When does higher working memory capacity help or

10 downloads 0 Views 1MB Size Report
Sciences, 2301 S. Third Street, Louisville KY 40292. ..... How might greater WM capacity at times benefit, but at other times hinder, insight problem- solving?
Running head: WORKING MEMORY AND INSIGHT

1

DeCaro, M. S. (2018). When does higher working memory capacity help or hinder insight problem solving? In F. Vallee-Tourangeau (Ed.), Insight: On the Origins of New Ideas. New York, NY: Routledge. This paper is not the copy of record and may not exactly replicate the final, authoritative version of the article. © 2018. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/

When Does Higher Working Memory Capacity Help or Hinder Insight Problem Solving?

Marci S. DeCaro University of Louisville

Author Notes Marci S. DeCaro, Department of Psychological and Brain Sciences. Address correspondence to Marci S. DeCaro, Department of Psychological and Brain Sciences, 2301 S. Third Street, Louisville KY 40292. Email: [email protected]

WORKING MEMORY AND INSIGHT Abstract Higher working memory capacity is associated with better performance on a range of tasks, including solving complex problems. However, an increased ability to focus attention on complex problem-solving approaches can also hinder associative thinking processes. Insight problem solving appears to rely on a combination of these working memory demanding and associative processes. As a result, sometimes higher working memory capacity helps, and sometimes it hinders insight. This chapter reviews the contradictory literature on working memory and insight, discussing when and how working memory may positively or negatively impact insight. This relationship likely depends on how individual differences and situational factors impact working memory, and the working memory requirements of the specific insight task. To fully support flexible and unconventional thinking, a better understanding of the cognitive processes underlying insight is needed. Keywords: insight, problem solving, working memory, executive attention

2

WORKING MEMORY AND INSIGHT When Does Higher Working Memory Capacity Help or Hinder Insight Problem Solving? Many great discoveries have been attributed to the process of insight. Thomas Edison suddenly thought to use carbon as a successful filament for the light bulb, after spending over a year fixating on platinum (Ohlsson, 2011). After struggling to find a way to distribute ink evenly in his printing press, Guttenberg gained insight by watching an analogous process in the wine press (Ohlsson, 2011; but see Weisberg, 2015). And Archimedes’s famous “Eureka” moment, realizing that volume could be measured by the displacement of water, came while seeing his own body displace water as he bathed (Siegler, 2000). Insight also occurs in daily experience. For example, students in the classroom demonstrate insight when they suddenly notice and use novel shortcut strategies to correctly solve mathematics problems, saving time and effort, and improving performance (DeCaro, 2016; Siegler, 2000). Insight is a specific form of creativity that is thought to happen when an individual reflects on a problem in new ways, after getting stuck in conventional ways of thinking. As these examples illustrate, insight often arises in a sudden moment of understanding, an “aha” moment (e.g., Ball, Marsh, Litchfield, Cook, & Booth, 2015; Kaplan & Simon, 1990; Weisberg, 2015). Because insight supports flexible and unconventional problem-solving approaches, many are interested in identifying the cognitive processes by which insight occurs, and what factors are most likely to foster these cognitive processes. A number of studies have addressed the role of working memory demanding versus associative processes in insight, which seem to be at odds (cf. Kahneman 2003). The current chapter examines this debate by reviewing research investigating the role of working memory (WM) and executive attention in insight problemsolving. WM enables problem solvers to deliberately focus on a problem and work with information relevant to the task at hand, while inhibiting unnecessary information (e.g., Barrett, Tugade, & Engle, 2004). WM is highly correlated with reasoning and intelligence (Conway, Kane, & Engle, 2003; Engle, Tuholski, Laughlin, & Conway, 1999; Unsworth et al., 2014). WM is also fundamental to a host of important tasks, ranging from scholarly activities such as reading comprehension, writing, and mathematics, to everyday activities such as following directions, planning ahead, and dealing with stress from life events (Conway et al., 2005; Engle, 2002). Although WM requires storing information temporarily, the ability to focus attention while inhibiting distractions (i.e., executive attention) drives the relationship between WM and performance on such tasks (Conway et al., 2003; Engle, 2002). For this reason, WM will be treated as essentially synonymous with executive attention in this chapter. WM varies across individuals—some people have higher trait WM capacity, and some have less. This capacity also varies within an individual, depending on situational factors such as time of day (Ilkowska & Engle, 2010). Measures of WM capacity require individuals to perform two tasks at once, holding information for one task in mind while actively processing information for the other task (Conway et al., 2005; Redick et al., 2012). For example, the commonly-used Operation Span task (Unsworth, Heitz, Shrock, & Engle, 2007) requires individuals to remember a series of 3-7 letters. In between the presentation of each letter, individuals are asked to determine whether a simple arithmetic problem (e.g., (1 × 2) + 1 = 3 ?) is true or false. Thus,

3

WORKING MEMORY AND INSIGHT individuals are asked to keep the letters actively available in a temporary memory store while also attending to the arithmetic problems. The strong association between performance on WM tasks and so many other activities is the basis for its privileged position in cognitive science: WM is considered central to human cognition (Ericsson & Delaney, 1999), and the performance of those higher in WM capacity sets the standard others should arguably follow (De Neys, 2006). Such positive findings are also the reason for a rising number of WM training programs (Klingberg, 2010; Melby-Lervåg & Hulme, 2012; Shipstead, Redick, & Engle, 2012). However, the relationship between WM and insight problem-solving is less clear. Consistent with the majority of research on WM, several studies show that greater WM capacity benefits insight (e.g., Chein &Weisberg, 2014; Chein, Weisberg, Streeter, & Kwok, 2010; De Dreu, Nijstad, Baas, Wolsink, & Roskes, 2012; Gilhooly & Fioratou, 2009; Ricks, Turley-Ames, & Wiley, 2007). However, there is a growing literature showing the opposite effect—that greater WM capacity hinders insight (e.g., DeCaro, Van Stockum, & Wieth, 2016; Van Stockum & DeCaro, 2014a; see also Jarosz, Colflesh, & Wiley, 2012a, 2012b, Reverberi, Toraldo, D’Agostini, Skrap, 2005; Wieth & Zacks, 2011). Some studies also reveal no relationship between WM and insight (e.g., Fleck, 2008; see also Lavric, Forstmeier, & Rippon, 2000). The purpose of this chapter is to outline factors and preliminary research questions that may help reconcile these contradictory findings. To begin to reconcile these findings, this chapter focuses on the role that individual differences in WM play in the strategies people use to solve problems. Greater WM capacity can lead individuals to select more complex, deliberate strategies to solve problems, whereas lower WM capacity often leads to simpler, shortcut, or more automatic strategies. These strategy choices have a different impact on performance, depending on whether a problem type benefits from more complex or associative strategies. Insight problem-solving is also thought to proceed through different phases. Some of these phases may be more reliant on WM, whereas others may benefit from more associative processes that operate largely outside of conscious control. Thus, WM capacity may be positively or negatively associated with problem-solving, depending on where a solver is in the process. Finally, situational factors can impact WM, and potentially change the relationship between WM and insight. For example, WM capacity is lower at one’s non-optimal time of day (Wieth & Zacks, 2011), when under the influence of alcohol (Jarosz, Colflesh, & Wiley, 2012), and immediately after completing attention-demanding (i.e., ego-depleting) tasks (DeCaro & Van Stockum, 2014b). Research indicates that individuals are more successful at solving insight problems in these situations compared to the converse (i.e., at one’s optimal time of day, when sober, or after completed a non-ego-depleting task). In contrast, individuals in other situations may pay closer attention to the steps of problem solving, such as when wearing a white lab coat signifying deliberate, analytical thinking (Van Stockum & DeCaro, 2014a) or when stating the problem steps aloud while working on a problem (e.g., Ball et al., 2015; Schooler, Ohlsson, & Brooks, 2003). This increased attention to the steps can hinder insight problem solving (but see Ball & Stevens, 2009). Hence, problem solving success seems to be contextually dependent.

