Sulzer-Azaroff B, de Santamaria M C
J Appl Behav Anal. 1980 Summer;13(2):287-95. doi: 10.1901/jaba.1980.13-287.
A "feedback package" system, designed to prevent occupational accidents and to fit directly into the normal operations of an industrial organization, was analyzed. Eighteen hazardous conditions in six production departments were assessed during seven observation sessions over a 12-week period, plus four follow-up observations over 4 months. The "feedback package" was presented in multiple baseline fashion, across subjects (department supervisors). It consisted of presenting the supervisor with copies of observational data, accompanied by a note which congratulated good practices and suggested ways for improving safety conditions, along with occasional comments from a senior executive. The results indicated that during the feedback phase, hazard rates were lower and less variable than during the baseline phase. Baseline data were highly variable with peaks ranging from 20 to 55 hazards per department. Following intervention, hazard frequencies dropped by 60%, averaged across departments, with decreases ranging from 29% to 88%. During treatment, data stabilized, with the highest frequency reaching 33. A modified feedback system was implemented by the organization following termination of the study, validating the assumption that such a system would tend to maintain.
对一个旨在预防职业事故并直接融入工业组织正常运营的“反馈包”系统进行了分析。在12周内的7次观察期间，对六个生产部门的18种危险状况进行了评估，另外在4个月内进行了4次跟进观察。“反馈包”以多基线方式呈现给各对象（部门主管）。它包括向主管提供观察数据副本，并附上一张便条，对良好做法表示祝贺并提出改善安全状况的方法，同时还有高级管理人员偶尔的评论。结果表明，在反馈阶段，危险率比基线阶段更低且变化更小。基线数据变化很大，每个部门的危险峰值从20到55不等。干预后，各部门平均危险频率下降了60%，降幅从29%到88%不等。在处理期间，数据趋于稳定，最高频率达到33。研究结束后，该组织实施了一个改进的反馈系统，验证了这样一个系统倾向于维持的假设。
