I have always had a hard time with competency models that attempt to define behavioural descriptions at different levels or points along a rating scale. The difficulty is defining logical, distinguishably different levels that enable people to clearly 'see' and understand the difference between the levels.
Does anyone have any views on this or methodologies that they have used that were successful?
Richard Rogers
5 Responses
Different Approaches
Here are some observations on this:
You could set them based on role or function. Not un-typically this would cover:
Strategic Level: Competencies or skills used in functions and roles where the focus is on future goals and targets based on a 2+ year time frame.
Tactical Level: Competencies or skills used in functions and roles where the focus is on converting ideas and proposals into real working plans which must be applied within a 6-12 Month time frame.
Operational; Competencies or skills used in functions and roles where the focus is on immediate actions and outputs.
Please note that this does not imply seniority. Many production managers occupy roles of an operational nature but are senior within the organisation.
An alternative to functional levels is application levels, this consists of the following:
Level 1 = The use of the skill, behaviour or competence on an individual basis.
Level 2 = The use of the skill, behaviour or competence within a group or team setting.
Level 3 = The use of the skill, behaviour or competence in a leader or management position in charge of other people and their application in this context.
Level 4 = The use of the skill, behaviour or competence in an organisational context, not just applying the skills toward team or groups but using them in relation to Divisions or Organisations.
Both can be successful when applied correctly. I hope this answers some of your question.
NVQ/SVQ levels
Have you considered the NVQ/SVQ framework which currently has five levels?
Level 1 – Competence in the performance of a range of varied work activities, most of which may be routine and predictable.
Level 2 – Competence in a significant range of varied work activities in a variety of contexts. Some of the activities are complex or non-routine, and there is some individual responsibility or autonomy. Working with others, perhaps through membership of a work group or team, may often be a requirement.
Level 3 – Competence in a broad range of varied work activities performed in a wide variety of contexts, and most of which are complex and non-routine. There is considerable responsibility and autonomy, and control or guidance of others is often required. This level approximates many supervisory functions.
Level 4 – Competence in a broad range of complex technical or professional work activities performed in a wide variety of contexts and with a substantial degree of performed responsibility and autonomy. Responsibility for the work of others and the allocation of resources is often present. This level approximates many management functions.
Level 5 – Competence which involves the application of a significant range of fundamental principles and complex techniques across a wide and often unpredictable variety of contexts. Very substantial personal autonomy and often significant responsibility for the work of others and for the allocation of substantial resources feature strongly, as do personal accountability for analysis and diagnosis, design planning, execution and evaluation. This level approximates many senior management functions.
A simple set of “levels” defined
Hi Richard,
Like you, I’ve seen “competency models” which are so hard to understand that they don’t serve any useful purpose, e.g. rating people on a scale of 1 – 100 in their ability at each of 200+ skills !! So, if you are looking for a simpler way to rate people’s abilities in a given range of skills, here’s one approach that has worked well in several places …
For each skill, or task, or technology, or activity – whatever – it is possible for people to be rated (or to rate themselves) on a scale from 0 to 4, using this sort of explanation :
0 – I don’t know anything about this.
1 – I can do this task when there is somebody to help / guide / lead me.
2 – I can do this task on my own.
3 – I can help / guide / lead other people in this task.
4 – I am a world-leading expert in this task.
You should find this helps to assess current skill levels, training needs, potential mentors, trainers & training material developers etc.
Assessment Vs Levels
I would agree with Nick’s method for assessing against competencies. But there is a distinction between my input and Nicks; mine seeks to define levels at which competencies exist, which I think is what you were after, whilst Nicks look at assessing against a given competence.
Let the data do the talking
The key is to let the data do the talking. What you are trying to find is the ‘just noticeable difference’ (JND) between one level and another – so code the behavioural data specific to each competency like this – start by identifying the most sophisticated cluster, then the least, then the next highest, then the least – and so on ’till all the data is ‘levelled’. Then the data will be telling you two critical things – how many levels there are (rather than how many you pre-determined there are) and how each level is different from the next. It may be that different competencies have a different number of levels; if this is too inconvenient to work with (most companies seem to think so) then start with a generic scale (5-point recommended) like some of the other answers suggest, and code to those – but then use the data you have IN each level to DEFINE each level; a good compromise between methodological purity and pragmatism. Bingo – you’ve got a meaningful scale. Frankly, I can’t understand how models cam be very useful without scales.