4

WORKING MEMORY AND INSIGHT This chapter reviews how each of these factors—individual differences in WM capacity, characteristics of the insight task, and the situational context—have separate and interactive effects on insight success. By considering these comprehensive influences on problem-solving, we might better understand how WM demanding and associative processes jointly influence insight. These considerations should help us predict successful insight more accurately, in a broader variety of problem-solving tasks and situations. Finally, by examining the role of WM in the context of insight, we might better elucidate the nature of WM, including its impact on processes supporting cognitive flexibility more generally. Working Memory and Insight Problem-Solving Higher WM capacity benefits performance in a wide range of studies—including those that examine reasoning and problem solving. However, this benefit has been found primarily with more analytic, incremental problems that rely on step-by-step procedures to reach a solution (Simon, 1978; Simon & Reed, 1976; Sternberg, 1982; Thomas, 1974). Incremental problems require executive attention to keep track of the goal and the sub-goals to progress through the problem effectively (Gilhooly & Fioratou, 2009; Hambrick & Engle, 2003; Hills, Todd, & Goldstone, 2010; Raghubar, Barnes, & Hecht, 2010). The relationship between WM and insight problem-solving is less straightforward. In the lab, insight problems typically lead solvers to conceptualize the problem in an incorrect, or unproductive way (i.e., a misrepresentation; Gilhooly & Murphy, 2005). This misrepresentation occurs because the problem solver thinks about the problem in a conventional manner based on prior experience (e.g., Ash & Wiley, 2006; Knoblich, Ohlsson, Haider, & Rhenius, 1999; Knoblich, Ohlsson, & Raney, 2001; Weisberg, 2015; Wiley, 1998) (see Table 1 for examples). To solve these kinds of problems, solvers typically try a conventional solution method, but fail to derive an answer. Instead, individuals must “think outside the box” and relax their conventions. Such flexibility can be challenging. There are two primary theories for how this re-representation process is accomplished, differing in their emphasis on WM demanding versus associative processes. According to the business-asusual view of insight, insight problems are solved in the same way as incremental problems (e.g., Ball & Stevens, 2009; Chein et al., 2010; Chronicle, MacGregor, & Ormerod, 2004; Chronicle, Ormerod, & MacGregor, 2001; Klahr & Simon, 1999; MacGregor, Ormerod, & Chronicle, 2001; Perkins, 1981; Thevenot & Oakhill, 2005, 2006, 2008; Weisberg, 2013). Individuals respond to a failed solution attempt by using attention-demanding search and retrieval processes to adjust their strategy and eventually reach a solution (Ball & Stevens, 2009; Chein & Weisberg, 2014; Davidson, 1995; Kaplan & Simon, 1990; MacGregor, Ormerod, & Chronicle, 2001). Higher WM capacity supports these processes, by better enabling individuals to represent and evaluate the problem, search long-term memory for possible solutions, inhibit irrelevant information, and keep track of previously-retrieved incorrect responses (e.g., Kane & Engle, 2003; Rosen & Engle, 1997; see Chein & Weisberg, 2014; Ricks et al., 2007).

5

WORKING MEMORY AND INSIGHT Table 1. Insight and Incremental Problems used by DeCaro, Van Stockum, & Wieth, (2016, Experiment 2) Insight Problems Socks If you have black socks and brown socks in your drawer, mixed in the ratio of 4:5, how many socks will you have to take out to be sure of having a pair the same color? Solution: 3 socks

Incremental Problems Cards Three cards from an ordinary deck are lying on a table, face down. The following information (for some peculiar reason) is known about those three cards (all the information below refers to the same three cards):  To the left of a queen there is a jack  To the left of a spade there is a diamond  To the right of a heart there is a king  To the right of a king there is a spade Can you assign the proper suit to each picture card? Solution: jack of hearts, king of diamonds, queen of spades

Lilies Water lilies double in area every 24 hours. At the beginning of the summer, there is one water lily on the lake. It takes 60 days for the lake to become completely covered with water lilies. On which day is the lake halfcovered? Solution: The lake is half-covered on the 59th day.

Crime The police were convinced that either A, B, C, or D had committed a crime. Each of the suspects, in turn, made a statement, but only one of the four statements was true.  A said, “I didn’t do it.”  B said, “A is lying.”  C said, “B is lying.”  D said, “B did it.” Who is telling the truth? And Who committed the crime? Solution: B is telling the truth, and A committed the crime

Triangle Show how you can make the triangle below point downward by moving only three of the circles. Solution:

Bachelor Five bachelors, Andy, Bill, Carl, Dave, and Eric, go out together to eat five evening meals (Fish, Pizza, Steak, Tacos, and Thai) on Monday through Friday. It was understood that Eric would miss Friday’s meal due to an out of town wedding. Each bachelor served as the host at a restaurant of his choice on a different night. The following information is known:  Carl hosted the group on Wednesday.  The fellows ate at a Thai restaurant on Friday.  Bill, who detests fish, volunteered to be the first host.  Dave selected a steak house for the night before one of the fellows hosted everyone at a raucous pizza parlor.

6

WORKING MEMORY AND INSIGHT Research in support of the business-as-usual view demonstrates a positive relationship between WM capacity and performance on insight problems. For example, Chein and Weisberg (2014) found a positive correlation between WM capacity and insight problem-solving, even on problems for which participants reported using an insight strategy (see also Chein et al., 2010; De Dreu et al., 2012; Gilhooly & Fioratou, 2009). Ash and Wiley (2006) demonstrated that higher WM capacity benefited insight problem-solving when the problem space was large (i.e., many moves were possible; see Figure 1), likely requiring individuals to maintain different failed solutions in WM. WM had no effect when the problem space was smaller. De Dreu et al. (2012) also found that asking individuals to perform a working memory-demanding dual task (remembering 5 digits on every trial) hurt insight performance, whereas a less demanding dual task (remembering 2 digits on every trial) did not impact insight problem-solving. Together, these findings suggest that WM is helpful, and possibly necessary, for creative insight (De Dreu et al., 2012). Figure 1. Example Problems used by Ash and Wiley (2006) Few Moves Available

Many Moves Available

Move 3 of the gray matchsticks to make 5 squares.

Move 3 matchsticks to make 5 squares.

Move 3 matchsticks to make 5 equilateral triangles.

Move the 3 gray matchsticks to make 5 equilateral triangles.

