The context of human error has been considered as one of the most important issues because of its enormous effect in both human and economic aspects. Accordingly, all humans have experienced human error. Errors can be made when people interact with machines and other complex systems. Herein, people have a tendency to do things which are contrary to their expectations and intentions. Depending on the complexness of a certain system and the human intentions that interacts with it, errors can be anything from a discomfort (which are often unnoticeable) to a actual tragedy or calamity.
Human error an happen in the plan, design, management, operation as well as maintenance of complex systems characterized by the modern life. Since humans, depend largely and increasingly on these systems to have a better living, it is transparent that human error has a potentialities and has a frequent source of risks to human life and welfare as well as the natural environment (Senders & Moray, 1991). Errors in human can result in both human and economic cost.
In line with the human factor, errors can result to anxiety, depression, disappointment and stress at work which may result in the inability of the individual to finish their tasks or responsibilities effectively and mat also lead to negative emotions. In terms of the economic aspects, the economic costs largely depend on the error numbers as well as the time spent in the diagnosis of such error and the recovery. Because of these enormous effects, researchers and scholars has spend their time in analyzing and understanding the context of human error.
There are various areas or fields in which human can occur, and one of these is the errors in Human-computer interaction (HCI). Primarily, the main goal of this paper is to determine the relations of human error with Human-computer interaction. Part of the discussion of the topic is the definition of human error and the types of human errors. This paper also attempts to provide practical examples of human errors and the future perspectives, reduction and prevention of human error. Pertinent conclusion will also be provided to summarize the findings of this report.
For many years, many scholars and researchers has been trying to determine the nature and definition of human error. Human error is regarded as departure from expected, desirable and acceptable practice on the individual part which can result in undesirable and unacceptable results or outcome. Accordingly, human error has played a very crucial role in various large-scale hazardous and life-threatening events and occurrences (Reason, 1990). Some researchers believed that human error might not be an accident because of itself but it might be a result of multiple factors which are not able to control (Reason, 1990).
As mentioned earlier, human error can error in various fields from military, politics, economic, computer systems, medical fields, engineering and others. It is said that errors, specifically the human errors are rare occurrences as compared with correct actions or successes. The kinds in which errors can take are restricted and may appear in variety of contexts, but only a few behavioral approaches seems to be accountable for all such errors (Rizzo, Bagnara & Visciola, 1987). Nonetheless, comparable types of errors can be seen in perception, speech, problem solving, decision making and actions.
However, human errors are less likely to happen tasks are automatic or skilled based such as driving, than when the action call for a rule or process to be considered. It is noted that those actions which call for the utilization of knowledge to solve new issues and conflicts are regarded as the most vulnerable tasks to human errors. The reason for this is because knowledge, like principle, should be translated into correct course of concrete actions, hence, error can be encountered in this translation process (Bogner, 1994).
An error has always been attributed to a mismatch of the computer system to human nature and capabilities (Rasmussen, 1987). To illustrate, controversies, technical problem, lost files, or other mistakes in relation to the use of computers are often blamed to either the machine or the human. Errors paved the way to the creation of back up files, for example, and the improvement and innovation of machines to assist humans in working or accomplishing their tasks faster and easier. Human error definition has not only been the subject of the interest of the scholars and researchers but also the types of human error.
According to some scholars there are two types of human errors in accordance with the systematic causes of failure. These types of human error include the active errors and latent errors. As Reason (1990) has noted, active errors are errors whose effects are immediately felt. Such errors include syntax error which prevents the efficient compilation or invalid algorithms. On the other hand, Reason (1990) has defined latent errors as errors in which adverse consequences may lie dormant within the system for a long period, and only becomes obvious when such is combined with other factors to hinder the defenses of the system.
Aside from Reason, Rasmussen (1982) has also been able to provide other types of human error and these include knowledge based, rule based and skill based human error (see Appendix 1). The taxonomy of human errors is needed when discussing or understanding errors that people encounter with the use of machine. Examples of errors include functionality problems, usability problems, inefficiency, and interaction problems. Functionality Problems Functionality problems refer to the mismatch between the task and the program.
This mismatch occurs when one’s plan has not been accomplished or one’s goal has not been achieved with the use of particular computer program. Functionality problems are classified according to technical criteria like origin in the process of software development and consequences for the action process (see Appendix 2; Brodbeck et al. , 1992). Usability Problems Although the functionality of a computer program is sufficient for a certain task, errors still occur. They can be caused by a mismatch between user and computer, called mismatch of usability.
From an action-theory perspective, mismatches of usability can be differentiated according to steps in the action process and different levels of action regulation. A common hypothesis in action theory is that actions are goal-oriented (Volpert et al. , 1987). Within this approach, the action process comprises goal and plan development, the execution of actions as well as monitoring, and feedback processes (Lewis & Norman, 1986). Similar to Hacker (1986), three levels of action regulation are distinguished within the framework of hierarchically organized action plans and goals.
There are at least three aspects of the knowledge base for regulation: knowledge of facts, knowledge of procedures, and understanding in the sense of mental models. This knowledge is used to develop goals and plans. Errors arising from usability problems include errors in knowledge, thought, memory, judgment, errors on the level of flexible action patterns habit, omission, recognition, and sensimotor (see Appendix 3. Knowledge errors occur when one is unable to do a task with the computer because one does not know certain commands, function keys, rules, and so forth.
Thought errors occur when goals and plans are inadequately developed or when wrong decisions are made in the assignment of plans and subplans although the user knows all the necessary features of the system. Memory errors occur when a certain part of the plan is forgotten and not executed, although the goals and plans were initially correctly specified. Judgment errors appear when one cannot understand or interpret the computer feedback after an input. Errors on the level of flexible action patterns occur when well-known actions are performed. Habit errors imply that a correct action is performed in a wrong situation.
Omission errors happen when a person does not execute a well-known subplan. This is most likely when the person is interrupted in an action plan. Recognition errors appear when a well-known message is not noticed or is confused with another one. Finally, sensorimotor errors are placed at the sensorimotor level. There is only one category here because, at this level, it is empirically difficult to differentiate among planning, monitoring, and feedback (see Appendix 4; Brodbeck et al. , 1992). Inefficiency Errors and inefficient behavior have a large conceptual overlap.
A detour to reach a goal may be conceptualized as an inefficiency but also as an error, because usually one’s goal is to proceed in the most straightforward manner. If there is any differentiation between the two terms, inefficiency could be formally defined as any deviation from an optimal action path (Volpert et al. , 1981). However, this conceptualization has its difficulties: it may be more cost effective, in the psychological sense, to use a strategy that is inefficient in the formal sense because setting up plans and differentiated calculations for them also imply psychological costs (Schonpflug, 1985).
Interaction Problems Interaction problems suggest that human error occurs not because of individual problems with the computer, but because of the mismatch between individuals. Although the individuals’ actions are more or less correct, an error occurs because of an organizational lack of coordination, unclear task allocation, or because of lack of communication between individuals (Brodbeck et al. , 1992).