Torre’s Job-Matching Model | Screening questions

4.7. Screening questions

4.7.1.

Purpose

Screening questions are a way to add requirements on a job post that are not part of the quantification of the normal job post requirements yet. With these questions, you are able to dig deeper into candidate's motivation for this specific job post for example.

4.7.2.

Data

4.7.2.1.

Talent seekers

Talent seekers can provide screening questions on the job post creation or while editing the job post. The questions can be given in different formats, and the talent seeker can specify what to use the screening question for.

Yes/No - Useful to rank or automatically filter applicants
Multiple choice - Useful to rank or automatically filter applicants
Checkboxes - Useful to learn about applicants
Open (written answer) - Useful to filter or learn about applicants
Open (video response) - Useful to have a better grasp of the applicants

For open questions with ranking purposes, the talent seeker must provide an expected good and bad answer. Candidates are able to see this information and use it as a guide when answering the question.

4.7.2.2.

Candidates

Candidates answer the screening questions as part of their application process.

4.7.3.

Algorithm validation

Daily and weekly metrics
Editorial reviews
Direct feedback from talent seekers
Direct feedback from candidates

4.7.4.

Description

We analyze the answers to screening questions through the following methods:

Yes/No - Ideal for ranking or automatically filtering applicants. We assess if the candidate's response aligns with the talent seeker's expected answer. A match scores 1.0; otherwise, the score is 0, leading to automatic filtering.
Yes/No - Ideal for ranking or automatically filtering applicants. We assess if the candidate's response aligns with the talent seeker's expected answer. A match scores 1.0; otherwise, the score is 0, leading to automatic filtering.

For Open (written answer) questions, useful for in-depth applicant evaluation, we analyze responses using LLMs based on these criteria:

Language: Ensuring the response is in the same language as the question.
Grammar: Evaluating the structural integrity and grammatical accuracy of the answer.
Punctuation: Assessing the appropriate use of punctuation marks.
Completeness: Checking if the answer is thorough and provides sufficient detail.
Consistency: Ensuring logical presentation and direct relevance to the question.
Professionalism: Gauging the tone to ensure it remains professional.
Quality: Assessing the overall writing quality and relevance to the question.

4.7.5.

Mathematical description

Let A be the answer given by the candidate.
Let E be the expected answer by the talent seeker.
Let S be the score for the question.

Stotal represents the total score for an open question.
Slang, Sgram, Spunc, Scomp, Sconsis, Sprof, Squal are the scores for language, grammar, punctuation, completeness, consistency, professionalism, and quality, respectively, each evaluated on the candidate's answer A through LLM analysis.