Note: In both of these problems, the insight is that the outer square/triangle forms a fifth triangle/square. However, findings from other studies run counter to those supporting the business-as-usual view of insight. These findings instead support a special-process view of insight, which states that insight problems generally differ from incremental problems in their underlying solution processes (Ball et al., 2015; Bowden, Jung-Beeman, Fleck, & Kounios, 2005; Chein & Weisberg, 2014; Ohlsson, 2011; Schooler et al., 1993; Seifert, Meyer, Davidson, Patalano, & Yaniv, 1995). According to the special-process view, insight problem-solving typically 7

WORKING MEMORY AND INSIGHT progresses through a series of stages (see Figure 2). First, solvers misrepresent the problem, leading to solution processes that result in an incorrect solution. Solvers then reach an impasse, a point at which no progress can be made until they restructure their initial representation of the problem (Ash & Wiley, 2006; Ohlsson, 1992). Figure 2. Phases of Insight Problem-Solving Representation Phase

Solution Phase

External problem statement is translated into a mental problem representation

Strategic navigation through faulty problem space

Impasse

Restructuring Phase

Success

Failure

Source: DeCaro et al. (2016). Adapted from Ash & Wiley (2006) and Wiley & Jarosz (2012b) Rather than relying on WM-demanding search and inhibition processes, the special-process view suggests that restructuring occurs via associative processes, such as spreading activation in semantic long-term memory. Solvers relax conventional constraints, and consider more peripheral aspects of the problem (Bowden et al., 2005; Knoblich et al., 1999; Ohlsson, 1992; Seifert et al., 1995). These processes are thought to operate largely outside of conscious attentional control (Bowden & Jung-Beeman, 1998; Bowden et al., 2005; Bowers, Regher, Balthazard, & Parker, 1990; Durso, Rea, & Dayton, 1994; Ohlsson, 1992; Schooler et al., 1993; Siegler, 2000). When the correct representation is reached, solvers often experience a sudden “aha!” moment (Ohlsson, 1992, 2011; Schooler et al., 1993; Smith & Kounios, 1996). Solvers have difficulty reporting the steps that led to the solution, presumably because the solution came as a result of unconscious, intuitive processes (Ball et al., 2015; Bowden & Jung-Beeman, 1998; Weisberg, 2015). Thus, according to the special-process view, WM should not benefit insight problem-solving, because insight profits from associative rather than attention-demanding processes. In support of this idea, Fleck (2008) demonstrated that higher WM capacity is positively related to incremental problem-solving, but not insight problem-solving. Similarly, Lavric, Forstmeier, and Rippon (2000) demonstrated that a WM-demanding dual task negatively impacted incremental but not insight problem-solving. Still other studies have demonstrated a negative impact of WM on insight problem-solving. For example, DeCaro, Van Stockum, and Wieth (2016) found that higher WM capacity was associated with less accurate insight problem-solving (see also Beilock & DeCaro, 2007; Van Stockum & DeCaro, 2014a). Other studies have shown that situational factors that reduce WM capacity improve insight performance (e.g., Jarosz et al., 2012; Reverberi et al., 2005; Wieth & Zacks, 2011). 8

WORKING MEMORY AND INSIGHT Working Memory Capacity and Problem-Solving Strategies How might greater WM capacity at times benefit, but at other times hinder, insight problemsolving? One of the ways higher WM capacity improves problem solving is by more capable use of attention-demanding strategies. Problem solving becomes more difficult when (a) the search space, or the number of possible solution paths, becomes wider, (b) the number of steps required to complete the problem increases, and (c) multiple sources of information must be held and manipulated simultaneously in the focus of attention (Ash & Wiley, 2006; Hitch, 1978). Greater executive attention is needed as the complexity of a problem is increased. Individuals with higher WM capacity are better able to devote these resources. Not only are individuals with higher WM capacity better able to devote executive attention to problem solving, but they are also often more likely to select complex strategies in line with this capacity (DeCaro & Beilock, 2010). For example, Beilock and DeCaro (2007, Experiment 1) instructed individuals to solve novel (“modular arithmetic”) mathematics problems. These problems taxed WM because they required multiple steps, including a borrow operation during subtraction, and were to be completed mentally, without the use of paper. After solving a subset of the problems, individuals were asked to describe the strategies they used to solve the problems. Individuals with higher WM capacity were more likely to describe using the correct, multi-step strategy to solve the problems. In contrast, individuals with lower WM capacity were more likely to describe using simpler, shortcut strategies that were less accurate than the correct strategy, but still more accurate than chance. In this case, as is the case across a wide range of common problem-solving tasks, higher WM was associated with using a more complex strategy, which benefitted performance (cf. Wiley & Jarosz, 2012b). This tendency to apply complex strategies does not always benefit performance (DeCaro, Carlson, Thomas, & Beilock, 2009; DeCaro, Thomas, & Beilock, 2008; Gaissmaier, Schooler, & Rieskamp, 2006; Wolford, Newman, Miller, & Wig, 2004). For example, Beilock and DeCaro (2007) conducted a second study using the water jug task, a common measure of insight problem-solving (Luchins, 1942). In this task, individuals are shown three hypothetical water jugs with varying quantities (e.g., Jug A=23, Jug B=96, and Jug C=3; see Figure 3) and asked how one might use these jugs to fill a goal jug to a certain capacity (e.g., 67). Beilock and DeCaro asked participants to solve these problems without the use of paper (increasing WM demand), and to use the simplest strategy possible. The first three problems could be solved using a complex, multi-step solution (i.e., B‒A‒2C; fill Jug B, pour that amount into Jug A, and then pour the remaining amount from Jug B into Jug C twice). The last three problems could also be solved using this complex strategy (e.g., Jug A=34, Jug B=72, Jug C=4; Goal=30). But, importantly, a much simpler strategy could also be used (e.g., A‒C). Insight is thought to occur when one breaks away from the practiced, complex strategy, to notice and use the shortcut. Beilock and DeCaro (2007) examined use of shortcut strategies for individuals who first had achieved mental set, solving the first three problems correctly. On the final problems, higher WM capacity individuals were more likely to use the complex formula, whereas lower WM capacity individuals were more likely to use the more efficient shortcuts. This finding demonstrates that having greater ability to focus attention might lead individuals with higher WM to fixate on a more complex, less adaptive problem representation (Chein & Weisberg, 2014). 9

WORKING MEMORY AND INSIGHT Figure 3. Example Water Jug Problem used by Beilock and DeCaro (2007)

Note: Solvers are asked to find the goal quantity using Jugs A, B, and/or C, using the simplest strategy possible. Here, the solution is B-A-2C. On the insight problems, both this complex strategy and a simpler strategy (e.g., A-C) are possible. The insight is to notice and use the simpler strategy when it is available. Thus, individual differences in WM capacity can both support and inhibit problem solving, depending on the strategies the solver selects. Higher WM enables individuals to better represent and evaluate the problem and search long-term memory for possible solutions (e.g., Rosen & Engle, 1997; see Chein & Weisberg, 2014; Ricks et al., 2007). Higher WM capacity also helps individuals focus on the goal of the task, inhibiting seemingly disparate, irrelevant ideas (Conway, Cowan, & Bunting, 2001; Kane & Engle, 2003). The difficulty may arise when higher-capacity individuals focus intently on the wrong information, such as an incorrect problem representation. Individual differences in WM capacity may therefore influence whether one persists in using WM-demanding strategies even when a simpler strategy is a better tack. This approach could have significant implications for insight problem-solving. However, the exact nature of these effects will be unclear unless researchers also consider the characteristics of the problem-solving task itself. Furthermore, one must consider the ways in which certain kinds of situations encourage higher-capacity individuals to be more flexible, or lower-capacity individuals to be less flexible. Characteristics of the Insight Task Higher WM capacity supports initial problem representation, search, and retrieval of solution possibilities. But higher WM can also lead one to persist in using complex strategies consistent with an incorrect problem representation, hindering re-representation necessary for insight. These strengths, and potential weaknesses, of higher WM capacity may have different impact on insight problem-solving depending on specific characteristics of the insight task. Two overlapping possibilities will be discussed here. First, higher WM capacity may lead to beneficial or detrimental effects during specific phases of insight problem-solving. Second, certain types of insight tasks may emphasize a particular phase (or phases) over others, changing 10

