Human-error taxonomies

Swain s taxonomy of human errors is rooted in the field of human reliability. It applies a mechanistic view of the human operator in the sense that the operator is regarded as a systems component in line with machinery. Human reliability is thus defined as the probability that a person first, performs an activity required by the system correctly and within a required time period and second, does not perform an extraneous activity that may degrade the system (Rosness in Suokas and Rouhiainen, 1993). Swain s classes of human errors constitute observable behaviour in relation to a planned work sequence. There are data banks on the frequency of typical human errors in different systems contexts. Data on human-error frequencies are applied as input to reliability models to calculate the overall probability of systems failures and accidents. 

Basis for classification Classes of human errors Source 

Cognitive stage Skill-based error Rasmussen et al.. 

Surry s decision model and Rasmussen s framework for cognitive stages have served as bases for the development of human error taxonomies. Reason distinguishes between errors at different performance levels  

Reason s error taxonomy that relates to the organisational hierarchy has been discussed in Section 5.2. 

---

Itoh, K., Omata, N. and Andersen, H.B. 2009. A human error taxonomy for analysing healthcare incident reports Assessing reporting culture and its effects on safety performance. Journal of Risk Research, 12(3-4), 485-511. 

We will revert to the different types of human-error taxonomies in Chapter 8. 

Rasmussen, J. (1982). Human Errors A Taxonomy for Describing Human Malfunction in Industrial InstallaHons. Journal of Occupational Accidents 4,311-333. 

This paper deals with the classification of an error type and the characteristics of human errors by each error type for the test and maintenance failures that have led to implanned reactor trips in Korean nuclear power plants. The classification of hmnan errors was basically performed on the taxonomy of Reason s basic error types (Reason 1990). Characteristics of the test and maintenance errors include the major contributing factors or error modes, and predictivity or identifiabil-ity of a potential erroneous action from the viewpoint of a human error prevention or management. 

Rasmussen, J. Human errors. A taxonomy for describing human malfunction in industrial installations. Journal... 

Rasmussen, J. 1981. Human Errors. A Taxonomy for Describing Human Malfunction in Industrial Installations. Ris0-M-23O4. Rosldlde, Denmark Risp National Laboratory. Rasmussen, J. and Taylor, J. R. 1976. Notes on Human Factors Problems in Process Plant Reliability and Scfety Prediction. Ris0-M-1894. Roskilde, Denmark Ris0 National Laboratory. 

For error analysis purposes, human error identification (HEI) approaches such as SHERPA (Embrey, 1986) can be used see Chapter 14 for a full description of SHERPA. SHERPA uses a behavioral taxonomy linked to an error mode taxonomy and is applied to an HTA of the task under analysis. The behavioral and external... 

Taxonomy-based HEI techniques use external error mode (EEM) taxonomies to identify potential errors within complex sociotechnical systems. Typically, EEMs are considered for each component step in a particular task or scenario to determine credible errors that may arise during human-machine interaction. Techniques such as the Systematic Human Error Reduction and Prediction Approach (SHERPA) (Embrey, 1986), the Human Error Template (HET) (Stanton et al., 2006), the Technique for the Retrospective and Predictive Analysis of Cognitive Errors (TRACEr) (Shorrock and Kirwan, 2002), and the Cognitive Reliability and Error Analysis Method (CREAM) (Hollnagel, 1998) all use domain-specific EEM taxonomies. Taxonomic approaches to HEI are typically the most successful in terms of sensitivity and are also the least expensive, quickest, and easiest to use however,... 

Error identifier techniques use prompts or questions to aid the analyst in identifying potential errors. Examples of error identifier prompts include Could the operator fail to carry out the act in time or Could the operator carry out the task too early or Could the operator carry out the task inadequately (Kirwan, 1994). The prompts are linked to a set of error modes and reduction strategies. Although these techniques attempt to remove the reliability problems associated with taxonomy-based approaches, they add considerable time to the analysis because each prompt must be considered. One example of an error identifier HEl technique is the Human Error Identification in Systems Tool (HEIST) approach (Kirwan, 1994). 

Shappell, S.A. Wiegmann, D.A. 1997 A Human Error Approach To Accident Investigation The Taxonomy Of Unsafe Operations. The International Journal Of Aviation Psychology, 7, 269-291. 

With regard to the aforementioned taxonomy human rehability analysis is a semi-specific method. This technique can be used to avoid faults and identify situations that are prone to induce inadequate actions. E.g. in Case 2 harassment might be encountered by explaining correct behavior to harassing people. 1 IRA could then be used to anafyze possible human errors in that task. One could then conclude that e.g. explanation should be performed by people with intercul-tural background as done in (Luzern 2013). 

In aviation, the taxonomy called Human Factors Analysis and Classification System (HFACS) has been established see Shappell Wiegmann (2000), to understand accidents based on several factors than mere human error. The HFACS taxonomy are based on the Swiss Cheese Model from Reason (1997), looking at accidents as unsafe acts based on preconditions, unsafe conditions and organisational influences, in order to avoid blaming the human element. However, the HFACS method has no exploration of resilience, and should be extended to cover resilient acts, based on preconditions of resilience, resilient conditions and High Reliable Organisational influences. 

In human factor analysis five human factors are considered based on an action based taxonomy as shown in Table 1. A Similar taxonomy is used in SPEAR and reported by Stanton et al. (2005). For each of the factors inherit error frequency/year/ plant was calculated based on the accidental analysis. Table 4 shows the basic human errors that led to an undesirable situations during past accidents. 

Concluding the statistical analysis of past accidents, frequencies and probabilities of human failure were obtained for selected human factor taxonomy. This analysis is strongly based on data obtained from past accidents, therefore it will provide the fewer estimates of HOF errors compare to normal operational situation where near-miss events are also an important aspect to be considered. 

Table 6.8 showed an overview of schemes or taxonomies for the classification of deviations. We will here develop this overview further by presenting taxonomies of human errors, Table 8.1. 

Garrett, J. W., Teizer, J. (2009). Human factors analysis classification system relating to human error awareness taxonomy in construction safety. Journal of Construction Engineering Managemen, 135, 754—763. 

Drury, C. G. (1991), Errors in Aviation Maintenance Taxonomy and Control, in Proceedings of the 35th Annual Meeting of the Human Factors Society (San Francisco), pp. 42-46. 

Isaac, A., Shorrock, S.T., Kennedy, R., Kirwan, B., Anderson, H., and Bove, T. (2002). Technical Review of Human Performance Models and Taxonomies of Error in Air Traffic Management (HERA), Eurocontrol Project Report, HRS/HSP-002-REP-01. Brussels European Organisation for the Safety of Air Navigation. 

---