Another difference between the 80-destinations machine and the 360 destinations machines was the method of payment. The 80-destinations machine accepted coins and banknotes whereas the new machine ac- cepted coins and electronic cards (pinpassen). Here again, interface technology changed but interface psychological knowledge that had been obtained by investigating old interface technology remains applicable, as can be seen in Figure 11.
9.8.2 Interface technology independence: the touch-screen vending machine
One of the main problems of the two generations of hard button train ticket vending machines discussed in this thesis proved to be the routing – i.e. steering the user through the sequence of sub-tasks that have to be performed to acquire a ticket. On hard button vending machines, for technical reasons, it is impossible to install controls in positions that are the best from a psychological point of view. The problem was tackled by presenting the controls belonging to one task very conspicuously; in one field and directing eye movements using carefully designed flashing
Figure 11. Minimal visual distance between destination and his code on the 360-destinations ticket vending machine
taken by the user are presented with minimal distance. The next step is positioned immediately adjacent to the fixation point of the previous step.
Other train ticket vending machines did not apply this solution (e.g.
Belgium and French railways).
Nevertheless routing problems were occasionally observed (0.08%, n=79, Verhoef 2000). However, the concept of visual distance also provided an explanation for these routing errors.
All routing errors were made after task 2, selecting a destination. When starting step two, the focus of the passenger is at the upper left hand corner of the screen. Suppose a passenger selects Hoensbroek at step two (see Figure 13), then their focus is on the upper right hand corner. After having made that selection, step five is presented at that position, seduc- ing some passengers into skipping steps three and four.
Figure 12. Small visual distance to next step on the screen ticket vending machine
Product, Destination, Class, Price, Number, step 1. step 2. step 3. step 4. step 5.
Figure 13. Large visual distance between the task route on the top and destination ‘Hoensbroek’
9.8.2 Domain independence: a coffee vending machine
Figure 14. ‘Place cup first’ and visual distance on a coffee vending machine
Designer’s proposal, Proposal after consulting the author,
large visual distance small visual distance
‘cup first’ and coin slot. ‘cup first’ and coin slot.
Another type of machine where visual distance has been applied was a coffee vending machine. Some of these machines do not have cups and the user has to position a cup under the coffee outlet; failure to do this results in coffee on the shoes of the user or in the drain. The left side of the Figure 14 shows how the designers of that machine initially presented information to prevent this happening. As it is unusual on vending ma- chines to have the users themselves positioning their cups, it was ex- pected that users will often forget to place a cup (Reason, 1990). As it is almost impossible to insert coins blindly, it was assumed that information near the coin slot would be noticed more frequently than information elsewhere on the machine. Therefore the proposal was made to reduce visual distance of the information ‘Place cup first’, by positioning that information as close as possible to the coin slot. The right side of Figure 14 shows the final design.
Another application of visual distance in this design is the position for controls needed to obtain one or more cups in one operation, without the machine providing a cup or can. For instance, filling personal china with 5 cups of coffee. In the original design the controls for this function were close to the coin slot. It was expected that inexperienced users mostly requiring coffee and a one cup, not bringing with them a cup and not expecting a machine to have a ‘fill your own cup/ can with the amount of more than one cup’ function, inadvertently would press the buttons for this function. No experiments were done to confirm this expectation.
Theory and practical experience were rather consistent in the outcome of such an experiment. Therefore, the author’s client, Van Nelle decided to increase ‘visual distance’ between coin slot and the ‘more cups’ controls and to reduce conspicuousness by reducing contrast.
9.9 Conclusion
A theoretical analysis including physiological and perceptual psychology suggested that visual distance is a fundamental property of a field with visual elements.
focus on a search direction culturally imposed (top left to bottom right).
Visual distance generally is not mentioned (Bastiaans, 1987; Bekking, Elburg and Weerdmeester, 1989; Odo, 1985).
Within the machine investigated the concept of visual distance could be applied to several interface technologies of the machine (several types of labels for buttons to press and unexpected information (out of order)).
When interface technology was changed from ‘panel with buttons and lights’ to ‘touch-screen’ technology visual distance was still applic able and was one of the main principles for the design. Unexpected problems with the operations could be understood using the concept visual dis- tance.
The concept was applied in several domains such as: interfaces for selecting a destination, the status of the machine, payment and drink vending machines.
The general conclusion is that theoretical considerations on visual dis- tance, an effect on human performance found in several experiments, and interface technology and domain independence, support the conclu sion that visual distance is a fundamental property of the visual field.