WORKING MEMORY AND INSIGHT the overall balance of WM-demanding and associative processing requirements for the task as a whole. Thus, the effect of WM capacity on a problem overall may differ depending on the characteristics of a particular insight problem. Working Memory and Phases of Insight Problem-Solving Recall that insight is thought to proceed through four phases: representation, solution, impasse, and restructuring (Figure 2). Each of these phases may benefit from, or be hindered by, greater WM resources in different ways. Representation Phase. Higher WM is likely important for the representation phase. Representing a problem requires one to interpret the problem statements and comprehend the task goals and rules (Gick & Lockhart, 1995; Hambrick & Engle, 2003; Mayer & Hegarty, 1996; Novick & Bassock, 2005; Wiley & Jarosz, 2012a). This process requires reading comprehension (Hambrick & Engle, 2003; Kintsch, 1998; Kinstch & Greeno, 1985), distinguishing and selecting relevant from irrelevant problem information (Passolunghi, Cornoldi, & De Liberto, 1999; Wiley & Jarosz, 2012a), and forming an initial mental model of the problem (Ash & Wiley, 2008; Thevenot, 2010). Higher WM supports all of these processes (Kintsch, 1998; Lee, Ng, & Ng, 2009; Thevenot, 2010). Therefore, individuals with higher WM capacity may be better able and/or quicker to form an initial problem representation (DeCaro et al., 2016; Gilhooly & Fioratou, 2009; Jones, 2003; Wiley & Jarosz, 2012b). In two studies, DeCaro, Van Stockum, and Wieth (2016) isolated the role of the initial problem representation to examine the impact of WM capacity on insight problem-solving. In the first study, DeCaro et al. asked participants to solve both insight and incremental matchstick arithmetic problems (Knoblich et al., 1999; see Figure 4). The initial problem representation was thought to be held constant in this problem-solving task: Participants were introduced to the problems at the beginning of the session, and all of the problems had the same instructions. The only difference between problems was the actual Roman numbers and operators used in each matchstick arithmetic problem. Higher WM capacity was associated with better incremental problem-solving, but lower insight problem-solving (see also Van Stockum & DeCaro, 2014a). Converging findings were reported in a second study, in which DeCaro et al. (2016) asked participants to solve insight and incremental word problems (see Table 1). Both problem types likely relied heavily on WM resources to initially represent each problem (i.e., to comprehend the text, select relevant information, and form a mental model; DeCaro et al., 2016). Higher WM capacity was associated with marginally better incremental problem-solving, but no relationship was found between WM and insight problem-solving. However, insight and incremental problem-solving performance was positively correlated, indicating that these tasks have some processes in common (e.g., problem representation). DeCaro et al. statistically controlled for this shared variance between insight and incremental problems, and revealed a negative relationship between WM capacity and insight problem-solving. These findings suggest that problem characteristics common to both incremental and insight problems (e.g., initial problem representation) will likely benefit from higher WM capacity. But when controlling for problem representation, negative effects of higher WM capacity can be seen. Together, DeCaro et al.’s

11

WORKING MEMORY AND INSIGHT (2016) findings suggest that the latter phase(s) of insight may be negatively impacted by higher WM capacity. Figure 4. Example Matchstick Arithmetic Problem used by DeCaro et al. (2016, Experiment 1) Transform this false arithmetic statement into a true arithmetic statement while adhering to the rules provided.

(a) only one matchstick can be moved (b) no matchstick can be discarded (c) upright sticks and slanted sticks are not interchangeable (d) result must be a correct arithmetic statement

Source: Knoblich et al. (1999). Note: In this constraint relaxation problem, the solution is to switch the plus sign into an equals sign.

Solution Phase. It is less clear which of these latter phases of insight—solution or restructuring—might be most negatively impacted by higher WM capacity. Higher WM individuals are better able to adhere to task goals, inhibit distractions, and execute multi-step strategies. Therefore, higher WM may be expected to specifically support the solution phase of problem solving. Support for this idea comes from Ash and Wiley (2006). Participants were given a variety of insight word problems that differed in terms of how much attention they demanded during the solution phase. As shown in Figure 1, the problems were manipulated between conditions so that there were either many possible solution moves, or just a few. Otherwise, the problems were identical between conditions. With this design, it can be assumed that problem representation and restructuring processes were equivalent between conditions—the only difference was the solution phase. Higher WM capacity benefited insight problem-solving when many possible moves were available, but there was no relationship with WM capacity when few moves were available. This finding indicates that higher-capacity individuals were quicker to navigate the solution phase and reach impasse, and move beyond impasse to insight (see Ash & Wiley, 2006). If so, then higher WM capacity may benefit progression through the solution phase of insight problem-solving, facilitating insight primarily by exhausting alternative solutions more efficiently, and beginning the impasse and restructuring process sooner. For some insight problems, a different outcome may be possible as well. Higher WM individuals may be more likely to persist in using complex approaches to such an extent that it slows their progress to impasse, and ultimately restructuring. This persistence may be especially likely if the incorrect problem representation cues individuals to use complex strategies that are inefficient 12

WORKING MEMORY AND INSIGHT (Beilock & DeCaro, 2007; Wolford et al., 2004). For example, in the Lilies problem (Table 1), solvers often initially attempt to use complex mathematical strategies. They may be especially likely to do so, and to persist in doing so, if they have the superior WM resources that support such a methodical approach. In this case, higher WM should slow progression through the solution phase and to impasse (see Wiley, 1998; Wiley & Jarosz, 2012a). However, this idea is untested. More research is needed to determine if, and when, higher WM hinders insight due to factors associated with the solution phase. Restructuring Phase. More research is also needed to clarify the role of WM in the restructuring phase. Some research suggests that WM capacity has no effect on the restructuring phase. For example, in Ash and Wiley’s (2006) study, WM capacity was not associated with insight on problems for which few moves were available. Ash and Wiley interpreted the lack of a positive correlation with WM to mean that restructuring relies on associative processes that operate outside of WM processes (Chein & Weisberg, 2014). However, WM could also have other effects on restructuring. One possibility is that individuals higher in WM capacity use restructuring processes that require a lot of attention, such as search and retrieval strategies (DeCaro et al., 2016; Fleck & Weisberg, 2004; Weisberg, 2006). If restructuring primarily relies on associative processes, as proposed by the special-process view, then such analytic restructuring approaches could hinder insight (see DeCaro et al., 2016). In contrast, if the business-as-usual view is correct, then WM demanding restructuring processes could actually be beneficial. For example, during the solution phase, solvers may encounter new information that prompts another search and restructuring process, leading to success without the experience of impasse at all (Chein & Weisberg, 2014; Fleck & Weisberg, 2004, 2013; Weisberg, 2015). In summary, further studies are needed to examine the impact of WM capacity at different phases of insight problem-solving, to test these different possibilities. Research suggests that higher WM capacity is likely to benefit the representation phase of insight problem-solving, if the representation phase is WM demanding (DeCaro et al., 2016). But, whether higher WM capacity hinders the solution or restructuring phases of insight is unclear. A negative effect would likely depend on (a) whether higher WM individuals are prompted to use a complex strategy, whether intentionally or inadvertently, and (b) whether this strategy conflicts with more optimal problem-solving processes (i.e., use of a simpler shortcut strategy in the solution phase, or use of associative processes in the restructuring phase). Overall Effects of WM on Insight Problem-Solving Therefore, the effect of higher WM on overall insight problem solving performance likely depends on the specific problem. For example, some problems may place heavier requirements on problem representation and, therefore, require greater WM resources (DeCaro et al., 2016). Other problems may instead anchor higher WM individuals on complex solution or restructuring methods, when simpler or more associative approaches will lead to greater success (Beilock & DeCaro, 2007). Some problems may not require impasse or associative processes to restructure at all—these may be solvable via analytic restructuring processes and benefit from higher WM capacity (Chein & Weisberg, 2014; Fleck & Weisberg, 2004, 2013; Weisberg, 2015). Overall 13

