In principle, you should first look at the statistical significance and not entering in discussing the magnitude or the sign. Nevertheless, there are recent criticisms on this approach (take p-values as a totem) and some journals even suggest publishing just the s.e. or i.c. The significance or insignificance can be driven by the quality of the data (e.g., big datasets include many observations and almost everything is significant)

Hello

the answer to your question assumes knowledge of the sample of workers with their categories. it is important to know the other independent variables as well because a variable with an insignificant coefficient becomes significant when we add other independent variables. you must also see the structure of your data. is it hierarchical? if so, contextual variables must be taken into account