A traditional view of automaticity holds that the ability to perform well-practiced skills without attention is adaptive because it frees mental resources to process other information. Without denying the benefits of automaticity, I show how it can also lead to error in domains of expertise such as driving, medical diagnosis, problem solving, and reading. I begin by discussing automaticity within the context of two theoretical frameworks, classical theories of skill acquisition and dual process theory. I then examine situational and individual-difference factors that make errors of automaticity more likely to occur. Next, using proofreading as a testbed, I demonstrate how knowledge, expectations, and other top-down constraints influence reading behaviors and comprehension. I then present two experiments to investigate the self-generation effect in proofreading: the hypothesis that it is more difficult to detect mistakes in one's own writing than in the writing of others. The reasoning behind this hypothesis is that overfamiliarity with self-generated text increases the probability that errors are overlooked or seen but undetected. Finally, I discuss implications of the research and argue that understanding the benefits and consequences of automaticity is critical to improve decision-making outcomes across a wide range of applied contexts.
Read