WORKING MEMORY AND INSIGHT performance will be determined by some function of all three phases of insight problem-solving considered together. It seems crucial to recognize this point and anticipate the potential impacts different problem configurations may have on the overall observations made in problem-solving research. If a particular problem benefits from higher WM capacity in any or all of these stages, then higher WM capacity should yield overall neutral or better performance. In contrast, problems that benefit from less complex, more associative processes in total may be hindered by greater WM capacity. DeCaro et al.’s (2016) findings, described above, support these ideas. Ash and Wiley’s (2006) findings might also be viewed as tentative support for this idea, when considering how their overall results may have arisen from a combination of effects from each problem-solving phase. First, Ash and Wiley found a neutral effect of WM on insight problems with few solution moves available. However, Ash and Wiley gave participants six different word problems: this added complexity likely required WM for initial problem representation. Thus, it remains possible that higher WM capacity simultaneously (a) benefitted problem representation (cf. DeCaro et al., 2016), (b) had a neutral impact on the solution phase (because few moves were available), and (c) had a negative (or neutral) impact on the restructuring phase. This pattern would likely yield an overall neutral association with WM capacity, when considered in aggregate across all phases of the problem. Second, accuracy on problems with many available moves was positively associated with WM. It is possible that WM (a) benefitted the representation phase, (b) benefited the solution phase, and (c) had a negative (or neutral) effect in the restructuring phase (consistent with the few-moves available condition). If two out of three phases benefited from higher WM capacity as just described, then the combined effect may be overall positive. These examples illustrate how the impact of WM capacity at each phase may lead to different additive or subtractive effects on insight problem-solving accuracy. This proposal is consistent with suggestions that insight may rely on a combination of WMdemanding and associative processes (Bowden et al., 2005; Chuderski et al., 2014; Martindale, 1995; Schooler, 2002; Weisberg, 2015; Wiley & Jarosz, 2012a). Yet, further research is needed to clarify how using WM-demanding processes may alter performance due to the configuration of problem characteristics. Moreover, some research demonstrates that the same problem may be solved by different people in different ways, depending perhaps on prior experience (Ash et al., 2009; Chein & Weisberg, 2014; Fleck & Weisberg, 2004, 2013). Thus, problem characteristics alone may reveal reliable patterns in relation to WM capacity, but other factors, such as individual differences and situational factors, will likely moderate these effects. Situational Factors Although WM capacity varies reliably among individuals, WM capacity also varies within individuals at any given moment due to situational factors. This section describes three categories of situational factors that impact insight in sometimes opposite ways: (a) situations in which WM is reduced, (b) situations in which individuals increase the WM devoted to problem solving, and (c) other general contextual factors that impact insight without necessarily directly targeting WM. All of these situational factors may interact with individual differences and 14

WORKING MEMORY AND INSIGHT problem characteristics to jointly determine the effect of WM capacity on insight problemsolving. Situational Factors that Reduce WM Many studies have shown that situationally decreasing WM improves insight. For example, Wieth and Zacks (2011) demonstrated that individuals solved insight problems more accurately at their non-optimal time of day (e.g., in the evening for a “morning type,” when inhibitory attentional control is reduced), compared to their optimal time of day. Jarosz, Colflesh, and Wiley (2012) found greater insight accuracy for participants given moderate amounts of alcohol compared to those who were sober. DeCaro and Van Stockum (2014b) found superior insight problem-solving when participants were ego-depleted by an attention-demanding task immediately prior. Reverberi et al. (2005) demonstrated that patients with brain damage to the lateral frontal cortex solved insight problems more accurately than normal controls. Ball et al. (2015) reported improved insight when participants engaged in articulatory suppression (i.e., repeating the numbers 1 through 7 over and over) or irrelevant speech (i.e., asked to ignore an irrelevant message, the number 1 through 7, repeated to them) conditions during problem solving (but see Ball & Stevens, 2009). Gasper (2003) showed improved insight after participants were primed by a positive mood state, known to reduce the focus of attention given to task performance. Although WM capacity was not directly examined in these studies, situational factors may interact with WM capacity to impact insight. One might anticipate that lower-capacity individuals would be most disrupted by situations that reduce WM capacity, but this is not what previous research demonstrates. Instead, the effects of reducing WM situationally are most likely to be seen for higher-capacity individuals (e.g., Beilock & DeCaro, 2007; Beilock & Carr, 2005; Gimmig, Huguet, Caverni, & Cury, 2006; Kane & Engle, 2000, 2002). Higher-capacity individuals rely on their superior WM resources to use more attention-demanding strategies. In WM-demanding situations, this extra capacity is taken away, leaving their performance too look like that of lower WM individuals (Beilock & Carr, 2005). As described previously, Beilock and DeCaro (2007) found that higher WM capacity led to less use of an insightful shortcut strategy on the water jug task. In addition to examining performance in a typical testing situation, Beilock and DeCaro also examined performance in a high-pressure condition. Certain high-pressure testing conditions lead to anxious thoughts and worries that compete for WM resources (Beilock, 2008). As a result, under pressure, higher-capacity individuals performed more like lower-capacity individuals, increasing their use of the insight solution. Thus, situational factors that reduce WM may improve performance for higher-capacity individuals who are otherwise likely to “over-think.” The situation decreases their use of attention-demanding methods that are counter-productive for insight problem-solving. Situational Factors that Increase WM Devoted to Problem Solving Although fewer in number, some studies indicate that insight problem-solving can be hindered by situational factors that increase the attention devoted to the task. For instance, Schooler, Ohlsson, and Brooks (1993) reported lower insight accuracy after participants stated aloud the 15

WORKING MEMORY AND INSIGHT steps they were taking to solve the problem or had to explain the steps they used after each problem. However, subsequent studies on this “verbal overshadowing” effect have shown mixed results—some replicating these findings (e.g., Ball et al., 2015), and others finding no effect of concurrent or subsequent verbalization (Ball & Stevens, 2009; Chein & Weisberg, 2014; Chein et al., 2010; Fleck & Weisberg, 2004, 2013; Gilhooly, Fioratou, & Henretty, 2010). Ball et al. (2015) note that the verbal overshadowing effect may be moderated by problem characteristics (i.e., whether the problems primarily require verbal or spatial WM resources) and situational factors (i.e., time on task). Van Stockum and DeCaro (2014a) examined the interaction between WM capacity and another situational factor known to increase attention towards performance—wearing a white lab coat. Previous research (Adam & Galinsky, 2012), demonstrated that the symbolic meaning associated with the clothing one wears can impact the executive attention one devotes to the task (an effect Adam and Galinsky termed “enclothed cognition”). A white lab coat, commonly worn by a doctor or scientist, is typically associated with deliberative or analytical thinking (Adam & Galinsky, 2012). Adam and Galinsky found that wearing a “doctor’s coat” improved performance on the Stroop task, a measure of attention. Van Stockum and DeCaro reasoned that, if wearing a white lab coat increases focused attention, then insight accuracy may suffer. Results supported this hypothesis, but primarily for participants lower in WM capacity. In a no-coat control condition, lower WM capacity was associated with better insight accuracy. However, in the coat condition, the performance of lower-capacity individuals dropped to the level of highercapacity individuals. These findings demonstrate that lower WM capacity individuals can be led to use more analytic approaches that hinder insight, depending on the problem-solving context. These results also support the idea that higher-capacity individuals are already using more deliberative task strategies, and therefore are unaffected when the situation primes this approach. Other Situational Factors There are other situational factors that do not necessarily impact WM, per se, but may still impact whether WM capacity is associated with performance. One is time on task. If higher WM individuals are more likely to use complex search and retrieval strategies to solve insight problems, then it is possible they could attain insight at higher levels if given more time (cf. DeCaro et al., 2009). De Dreu et al. (2012) found that time on task was positively related to insight performance for higher WM individuals, but unrelated to time on task for lower WM individuals. Ball et al. (2015) reported that talking aloud during insight problem solving hindered performance up to the midpoint of the time given (i.e., 3.75 min). But performance increased during the latter half of the task, revealing no differences between the talk-aloud condition and those working silently by the task’s end (7.5 min). Thus, increased focused attention may be particularly detrimental early in the insight task. With sufficient time, higher WM individuals may be able to recover from initial setbacks or improve upon faulty problem-solving approaches. Another possibility is that higher WM capacity individuals perform better on insight tasks if they are given hints, are told that the tasks are measures of insight, or are told that less complex solutions will benefit performance. For example, Chein, Weisberg, Streeter, & Kwok (2010) found a positive relationship between WM capacity and insight accuracy. But participants were provided with a crucial hint, potentially changing the nature of the problem situation. Conway, 16

