The answer to the question as to the principal causes of error and rework in the construction industry is not straightforward. Reason (1995) describes errors as arising in one of three ways:
• failures of intention where the plan is inadequate (failure to plan). Reason (1995) terms these failures of intention ‘mistakes’ and distinguishes between : rule based mistakes which involve the misapplication of a rule, the application of the wrong rule, or not applying the right rule for a given context; and knowledge based mistakes which occur in novel situations and require significant cognitive processing, which is prone to bias.
• failures in execution (slips and lapses), where a plan is adequate but the actions do not go as planned. These can be further sub-categorised as failures of recognition; attention; memory; or selection. Slips and lapses usually occur in familiar surroundings while someone is performing a routine task.
• deliberate violations which can arise from: routine violations, or cutting corners; optimising violations, to address a personal objective (such as going home early) rather than task objectives; and situational violations which ‘offer the only path available to getting the job done, or when the procedures are seen to be inappropriate for the present situation’.
Each of these causes has different psychological origins and requires different counter measures (see Michie et al. 2011 for a discussion on diagnosing and encouraging behaviour change).
These classifications provide a comprehensive basis to analyse the reason an identified error was made. However, in a complex supply chain the root cause of an error may not be immediately apparent. Accordingly, researchers have also analysed error and rework caused by the nature of the deviation: carelessness, ignorance, or recklessness; by their consequences or causes (Reason 1995); by type of deviation; by construction phase origin; or by cause type – error, omission or change. There is little agreement in the literature as to which is the most appropriate approach to recording errors. Indeed, Reason (1995) suggests that there is no one taxonomy which can serve all needs in describing errors, and as such, the search for a unifying taxonomy may ultimately be fruitless.