Organize a hazard inventory without losing useful detail

Start with a mixed pile of notes

A learner has six notes from an invented training workshop: two refer to the same delivery crossing, one concerns the packing desk, one mentions a visitor route, and two describe different activities in the same room. They are filed under labels such as “yard,” “people” and “important.” The learner cannot tell whether repeated words describe duplicate observations or different situations.

CCOHS treats hazard identification as finding and recording possible hazards. The organizing problem here comes afterward: retaining enough context to retrieve and compare the notes accurately.

Choose three useful organizing keys

For this case, each note receives a task, a location and an affected group. The learner also retains its original identifier. These fields answer different questions. Location finds related work in the same area; task finds similar activity across areas; affected group prevents a visitor concern from disappearing inside a worker-only summary.

Do not force a note into a single category if that destroys meaning. A delivery crossing can concern both receiving activity and visitors. Linking it to two useful views is different from counting it as two independent findings.

Work through the inventory table

N1 and N2 may describe one event, while N3 introduces a different activity and group. Shared location is not sufficient reason to merge all three.

Write a merge rule you can defend

The learner's original rule was “merge anything with the same location.” Replace it with a narrower question: do these notes describe the same event or condition in the same context, or do they merely share a place name?

If the answer is uncertain, retain both identifiers and mark the relationship as unresolved. Deleting one note to make a spreadsheet look tidy can erase a relevant distinction. Equally, counting two photographs of one observation as two findings can exaggerate the inventory. A link between records often preserves more information than a forced merge.

Test whether the inventory answers a real question

Ask the fictional learner to retrieve every note affecting visitors. Then ask for every note related to receiving. The overlapping answers should make sense without changing the underlying count of distinct observations.

Finally, select one record and explain why it is separate from its nearest neighbour. If the answer is only “it was on another page,” the organizing logic needs work. If the answer identifies a different task, group or time-dependent condition, the distinction may be useful. This retrieval test makes the inventory serve analysis instead of simply accumulating rows.