WORKING MEMORY AND INSIGHT Cowan, and Bunting (2001) and Colflesh and Conway (2007) demonstrated the impact of task instructions on attentional focus for participants with higher and lower WM capacity. Conway et al. asked participants to focus on a message played to them in one ear, while ignoring the message played in the other ear. As instructed, higher-capacity individuals were better at inhibiting the message: Higher-capacity individuals were less likely than lower-capacity individuals to hear their name played in the message they were asked to ignore. Colflesh and Conway instructed participants to listen to both messages, requiring them to divide their attention. This time, higher-capacity individuals were more likely to report hearing their name than lower-capacity individuals. Thus, higher WM individuals could more flexibly spread their attention when instructed to do so. A key issue with insight tasks is the misrepresentation of the problem. If individuals are aware they are likely to incorrectly frame the problem, then they may be quicker to flexibly restructure (see Ash, Cushen, & Wiley, 2009). Higher WM may support the ability to do so. These findings illustrate the importance of how the problem is framed or presented by the experimenter (Kaplan & Simon, 1990). Experience also influences the problem solver’s flexibility. Wiley (1998) demonstrated that baseball experts were anchored by baseball-related words, leading them to fixate on an incorrect problem representation and perform worse on an insight task. Ricks, Turley-Ames, and Wiley (2007) further demonstrated that baseball experts who were higher in WM were the most likely to exhibit negative effects of expertise on insight, likely because they focused their attention more strongly on the incorrect, baseball-related problem representation. In mathematics, prior experience can also lead individuals to rigidly adhere to strategies they have used in the past, overlooking more insightful problem approaches (DeCaro, 2016; McNeil, 2008, 2014; Siegler, 2000). Any other factor that affects WM or flexibility could also potentially impact insight performance. For example, Lin, Hasher, and Zacks (2007) demonstrated that older adults were better at an insight task than younger adults, specifically when distracting hints were embedded in a prior reading task. Older adults were less likely to filter out the distracting (but useful) words. White and Shah (2006) demonstrated that individuals with Attention-Deficit/Hyperactivity Disorder (ADHD) performed worse than non-ADHD individuals on a typical insight task (the Remote Associates Task). Cushen and Wiley (2011) found that early age bilingualism, which is associated with improved ability to switch between task representations (cf. Ricciardelli, 1992), was beneficial for insight. Taken together, these findings demonstrate that a myriad of situational and individual differences factors could impact insight, by affecting how attention and working memory are devoted to the task. Measurement of Working Memory Capacity One other, often overlooked, factor that varies across studies is how WM capacity is measured (Chuderski et al., 2014). First, some WM measures emphasize domain-specific verbal or spatial WM resources (e.g., Baddeley & Hitch, 1974; Hambrick & Engle, 2003; Shah & Miyake, 1996). For example, Chein et al. (2010) found a positive association between spatial WM capacity and performance on the nine-dot problem, a problem that is primarily spatial in nature. The authors argue that it is the domain-specific (i.e., spatial) aspect of the WM task that predicted performance. Problem accuracy was unrelated to verbal WM capacity, and when the domain17

WORKING MEMORY AND INSIGHT general variance between both tasks was accounted for, the relationship with insight problemsolving was no longer found. Second, other measures emphasize WM capacity as primarily driven by domain-general executive attention (Engle, 2002). Several measures of WM capacity align with this conceptualization, namely the complex span tasks (e.g., Operation Span, Reading Span, Symmetry Span; Redick et al., 2012). Performance on these measures is highly correlated, even though they use different secondary tasks (i.e., mathematics problems, semantic judgments, and spatial memory, respectively; Redick et al., 2012). It is this shared variance that typically predicts performance, indicating that the predictive power of these WM tasks stems from domain-general executive attention resources (Kane et al., 2004). Many studies examine the impact of working memory on insight using complex span tasks (e.g., Ash & Wiley, 2006; Beilock & DeCaro, 2007; DeCaro et al., 2016; Chein & Weisberg, 2014; Gilhooly & Murphy, 2005; Ricks et al., 2007). Therefore, these studies primarily emphasize the relationship between insight and executive attention. Third, some studies use more than one measure, complex span or otherwise (e.g., Ash & Wiley, 2006; Beilock & DeCaro, 2007; Chein & Weisberg, 2014; DeCaro et al., 2016; Fleck, 2008; Gilhooly & Murphy, 2005). Sometimes these studies examine both domain-specific and domaingeneral aspects of WM (e.g., Chein & Weisberg, 2014), and sometimes they simply combine both measures into a composite executive attention measure of WM (e.g., Ash & Wiley, 2006; Beilock & DeCaro, 2007; DeCaro et al., 2016). Studies that create composite scores also differ in how they create these composites (e.g., averaging raw or standardized scores). Fourth, studies also differ in whether they include covariates in their analyses (e.g., measures of attention; Chein & Weisberg, 2014), along with WM measures. For example, Gilhooly and Fioratau (2010) measured WM capacity using four measures, two verbal and two spatial. One of these was a complex span measure. They found that WM capacity was positively associated with insight problem solving, but only after controlling for measures of switching and inhibition. Gilhooly and Fioratau argued that this method captured the short-term storage capacity aspect of WM. Similarly, Chuderski et al. (2014) included multiple measures of WM, attention, and reasoning in a structural equation model, to determine the interplay of these variables. Thus, the way WM capacity is conceptualized and measured, and the covariates used in a study, can change the interpretation of WM’s effects on performance. However, it seems unlikely that inconsistency in measuring WM completely explains the contradictory relationship reported between WM and insight. Studies finding both positive and negative effects of WM capacity use a wide variety of measures. In addition, other studies implementing a dual task methodology, or other situational factors that impact WM, show contradictory results similar to individualdifference studies, indicating this pattern of results goes deeper than a simple measurement issue. Thus, although measurement of WM capacity is worth considering and across studies, it seems more likely that disparate effects on insight are driven by individual differences, task characteristics, and situational factors. Nevertheless it would be wise to adopt more standard procedures to measure WM, or routinely report and consider the potential moderating implications of the measures used for one’s research.

