Relationships between ability requirements and human errors in job tasks.

Abstract

This article describes a program of research investigating the relations between abilities required to perform job tasks and human error probabilities while performing these tasks. Study 1, using objectively derived human errors made in performing Air Force job tasks, evaluated an ability-requirements classification methodology and found that error rates were related to the numbers, types, and levels of different abilities required. In Study 2, subject matter expert panels used the methodology to identify the ability requirements of tasks in 2 different jobs found in each of 2 types of nuclear power plants. This study, using a task-anchored scale of human error probability, replicated the findings of Study 1, identified 2 ability requirements that cross-validated in predicting task-error probabilities across both jobs and plants, and identified other abilities specific to error probabilities in each job.

Job requirements for control room jobs in nuclear power plants

“…One of the few available studies concerning the relevance of ability requirements for safety in nuclear power plants was conducted by Buffardi et al (2000). This study confirmed that the abilities required for tasks in nuclear power plants are related to human error probabilities, such that error probabilities were found to increase when a larger number of abilities or a higher ability level were required for a task.…”

“…Interrater reliabilities for single abilities in past studies tended to be in the .80s and .90s when 15 or more raters were available (cf. the research reviewed by Buffardi et al, 2000, or by Fleishman andMumford, 1991). In addition, the scales show substantial evidence of construct and predictive validity (Fleishman and Mumford, 1991;Kleinmann et al, 2010).…”

“…The scales include construct definitions, distinctions from similar abilities, definitions of high and low levels of each ability requirement, and task anchors to provide raters with examples of everyday tasks that reflect high, moderate, and low levels of each ability (cf. Figure 1 or Buffardi et al, 2000, for an example of such a scale; or Fleishman et al, 1999, for additional examples). The scale values of these task anchors were determined empirically and were selected due to their high reliability (Fleishman and Mumford, 1991;Fleishman and Railly, 1995).…”

“…An important source of human error is said to be a poor match between employees' abilities and relevant job requirements (Buffardi et al, 2000). Human abilities are important in several different ways.…”

“…Buffardi et al, 2000). Some of these sources concern aspects of the work environment or of the job, such as the plant type (pressurized water reactor vs. boiling water reactor), organizational safety climate, the type of job within a plant, or the operating conditions of a plant, while others concern the characteristics of the employees who perform the jobs in a Job requirements… 5…”