Usability belongs to the most imperative requirements that a product is expected to fulfil, and it has accordingly received much attention in research. It is difficult to gather up the results of these studies, because of the great variability of products and their uses. In the following, we just try to present a few general patterns that can be seen in recent studies of usability. Finally there is a summary of the pertinent research methods.
Usability (or functionality) means the product's suitability to its use. Later on, we shall try to enumerate most important of those factors which make possible using a product or can support or restrict the act of using a product. Before this, however, it is necessary to note that the concept of 'using a product' can mean very different things when we speak of different products.
In the list that follows, a few products are categorized into three classes, each of which have an essentially different mode of use and usability:
The list is not intended for any comprehensive classification of products, because the division active/passive depends on how you look at things. The purpose of the list is just to illuminate the great variety of uses that products can have. 'Using a product' means very different things for different products, and a detailed study of it is possible only by focusing it to a specific class of products. However, a few general concepts and structures exist that can be used in most studies of product use and usability, and they will be discussed in the following.
In analytic study of a concept such as 'usability' the first target usually is to find out which are the most important constituents of the concept. The second task then is to register or measure these from the object of study in order to find out the exact relations between them. For the concept of usability, which was above defined as "the product's suitability to its use", the question is, which are these "important constituents" which together define the usability of the product? If there are among them such properties of the product that designers can manipulate, this knowledge would be valuable when designing new products to be as usable as possible.
Obviously the characteristics of a product can predefine its final usability in a real situation only when there are not too many other factors that affect the use of the product. These other factors could be that the users of the product vary or their way of using the product varies too much. This does not happen with automatic machines: no user can affect their operation. Neither does the use of consumables much vary, and the measurable qualities of a consumable product can thus quite well predict its factual usability.
Things are different for a product of the interactive type: its existing characteristics cannot wholly predict its usability, because the responsibility of getting the product to work is shared: it depends not only on the qualities of the product, but on its user as well. If the product gives less than optimal service, where is the reason? Is it a badly designed or faulty product, an incompetent user, or the fact that the wrong kind of product has been selected for this user? We have to investigate all these possible explanations before we can give a verdict on the usability of an interactive product.
It is evident that usability criteria are different for different types of products, and for interactive products the number of important aspects is larger than for, say, self-operating machines. However, there are several types of criteria that are common to all types of products. On the list below, they are the four first named. Three last-named criteria are only relevant for interactive products. Beside those on the list, there certainly are many miscellaneous criteria specific to the product type.
Below are given a few suggestions for defining and measuring the mentioned dimensions of usability.
Utility, the product's ability to be used to achieve the intended goal, means primarily that the product is able to give exactly the right kind of service, i.e. what is expected from it. You could also say that the product is appropriate or valid for its purpose. In tailor-made products this requirement normally is met almost per definition, given that the customer can afford to buy such a product. That is not always the case, which can give reason to various optimizations.
Another problem occurs with industrial products. A big manufacturer often wants to sell identical products in many countries where the customers' needs and expectations vary. The local climate varies, likewise the voltage of mains, language of dials, meters and menus, the operating system of PC, direction of traffic and the styles of living in general, all of which can lower the utility of a globally standardized product. Nielsen (1993 b, 150) takes up also the question of "social acceptability" and names as an example a French-made teaching package, which does not allow the student to add his own viewpoints to the material - that option is reserved for the teacher. Nielsen thinks that such a design can be socially acceptable in France where the teachers want to maintain their authority, but Scandinavian students would not accept it because it limits their creative thinking.
Note that utility is a property of the product alone, and it does not include the user's ability of using the product (which latter one belongs to the criteria for interactivity, later on).
Capacity means that the output from the product is sufficient to the purposes for which the product is intended. For products of the active type, the word effect is often used, too. Related concepts are efficiency and productivity, both of which can be measured as the quotient of input and output. An efficient product gives the desired benefits and needs for it no more than moderate inputs (of e.g. energy, service and maintenance).
Capacity is easily quantifiable (as a contrast to utility). To measure capacity, effect or efficiency, you have to define the output or benefit from the product with measurable units, often called service units, which are, of course, specific for each type of product. For example, the service unit for laundering machines could be a kilogram of cleaned clothes. The service unit of an utensil can simply be using it once, or an hour of use.
