Measurements of causes

Measurements of causes are also divided into two types, namely:

• Measurements of actual causes, and
• Measurements of potential causes.

Measurements of actual causes analyse the causes of actual losses resulting from reported incidents.  These are often used to establish critical targets for safety and loss control work.

The percentage of back injuries due to improper lifting, the number of vehicle collisions due to brake failure, and so forth.

For the most part, cause analyses have historically related to the immediate causes involved with accidents and have been, in effect, analyses of symptoms rather than the root causes.

In the search of predictive and before-the-fact measurement systems, an increasing number of loss control leaders are measuring factors related to substandard practices and conditions (potential loss causes) that can be detected before loss occurs.  Several of the more interesting approaches are:

• Behaviour sampling, and
• Environmental sampling.

Measurements that utilise incident-related causal data are essentially after-the-fact and reactive because of their dependency on events that have already resulted in loss.

Random sampling is an accepted management tool that has a solid background of use for measuring product quality and work effort.  It has also been employed in recent years to measure the degree of safe behaviour of groups of workers.  

Utilising a list of unsafe behaviours, a trained specialist can observe the unsafe behaviours; a trained specialist can observe the workers and accurately estimate the safe and unsafe behaviour of the group being studied.  

Care must be taken to make sure that the principles of random selection are followed and that observations are made before the observer is detected.  Bias is minimised by using a trained specialist, and error-free results become a function of training the observers to see the violations included in the system.  

One of the important “musts” is that management must recognise and accept the tool as a reputable measuring device.

Random sampling can also be applied to physical conditions in a loss control programme in a similar fashion to its application in quality control.

A housekeeping rating could be established through the application of this technique.

Good housekeeping is one of the most sought-after goals in any management efficiency campaign.  

By applying the safe and unsafe determination to a listing of items that would include a further breakdown of conditions related to machinery and equipment, stock and materials, aisles, floors and buildings, etc., a very similar system could be developed.  Care would have to be exercised to maximise objectively.

Quantified risk assessments should be aimed at meeting two important needs:  

1. To determine the relative seriousness of hazards, in order to guide management in assigned priorities to preventive work, and  

2. To measure the justification for the estimated cost of contemplated remedial action.  

To accomplish these needs one should use a formula that calculates the risk of a hazardous situation and provides a numerical evaluation for the urgency of corrective attention to the item.  

The calculated risk scores can then be used to establish corrective action priority.  Another formula should weigh the estimated cost and effectiveness of the contemplated corrective action and give a quantitative estimate of the justification for the cost.  

Figure 2 shows the areas of this causation model addressed by measurements of causes.