Datalytics Datalytics

NOM-035 · Reference guides

Which guide applies to you

The three NOM-035 guides are not interchangeable. Which one you run depends on the size of the workplace, and two of them are run together.

Decide it by headcount

Count the workers at the workplace, not the company. A company with three plants decides plant by plant.

Workers at the site What you run Why
Up to 15 Guide I Only the identification of people who experienced a severe traumatic event. No psychosocial risk questionnaire.
16 to 50 Guide I + Guide II The above, plus identifying and analysing psychosocial risk factors across the workforce.
More than 50 Guide I + Guide III The same, with the extended questionnaire, which adds an evaluation of the organizational environment.

Confirm the thresholds against the current text of the standard before planning a rollout.

How II and III differ

They are the same kind of instrument. III is longer and measures one more thing.

Aspect Guide II Guide III
Who it applies to Workplaces of up to 50 workers Workplaces of more than 50 workers
What it evaluates Psychosocial risk factors Psychosocial risk factors and the organizational environment
Scoring tables Tables 2 and 3 Tables 5 and 6
What it produces A score by domain, by category, and a final score The same, plus the organizational environment breakdown

Why Guide I always applies

Guide I does not evaluate the work. It identifies people. It asks whether someone experienced or witnessed a severe traumatic event — a serious accident, an assault, a kidnapping, a colleague's death — and whether they still have symptoms.

That is why it applies at every size, including where there is no psychosocial risk questionnaire at all. Someone with symptoms needs clinical evaluation, and that obligation does not depend on how many people work there.

It produces no score and no risk level. It produces a list of people to refer, and the record that they were referred.

How it works in practice

  1. 01

    Decide the scope

    Workplace by workplace. Run it across the whole company at once and a 20-person plant's results dissolve into a 400-person one's, and stop telling you anything.

  2. 02

    Survey everyone, or a representative sample

    The standard permits sampling at large sites. A badly built sample is worse than none: over-represent one area and the final score describes that area rather than the workplace.

  3. 03

    Guarantee anonymity, and be seen to

    People answer differently when they think the answer carries their name. If they do not trust the anonymity, the result comes out better than the reality and the diagnosis is worthless.

  4. 04

    Keep the record

    Which questionnaire, to whom, when, and what came out. That is what gets reviewed afterwards.

After you run it

You have numbers by domain, by category, and a final score. The hard part starts there: what they mean, and what each level obliges you to do.

How to read the results

With Datalytics

All three guides are loaded and scored against the official tables. Pick the workplace, send the link, and the report assembles itself as answers come in.

book a demo