Capacity is an excellent statistic for optimizing a product, i.e. for finding the best relation between benefit and cost, or between benefit and environmental damage. It can also be used for evaluating and comparing rival products or projects, and also for comparing services which give the same benefit to the user. For example, when comparing person cars, we can include in the comparison also other means of transport like bus and train. This can help us point out and evaluate new alternatives which may be environmentally better than conventional products.
For passive products the words "capacity" or "effect" are seldom used, because there is no output from a product other than the product itself. Consumables and components are thus usually measured simply as the number of units, kilograms etc.
Capacity for container-type products is typically measured as volume with certain properties, such as the following:
Usability of buildings, also called their "functionality", has been one of the main points in architectural theory since Vitruve in antiquity. A more exact theory of Functionalism was created first in the years 1920 in the Bauhaus school, and the findings of the studies were often published as standards, which enumerate the space needs of various uses of rooms, often clarified with drawings like on the right (from the book Bauentwurfslehre, 1936 by Neufert). It shows the space needs in a hospital ward room.
Note that the shells that are intended to protect people, in other words buildings and clothes, have beside their physical function of being of right size, warm, watertight etc. often very important function as a message, discussed elsewhere.
When assessing the usability of furniture, aesthetic and semiotic aspects can have some weight, too, but usually the focus is on the ergonomic requirements of use: right dimensions and the limits of their adjustment for persons of different stature, ease of moving the piece of furniture, smoothness of its surfaces, the need for cushioning, load-bearing capacity of seats and shelves, stability against tipping over, and resistance against scratches, heat, alcohol and other liquids.
On the right is an excerpt from General ergonomic requirements for office chairs and desks (from Berglund, 1976 p. 45). The explanations for the picture can be found on the page Theory of Furniture. Of course, the standards also include exact testing methods for all these requirements.
Subjectively pleasing. Products can give many types of pleasures to their owner or user, see a list of these on the page Evaluating Normative Proposals and special pages about the meaning and beauty of products. Many of these pleasures are related only to the properties of the product such as appearance, taste, smell etc., but many concern the use of the product which interests us here.
The pleasures or vexations of use are most marked in interactive products, because the user's contact with these products is closer and more intensive. Some researchers have named as dimensions in the pleasure of use such feelings as security, assurance, confidence, pride, excitement and satisfaction, see the paragraph Usability and emotions, on the page Usability of Interactive Products.
Reliability, i.e. a low rate of faulty products, breakdowns and refusals of service is an important criterion for all types of products. When using this concept, it is essential to define the time perspective. Usual alternatives are:
Note that user complaints are usually most frequent in the running-in phase of a new product when the novelty still suffers from "children's diseases", and once more when the useful life of the product turns to its end. In other words, their frequency more or less follows the "bathtub curve" on the right (from Abbott, 1989 p. 127).
Failures that can involve risk of injury to people are discussed on a separate page Safety of products.
Easy to learn using the product is a quality (like the following three, too) that we can require from interactive products. How it can be defined and measured either with prototypes or with actual products, is discussed in the paragraph Usability as a measurement on the www-page Usability of Interactive Products written by Turkka Keinonen.
Easy to remember is another, comparable concept. After a period of not having used the product, users can have problems in starting to use it again, which can be measured, but these difficulties can be alleviated with lucid design of the interface and a clever automatic help function of the machine. For a discussion, see the link above.
Low rate of user errors. Beside the number of errors that users make, important aspects are the ability to recover from errors and continue normal operation, and the risk of catastrophic errors, which destroy all or much of the user's finished work. Frequency of errors can easily be measured either with prototypes or with genuine products, and usually they will be fewer when the mechanisms for controlling the machine are logical and clearly marked, and when the instructions are intelligible. Keinonen (1998) remarks that the user must be able to immediately notice and undo erroneous actions without doing the whole thing again, and there should be a system for error handling which gives warnings before hazardous commands and gives information about actions that cannot be cancelled. For a deeper discussion, see the link above.
UsabilityNet: A European Union project that provides usability and user centred design resources to practitioners, managers and EU projects.
August 3, 2007.
Comments to the author:
Original location: http://www2.uiah.fi/projects/metodi