Through a series of case studies, we will consider different types of algorithmic bias and debunk several misconceptions. We then consider several concrete steps towards addressing bias.