Core Concepts: SkillRank
Lee Iverson is the CTO at Ibbaka - view his Skill Profile
Definition of SkillRank
SkillRank on Ibbaka Talent is an ongoing estimate of the likelihood that a person has a particular skill at a particular level of expertise.
We incorporate a variety of different sources of information in order to arrive at this value, but the primary sources are self-assessment and peer assessments.
Since self-assessment can be unreliable, it is weighted less than peer assessments in the calculations and used only for presentation purposes once we have a certain number of peer assessments available.
Underlying the SkillRank values shown in the app is a Bayesian likelihood model that includes a set of prior probabilities and a measure of the potential accuracy (i.e. the influence on the final likelihoods) of both self and peer assessments. When considering peer assessments, the accuracy is based on the expertise level of the assessor on the same skill. e.g. Someone who has a SkillRank of Expert is assumed to be better able to assess that skill in others.
What arises from these likelihood calculations is a measure of the likelihood that a Talent user has this skill with this expertise for any combination of skill and expertise level. With this information, we can present both the derived SkillRank level (the most likely expertise level for that skill) and the confidence that the skill is at that level or greater.
These likelihood values feed into many of the other calculations and insights that Ibbaka Talent can provide. Some examples:
The match score (i.e. likelihood) that a Talent user has a skill at a level required by some role;
The role match score we calculate when asked to determine how well a person might match the full set of skill requirements for a Role or Job (i.e. the accumulation of skill match scores for all of the Must Have, Should Have and Nice to Have skills for that Role or Job); and
The likelihood that a company has the skill coverage for a Role or set of Roles as defined within a competency model.
What are your thoughts on self-assessments and peer reviews? We would love to hear from you.