The different subtasks that make up the task to be analyzed are defined, and then each of the coded postures is assigned to the corresponding subtask. This way, the risk associated with the postures of various actions of the worker can be distinguished (e.g. while clinching on a horizontal and a vertical element, etc.).