Transforming the art of feedback with a touch of AI

If there's one thing I find difficult when assessing a candidate's or employee's skills, it's providing feedback that's accurate, true and, most importantly, acceptable. It's a real challenge every time, a bit like navigating in a stormy sea, in the middle of reefs, when it's foggy.

So the adventure began with a quest to make the European Qualifications Framework (EQF) scales of autonomy and responsibility more accessible and applicable to our candidates. The EQF scales, while comprehensive, often seemed to float in a realm of abstraction, completely disconnected from the concrete perceptions each person might have of how skills are put into practice. The challenge was to ground these scales in a tangible reality, making them directly relevant to both candidates and the roles they were seeking. The goal was to move from a kind of technologist jargon to an expression in everyday French or English that everyone could understand.

That's where ChatGPT came in, a tool that wasn't originally designed for HR, but which proved remarkably adaptable. The first breakthrough came when I asked ChatGPT to adapt the EQF scale for specific skills, simply identified by name. This brought us closer to clarity, but it was only the beginning.

As promising as this first result was, I immediately sensed the potential for something bigger and much more grounded in reality. By gradually incorporating additional elements - job descriptions, full skill sets, other contextual information, even personal comments - the feedback document began to take on a new shape. It was transformed from a simple rating to a complete and nuanced reflection of the candidate's true position.

The impact of clarity

The results were immediate. Candidates, previously lost in the blur of generic feedback, now had clearer reference points. They began to take much more ownership of the feedback, asking pertinent questions, clarifying points of agreement and disagreement, and being much more constructive. These documents did more than evaluate; they illuminated a path forward, offering candidates not only a mirror of their current state, but also a map to their potential.

This led to the idea of extending this to other situations where there's a need to give feedback or start a discussion about feelings about skill levels. We began to use it in coaching interviews, skills assessments, mid-career reviews, and so on. In the latter context, one element was crucial. I was faced with a very sensitive situation of profound disagreement over a skills assessment, with a person who was particularly frustrated by failed career progression. This otherwise competent person tended to be in denial about his points of progress, but also not really aware of his real points of support. The situation seemed to have reached an impasse, a dialogue of the deaf gradually taking hold, as it often does in such cases.

I then invited the person to reconstruct together the different levels for a certain number of critical skills, integrating elements of his or her business domain that he or she alone was able to formulate correctly in the ChatGPT interface.

Between the two of us, we were able to create precise, well-formulated scales, clearly referenced to the standardized EQF levels, but anchored in this business reality. This was a critical turning point that not only raised awareness on the part of the person being assessed, but also allowed me to refine my perception of the person's skills.

A few days later, when I was facilitating an Annual Performance Review (APR) training session, we used the same approach with the participants to simulate a disagreement with a peer about a skills assessment. Once again, this provided a really fresh and interesting perspective.

Sharing a new approach to assessment with AI

Inspired by the success of this approach, I felt compelled to share the perspectives and methods with my HR colleagues. The goal was not just to introduce a new tool, but to demonstrate a new way of thinking about AI in HR. It was about turning abstract metrics into personalized, actionable insights.
 

The journey is far from over. The HR and skills assessment landscape is constantly evolving, as are the tools and methods we use. Integrating the EQF scales into ChatGPT's capabilities is just one step in a continuous quest for improvement.

I invite my colleagues, both in HR and across industries, to explore this approach. Think about how AI can improve your assessment and feedback processes. Think about the potential of clearer, more actionable feedback to transform the candidate experience.

What began as a quest to demystify the EQF scales has evolved into a broader exploration of how AI can improve HR practices.

Working with an Appraisee

Here is an example from a real job. These are the first two prompts of a joint session with a person I was assessing. At the beginning of this work, we had our differences and the situation seemed to have reached an impasse...

The formalization of a reference document had a very significant leverage effect.

-->>

User

Task: Extract an 8-point scale and reformulate it in table format.

- Analyze the following page: https://europa.eu/europass/en/description-eight-eqf-levels
- Identify the autonomy and responsibility scale
- Reformulate the 8 levels of this scale into more concrete and simpler terms
- Format these eight levels into a table

ChatGPT

In response to your request, I have consulted the specified page and summarized the information on the eight levels of autonomy and responsibility according to the European Qualifications Framework (EQF). However, I cannot directly reproduce the content of the source. I will provide a simplified summary of the levels of autonomy and responsibility in table format, reflecting the essence of each level without copying the original text. For a detailed description, please go directly to the Europass page describing the eight EQF levels.

This table is a general interpretation of the levels of autonomy and responsibility described in the EQF. To access the specific and complete content, please visit the Europass website mentioned above.

User

Reformulate this scale into 8 levels, in tabular form, for the competence :
- "managing priorities

Characterized by
- Distinguishing critical from non-critical;
- Identifying what will facilitate the achievement of the goal and what will slow down the action;
- Taking a global view of stakeholder priorities;
- Managing priorities in collaboration with others;
- Facilitate collective action;
- Manage your own time flow.

Role:
- Maintenance of mobile machinery (dumpers and hydraulic shovels) in an open pit mine.

ChatGPT
In order to reformulate the "managing priorities" competency scale for the role of maintenance of mobile equipment in an open-cast mine, taking into account the characteristics given, here's a proposal in tabular form. This 8-level scale is adapted to assess the specific ability to separate what is critical from what is not, to identify facilitators and barriers to achieving goals, to have a global vision, to manage priorities cooperatively, to facilitate collective action, and to control one's own time flow.

This scale is a guide for assessing and developing priority management skills in the specific context of mobile equipment maintenance in the mining industry.