Country
You are in the Nigeria edition. Type your country to switch.
Nigeria
Loading every country…
No country matches. Check the spelling, or choose "Any country" for the Global edition.
Your learning profile
Human resources and AI · Lesson 2 of 12
AI in sourcing and screening
Sourcing and screening is where AI first arrived in HR, because a single advert can draw hundreds of applications. This lesson explains what matching engines, CV parsers, rankers, applicant assistants and skills inference actually do, how proxy discrimination, feedback loops and gaming creep in, and how to check a screening tool with selection rates by group before it is switched on.
This lesson is complete in writing. A video for it is on its way.
What you will learn
- By the end you can describe how candidate matching, CV parsing, ranking, applicant assistants and skills inference work.
- You can explain proxy discrimination, feedback loops and gaming, and say how each enters a screening tool.
- You can compute selection rates by group and the impact ratio between them from a hiring funnel.
- You can list what to check before a screening tool is switched on.
Key terms in this lesson
- Skills inference
- Using AI to estimate the skills a person has from their CV, job history, learning records or work outputs, often to recommend roles, courses or internal moves.
- Proxy discrimination
- Unfair treatment that arises when a model uses data such as postcode, school or career gaps that stand in for a protected characteristic it was never given.
- Proxy variable
- An input that stands in for another characteristic because the two are closely linked, for example a postcode that closely tracks income or ethnicity.
- Feedback loop
- A cycle in which a model's own past decisions become its next training data, so whatever pattern it started with grows stronger with each retraining.
- Selection rate
- The share of a group that passes a stage of a process: the number selected divided by the number who applied or were considered at that stage.
- Impact ratio
- A group's selection rate divided by the selection rate of the most selected group; a ratio well below 1 points to possible adverse impact.
- Knock-out question
- A screening question, often asked by software in an application, where one answer removes the applicant from the process automatically, so it acts as a selection decision.
- Step 1: What the tools do at the top of the funnel, not started
- Step 2: Three ways screening tools go wrong, not started
- Step 3: A worked funnel: selection rates and the impact ratio, not started
- Step 4: What to check before the tool goes live, not started
- Step 5: Applicant assistants and the candidate's experience, not started
About 10 min of reading
Step 1 of 5
What the tools do at the top of the funnel
Opening this step…
Step 2 of 5
Three ways screening tools go wrong
This step opens when you finish the one above.
Step 3 of 5
A worked funnel: selection rates and the impact ratio
This step opens when you finish the one above.
Step 4 of 5
What to check before the tool goes live
This step opens when you finish the one above.
Step 5 of 5
Applicant assistants and the candidate's experience
This step opens when you finish the one above.
Key takeaways
The key takeaways open when you finish every step.
Lesson quiz
6 questions. You need 85% to pass.
The quiz opens when you finish every step.
Read first
Stockertim Learn is education. It explains how investing works and never recommends buying or selling anything.
Popular this week: Basics: BeginnerNGX & Africa: Beginner