18

WORKING MEMORY AND INSIGHT Conclusion Creative insight is critical to the evolution of ideas, both for society at large and in daily experiences. If the factors that enhance or hinder insight can be determined, we may better develop interventions to support this process. This chapter examined the impact of WM on insight, documenting evidence that higher WM capacity can both help or hinder insight problemsolving. These contradictory results highlight the need for additional research that does not simply ask whether WM impacts insight, but why and how these effects occur. The current chapter reviewed three primary boundary conditions for the effect of WM on insight: individual differences in WM capacity, characteristics of the problem-solving task, and situational factors. Whether higher WM will help or hinder insight likely depends on an interaction between these factors. For example, higher WM capacity sometimes leads individuals to utilize complex problem-solving strategies in situations where less demanding or associativebased strategies are more useful. However, in situations where WM capacity is reduced (e.g., in a distracting high-pressure situation; Beilock & DeCaro, 2007), typically higher-capacity individuals may respond more like lower-capacity individuals. Thus, individual differences and situational factors—and their interaction—are both important. As outlined in this chapter, whether higher trait and/or state WM will benefit or hinder insight also depends on the nature of the insight task. Insight tasks tend to fall into a category that is distinct from more incremental, non-insight tasks (Gilhooly & Murphy, 2005). However, insight tasks also vary widely. One way to further characterize these tasks may to be to determine the cognitive processes required or used at each phase of the problem. Specifically, one might ask whether WM supports or hinders progress during the representation, solution, and restructuring phases (Figure 2). At the representation phase, higher WM capacity supports the ability to comprehend the problem statements and select relevant problem information (e.g., Hambrick & Engle, 2003; Wiley & Jarosz, 2012a). But not all problems place heavy demands on representation (DeCaro et al., 2016). And, higher-capacity individuals may also be more likely to attempt solutions using a complex strategy (e.g., Beilock & DeCaro, 2007) or a strategy based on prior experience (Ricks et al., 2007). Higher-capacity individuals are also better able to maintain task goals and execute multi-step strategies (Conway et al. 2001; Kane & Engle, 2003; Wiley & Jarosz, 2012b). Solvers may therefore traverse the solution phase more quickly with higher WM resources available (Ash & Wiley, 2006). But higher WM may also lead individuals to persist too long in using incorrect or suboptimal complex solutions —slowing progress towards impasse and more successful restructuring of the problem. Finally, higher WM may help during restructuring, if this process is best accomplished via methodical search and retrieval processes (cf. Fleck & Weisberg, 2004; Weisberg, 2006). However, these WM-demanding processes may override more optimal associative processes, such as spreading activation in long-term memory, or inhibit access to more weakly activated strategies. This increased focus of attention may reduce the chance of successfully “thinking outside the box” (DeCaro et al., 2016; Ohlsson, 2011). Thus, overall success at an insight problem requires success across each of these phases of insight, and higher WM may help or hurt at each step. Interestingly then, one way to support 19

WORKING MEMORY AND INSIGHT insight may be to determine when WM is most useful for the particular task one is facing. For example, when seeking a new insight, one may need to first attend to the specific details of the problem (i.e., represent the problem) using the full WM resources available. Then, when deriving possible solutions, or after reaching an impasse, one might either try to persist in generating solution possibilities or put oneself in a situation where more associative processes can take over. In other words, one might try to reach an insight through WM-dependent or associative processes, or attempt both. This approach is similar to altering the instructions for an insight task. If one is instructed to look for more flexible problem solutions, this framing may allow one to flexibly alter the mode of attention (cf. Colflesh & Conway, 2007, Ash et al., 2009). By carefully analyzing the cognitive processes important at each stage, one might learn to reach insight more efficiently. And as research accumulates, we will be better able to predict when WM-demanding and associative approaches will best lead to a solution. Additional research is also needed to further study the interaction between the factors outlined in this chapter and other individual differences known to impact cognitive flexibility more generally. For example, bilingualism leads to greater flexibility in switching between task representations in general, and better insight in particular (Cushen & Wiley, 2011). However, the latter finding has only been shown in one study, using an insight task that typically shows neutral to positive effects of WM capacity (i.e., insight word problems; Cushen & Wiley, 2011). Thus, it is unclear whether bilingual individuals will be better at all insight tasks (e.g., on tasks for which the representation phase is less WM-demanding) and in all situations (e.g., when WM is coopted). More generally, an important question for future research is how certain individuals are able to flexibly allocate attention when it is needed, and reduce attention when it is not. Such studies may more generally help inform us about the nature of cognitive flexibility as well as the control of WM. Together, answers to these questions will further inform us about the impact of cognitive control during innovative thinking: Although WM-demanding processes are often beneficial for so many skills, reliance on these resources can inhibit associative processes sometimes necessary for insight.

20

WORKING MEMORY AND INSIGHT References Adam, H. & Galinsky, A. (2012). Enclothed cognition. Journal of Experimental Social Psychology, 48, 1225–1398. doi:10.1016/j.jesp.2012.02.008 Ash, I. K., & Wiley, J. (2006). The nature of restructuring in insight: An individual differences approach. Psychonomic Bulletin & Review, 13, 66–73. doi:10.3758/BF03193814 Ash, I. K., & Wiley, J. (2008). Hindsight bias in insight and mathematical problem solving: Evidence of different reconstruction mechanisms for metacognitive versus situational judgments. Memory & Cognition, 36, 822–837. doi:10.3758/MC.36.4.822 Ash, I. K., Cushen, P. J., & Wiley, J. (2009). Obstacles in investigating the role of restructuring in insightful problem solving. The Journal of Problem Solving, 2, 6-41. Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. Bower (Ed.), Recent advances in learning and motivation: Vol. VIII (pp. 47-90). New York: Academic Press. Ball, L. J., Marsh, J. E., Litchfield, D., Cook, R. L., & Booth, N. (2015). When distraction helps: Evidence that concurrent articulation and irrelevant speech can facilitate insight problem solving. Thinking and Reasoning, 21, 76-96. doi:10.1080/13546783.2014.934399 Ball, L., & Stevens, A. (2009). Evidence for a verbally-based analytic component to insight problem solving. In N. Taatgen & H. J. van Rijn (Eds.), Proceedings of the Thirty-First Annual Conference of the Cognitive Science Society (pp. 1060-1065). Austin, TX: Cognitive Science Society. Barrett, L., Tugade, M. M., & Engle, R. W. (2004). Individual differences in working memory capacity and dual-process theories of the mind. Psychological Bulletin, 130, 553–573. doi:10.1037/0033-2909.130.4.553 Beilock, S. L. (2008). Math performance in stressful situations. Current Directions in Psychological Science, 17, 339-343. doi:10.1111/j.1467-8721.2008.00602.x Beilock, S. L., & Carr, T. H. (2005). When high-powered people fail: Working memory and “choking under pressure” in math. Psychological Science, 16, 101-105. Beilock, S. L., & DeCaro, M. S. (2007). From poor performance to success under stress: Working memory, strategy selection, and mathematical problem solving under pressure. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33, 983–998. doi:10.1037/0278-7393.33.6.983 Bowden, E.M., & Beeman, M.J. (1998). Getting the right idea: Semantic activation in the right hemisphere may help solve insight problems. Psychological Science, 9, 435–440. doi:10.1111/1467-9280.00082 Bowden, E. M., Jung-Beeman, M., Fleck, J., & Kounios, J. (2005). New approaches to demystifying insight. Trends in Cognitive Sciences, 9, 322–328. doi:10.1016/j.tics.2005.05.012 Bowers, K. S., Regehr, G., Balthazard, C., & Parker, K. (1990). Intuition in the context of discovery. Cognitive Psychology, 22, 72–110. doi:10.1016/0010-0285(90)90004-N Chein, J. M., & Weisberg, R. W. (2014). Working memory and insight in verbal problems: Analysis of compound remote associates. Memory & Cognition, 42, 67–83. doi:10.3758/s13421-013-0343-4 Chein, J. M., Weisberg, R. W., Streeter, N. L., & Kwok, S. (2010). Working memory and insight in the nine-dot problem. Memory & Cognition, 38, 883–892. doi:10.3758/MC.38.7.883 Chronicle, E. P., MacGregor, J. N., & Ormerod, T. C. (2004). What makes an insight problem? The roles of heuristics, goal conception, and solution recoding in knowledge-lean 21

