#function #call #test #dependencies Based on a brief conversation I had, it seems that, recognizing when a function calls for an integration test boils down to identifying how its content is treating its dependencies. If the dependencies are called and one output is made to be input, now we have a flow between the two components and the function we are looking at defines a scenario and an interface where these two elements are made to interact. In this case, we need to test the integration. If the dependencies are just called and their output is not used among them, then, it is a simple case of a components with dependencies to mock. A unit test.