Concept Formation and Categorization in Dynamic Cognitive Tasks
What are the cognitive processes that determine successful problem solving during goal-directed behavior? How do people determine and selectively employ only the relevant aspects of their knowledge about common objects in the world when they attempt to accomplish an everyday task?

People can entertain multiple identities for the same object (typical or atypical) depending on their current circumstances (Malt & Sloman, 2007). In fact, category membership for artifacts may be defined by the user’s goal at any given time and this membership status changes constantly within the flexible and dynamic nature of human cognition and under the influence of social and cultural factors, as well as specific situational constraints (see Barsalou, 1982, 1983, 1991). The presence of a goal (e.g., inserting a nail into the wall) requires the retrieval of goal-relevant information within semantic memory that will facilitate the achievement of that goal (e.g., using a hammer). Importantly, this activation of semantic knowledge is modulated by the availability of ideal means in the environment toward goal achievement (e.g., inserting a nail into the wall when a hammer is unavailable).

Screen Shot 2012-09-04 at 9.02.58 PM
In past work I examined a novel and intriguing hypothesis, namely that one’s tendency to organize knowledge about the world in concepts and categories (e.g., birds, animals, furniture) in a particular way can predict problem-solving performance. Although concept formation and problem solving are areas not typically examined in conjunction with each other, I have presented a new theoretical model of goal oriented action that is anchored in categorization processes (Chrysikou, 2008). In particular, I asserted that success in problem solving depends on the solvers’ ability to construct goal-derived categories, namely categories that are formed ad hoc to serve specific goals (e.g., things-that-float). I further provided for the first time experimental evidence suggesting that goal-derived categorization is implicated considerably in problem solving: In a series of behavioral experiments I have shown that training individuals to categorize common objects in alternative ways (e.g., using a shoe as a hammer) significantly increases participants’ solution rates in unrelated insight problem solving tasks which require participants to think broadly about a problem by re-categorizing its elements (Chrysikou, 2006; see Figure 1).

What are the cognitive and neural mechanisms underlying these effects? Importantly, what is their duration and generalizability to other tasks? Ongoing research in my lab suggests that the effects of this kind of categorization training may be associated with changes in participants’ attentional states (or mindsets) and can generalize to other domains by affecting performance on a broad range of cognitive tasks (Chrysikou et al., in preparation). Within this line of work, the questions I aim to pursue in the future focus on the characterization of flexible cognitive control in dynamic everyday tasks. An additional line of inquiry refers to the cognitive processes that underlie the activation of specific properties of an object and determine its function: What is the influence of typical or atypical contexts in determining the function of an object? How does the presence of a goal affect the relative dominance of certain aspects of the object’s representation? Finally, I am interested in identifying factors or strategies that either enhance or impede one’s efforts to reach an intended goal. For example, in my earlier research I have examined the conditions under which the presence of examples leads participants to ‘fixate’ on an incorrect solution during problem solving and explored effective ways to counteract these effects (Chrysikou & Weisberg, 2005). Current work in my lab explores whether the modality of the stimulus input (verbal or pictorial) influences the speed and efficiency in determining aspects of an object that are relevant for a specific task. Among the future research aims of the lab is to explore the educational implications and possible applications of these findings for children and young adults.