Equality and Human Rights Commission

This covers:

• Factors
• Bias in factor choice
• Factor definition
• Numbers of levels
• Scoring
• Weighting

Factors

• Factors are clearly identifiable aspects of jobs that can be defined and measured and which provide you with the basis for assessing and comparing the relative overall worth of different jobs.
• Examples of factors are responsibility for people', 'knowledge', 'communication skills', 'physical demands', 'emotional demands', 'mental skills' and 'initiative'.
• Except in very broad terms, such as effort, skill, decision-making, there is no standard set of factors applicable to all jobs. The choice of which factors to use is crucial since the final rank order of jobs is most heavily affected by the selection of factors.

Bias in factor choice

• The exclusion of a factor that is important for a particular job will result in it being undervalued relative to other jobs.
• If factors are excluded which occur in predominantly female jobs, this will result in those jobs being placed unfairly at the bottom of a grading structure.
• 'Working conditions' and 'physical strength' are often included in schemes covering manual workers; both these factors will appear in jobs performed by men. On the other hand, factors associated with work done by women, for example, 'manual dexterity, 'caring skills' and 'working with people', may be omitted.
• It is essential to choose factors representative of the whole range of work being evaluated. The example in table 1 shows the impact of a biased set of factors.

Table 1: an example of how biased factors can produce biased outcomes

Factors	Maintenance Engineer	Company Nurse
Skill
Experience in job	10	1
Training	5	7
Responsibility
For money	0	0
Equipment and machinery	8	8
For safety	3	6
For work done by others	3	0
Effort
Lifting equipment	4	2
Strength required	7	2
Sustained physical pressure	5	1
Conditions
Physical environment	6	0
Working position	6	0
Hazard	7	0
Total	64	29

Note: each factor is scored on a scale from 1 to 10. For simplicity no weights have been applied.

The set of factors in table 1 is discriminatory on gender grounds because:

• It contains many characteristics of the predominantly 'male' job of maintenance engineer and very few of the predominantly 'female' job of company nurse.
• Some of the characteristics of the male job are duplicated; for example, 'strength required' to some extent duplicates 'sustained physical effort', with the result that a high score on one would be associated with a high score on the other.
• The same is true of 'lifting requirement' and 'strength required'.
• Note also that the difference in scores on the factor 'experience in job' completely outweighs the more significant difference in the factor 'training'.

Table 2 shows you how choosing a set of factors that incorporates all the important and relevant differentiating characteristics of the jobs to which the scheme is going to be applied, produces a more equitable outcome.

Table 2: an example of how a different choice of factors produces a more equitable outcome

Factors	Maintenance Fitter	Company Nurse
Basic knowledge	6	8
Complexity of task	6	7
Training	5	7
Responsibility for people	3	8
Responsibility for materials and equipment	8	6
Mental effort	5	6
Visual attention	6	6
Physical activity	8	5
Working conditions	6	1
Total	53	54

Note: each factor is scored on a scale from 1 to 10. For simplicity no weights have been added.

You should also check the factor scores of the jobs performed predominantly by female employees. If there are a lot of low scores or if the set of factors makes no provision for scoring aspects of the female jobs, then the set of factors is discriminatory and you should change them.

Factor definition

• Where large numbers of jobs are involved, factors are inevitably generic, so it is essential that you provide definitions of the meaning and scope of each factor. You should look closely at the definitions to ensure that sex or other prohibited forms of bias do not occur.
• For example, a definition of 'working conditions' which included shift work could discriminate against women. A definition of 'experience' which included continuous length of service could discriminate in terms of both gender and age.
• For the purposes of job evaluation, if the job could be learned in two months then the fact that a jobholder has twenty years service is not relevant.
• If 'numeracy' is a factor, a proper definition of what is encompassed in this factor will help to avoid subjective judgments, for example, that women are less numerate than men.

Numbers of levels

Individual factors often have a number of levels within each factor. For example, 'initiative' could have five levels ranging from the lowest, defined as 'following detailed instructions under close supervision', to the highest, defined as 'working within overall policy and having very wide discretion over a broad range of activities with minimal managerial direction'.

Levels should represent clear and recognisable steps in terms of the demands made on the workers, as equal in distance apart as possible.

If each level is equally scored, for example 8 levels = 80 points, 5 levels = 50 points, then the number of levels applied to each job is a form of implicit weighting. In order to avoid sex bias in the number of levels you need to ensure that factors characteristic of 'male' jobs do not have more levels than those factors characteristic of 'female' jobs.

Scoring

• Scoring is the method of attaching values to the various levels within each factor so that a total score for each factor and therefore for the overall job can be calculated.
• The method of scoring for each factor should be reasonably similar, otherwise factors with the same or similar numbers of levels can result in widely differing scores.
• For example, a factor of 'responsibility for financial resources' may have five levels with scores increasing by multiples of five giving a maximum score of 25. However, the factor of 'responsibility for caring' could also have five levels but with scores increasing in multiples of 3, giving a maximum score of 15. The effect of this is that a form of implicit weighting has been applied to the 'responsibility for financial resources factor'.
• You need to ensure that factors characteristic of male dominated jobs do not have a wider dispersion of scores than factors characteristic of female dominated jobs or that you can objectively justify the setting of different levels and scoring.

Weighting

When you have identified the important components of the jobs in your organisation and have converted these into job factors with scales for measuring those factors, you will find that they are not all equally important. It is normal practice to apply weightings to the factor scores so as to reflect the relative importance of the various factors. An important factor may be weighted 10%, whilst an unimportant factor may be only weighted 5%.

Deciding what the weightings should be is a highly subjective process, so it is easy for sex discrimination to enter at this stage. You need to bear in mind that in the event of any challenge, you would have to justify the particular set of weightings by reference to the importance of the factors to the organisation as a whole.

You should not give extreme weights (either very high or very low) to factors that are exclusively found in jobs performed predominantly by one sex. An analysis showing the percentage of total points (after weighting) attributable to each factor will highlight factors with heavy weightings and provide you with a comparison between factors. You can then look again at the factors with the heaviest and the lightest weightings to check that these are not likely to penalise the jobs predominantly held by individuals of one sex.