Correlates of work injury frequency and duration among firefighters.

This study examined demographic, personality, and economic incentive correlates of workplace injuries suffered by 171 firefighters over a 12-year period. Results showed that female firefighters experienced more injuries than male firefighters. Several Minnesota Multiphasic Personality Inventory (MMPI) scales (Conversion Hysteria, Psychopathic Deviate, and Social Introversion) were positively related to injury frequency. Regression analyses revealed that age, tenure, gender, marital status, type of injury, and wage variables were significant when predicting the duration of injuries as well as an interaction between marital status and gender. Two MMPI scales (Psychopathic Deviate and Schizophrenia) were also significantly related to injury duration. Indemnity cost estimates were calculated. The results underscore the importance of distinguishing the duration of injury from the occurrence of injury.

Workers' compensation claims from latex glove use: a longitudinal analysis of Minnesota data from 1988 to 1997.

Allergic reaction experienced by health care workers from latex glove use has increasingly become the focus of researchers evaluating occupational sources of injury in health care settings. Many studies have attempted to estimate the prevalence of latex sensitization among health care workers by using various methods, but the findings have been inconsistent and do not predict reactivity. This study used workers' compensation data from Minnesota from 1988 to 1997 to assess allergic reactivity rates, injury severity, and costs associated with latex allergic reactions. The average reactivity rate was 7.1 claims per 100,000 health care workers, and total cost associated with the claims averaged $0.295 per health care employee. Using empirical cost data from another study, it was found that it is not cost-beneficial for health care institutions to globally adopt latex glove-free policies solely on the basis of workers' compensation costs.

Performance evaluation in work settings.

Recent research from 1993 on performance evaluations in work settings is reviewed and integrated with the prior reset and historical bases. Contemporary research reflects several themes: General models of job performance are being developed, the job performance domain is being expanded, research continues to explore the psychometric characteristics of performance ratings, research is developing on potential bias in ratings, rater training is examined, and research continues in terms of efforts to attach utility values to rated performance. We conclude that research is progressing in traditional content areas as well in the exploration of new ground. Researchers are recognizing that job performance is more than just the execution of specific tasks and that it involves a wider array of important organizational activities. There is also an increased optimism regarding the use of supervisory ratings and recognition that such "subjective" appraisal instruments do not automatically translate into rater error or bias.

Development of physical ability tests for police officers: a construct validation approach.

A construct validation approach was followed to affirm that 8 physical ability test events were significantly related to two important constructs underlying the job performance of police officers: strength and endurance. A sample of 115 incumbent police officers took 8 physical ability tests and were rated by supervisors on their physical performances in their job. LISREL methods were used to test the model specified, and a reasonable fit was achieved. Portions of the model were tested on an independent sample of 161 applicants; the fit of the model was again acceptable. A nomological network of relationships, in which strength and endurance factors correlated in expected directions with other physiological and demographic variables, was hypothesized and tested. Finally, the data were examined for potential gender differences and bias. Considerable differences were shown between men and women on both test and performance variables, and women would be overpredicted if a common regression line were used for selection purposes.