WORKING MEMORY AND INSIGHT problems. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30, 14-27. doi:10.1037/0278-7393.30.1.14 Chronicle, E. P., Ormerod, T. C., & MacGregor, J. N. (2001). When insight just won’t come: The failure of visual cues in the nine-dot problem. Quarterly Journal of Experimental Psychology, 54A, 903-919. Chuderski, A. (2014). How well can storage capacity, executive control, and fluid reasoning explain insight problem solving. Intelligence, 46, 258-270. doi:10.1016/j.intell.2014.07.010 Colflesh, G. J. H., & Conway, A. R. A. (2007). Individual differences in working memory capacity and attention in dichotic listening. Psychonomic Bulletin and Review, 14, 699703. doi:10.3758/BF03196824 Conway, A. R. A., Cowan, N., & Bunting, M. F. (2001). The cocktail party phenomenon revisited: The importance of working memory capacity. Psychonomic Bulletin and Review, 8, 331–335. Conway, A. R. A., Kane, M. J., Bunting, M. F., Hambrick, D. Z., Wilhelm, O., & Engle, R. W. (2005). Working memory span tasks: A methodological review and user’s guide. Psychonomic Bulletin & Review, 12, 769–786. Conway, A. R. A., Kane, M. J., & Engle, R. W. (2003). Working memory capacity and its relation to general intelligence. Trends in Cognitive Sciences, 7, 547-552. doi:10.1016/j.tics.2003.10.005 Cushen, P. J., & Wiley, J. (2011). Aha! Voila! Eureka! Bilingualism and insightful problem solving. Learning and Individual Differences, 21, 458-462. doi:10.1016/j.lindif.2011.02.007 De Neys, W. (2006). Dual processing in reasoning: Two systems but one reasoner. Psychological Science, 17, 428-433. DeCaro, M. S. (2016). Inducing mental set constrains procedural flexibility and conceptual understanding in mathematics. Memory & Cognition, 44, 1138-1148. doi:10.3758/s13421-016-0614-y DeCaro, M. S., & Beilock, S. L. (2010). The benefits and perils of attentional control. In M. Csikszentmihalyi & B. Bruya (Eds.), Effortless attention: A new perspective in the cognitive science of attention and action (pp. 51–73). Cambridge, MA: MIT Press. DeCaro, M. S., Carlson, K. D., Thomas, R. D., & Beilock, S. L. (2009). When and how less is more: Reply to Tharp and Pickering. Cognition, 111, 397–403. doi:10.1016/j.cognition.2009.03.001 DeCaro, M. S., Thomas, R. D., & Beilock, S. L. (2008). Individual differences in category learning: Sometimes less working memory capacity is better than more. Cognition, 107, 284–294. doi:10.1016/j.cognition.2007.07.001 Van Stockum, C. A., & DeCaro, M. S. (2014a). Enclothed cognition and controlled attention during insight problem-solving. The Journal of Problem Solving, 7(1), 73-83. doi:10.7771/1932-6246.1164 Van Stockum, C. A., & DeCaro, M. S. (2014b, November). The Upside of Down-Regulated Self-Control: Depletion Enhances Insight. Poster presented at the Annual Meeting of the Psychonomic Society, Long Beach, CA. DeCaro, M. S., Van Stockum, C. A. Jr., & Wieth, M. (2016). When higher working memory capacity hinders insight. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42, 39-49. doi:10.1037/xlm0000152 22

WORKING MEMORY AND INSIGHT De Dreu, C. K. W., Nijstad, B. A., Baas, M., Wolsink, I., & Roskes, M. (2012). Working memory benefits creative insight, musical improvisation, and original ideation through maintained task-focused attention. Personality and Social Psychology Bulletin, 38, 656669. doi:10.1177/0146167211435795 Durso, F.T., Rea, C.B., & Dayton, T. (1994). Graph-theoretic confirmation of restructuring during insight. Psychological Science, 5, 94–98. doi:10.1111/j.14679280.1994.tb00637.x Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11, 19–23. doi:10.1111/1467-8721.00160 Engle, R.W., Tuholski, S.W., Laughlin, J.E., & Conway, A.R.A. (1999). Working memory, short term memory and general fluid intelligence: A latent variable approach. Journal of Experimental Psychology: General, 128, 309–331. Ericsson, K. A. & Delaney, P. R. (1999). Long-term working memory as an alternative to capacity models of working memory in everyday skilled performance. In Miyake, A. & Shah, P. (1999). Models of working memory: Mechanisms of active maintenance and executive control. New York: University Press. Fleck, J. I. (2008). Working memory demands in insight versus analytic problem solving. European Journal of Cognitive Psychology, 20, 139–176. doi:10.1080/09541440601016954 Fleck, J. I., & Weisberg, R. W. (2004). The use of verbal protocols as data: An analysis of insight in the candle problem. Memory & Cognition, 32, 990–1006. doi:10.3758/BF03196876 Fleck, J. I., & Weisberg, R. W. (2013). Insight versus analysis: Evidence for diverse methods in problem solving. Journal of Cognitive Psychology, 4, 436–463. doi:10.1080/20445911.2013.779248 Gaissmaier, W., Schooler, L. J., & Rieskamp, J. (2006). Simple predictions fueled by capacity limitations: When are they successful? Journal of Experimental Psychology: Learning, Memory & Cognition, 32, 966–982. doi:10.1037/0278-7393.32.5.966 Gasper, K. (2003). When necessity is the mother of invention: Mood and problem solving. Journal of Experimental Social Psychology, 39, 248-262. doi:10.1016/S00221031(03)00023-4 Gick, M.L., & Lockhart, R.S. (1995). Cognitive and affective components of insight. In R.J. Sternberg & J.E. Davidson (Eds.), The Nature of insight. (pp. 197–228). Cambridge, MA: MIT Press. Gilhooly, K. J., & Fioratou, E. (2009). Executive functions in insight versus non-insight problem solving: An individual differences approach. Thinking and Reasoning, 15, 355–376. doi:10.1080/13546780903178615 Gilhooly, K. J., Fioratou, E., & Henretty, N. (2010). Verbalization and problem solving: Insight and spatial factors. British Journal of Psychology, 101, 81-93. doi:10.1348/000712609X422656 Gilhooly, K. J., & Murphy, P. (2005). Differentiating insight from non-insight problems. Thinking and Reasoning, 11, 279–302. doi:10.1080/13546780442000187 Gimmig, D., Huguet, P., Caverni, J. P., & Cury, F. (2006). Choking under pressure and working memory capacity: When performance pressure reduces fluid intelligence (Gf). Psychonomic Bulletin & Review, 13, 1005-1010. doi:10.3758/BF03213916

23

WORKING MEMORY AND INSIGHT Hambrick, D. Z., & Engle, R. W. (2003). The role of working memory in problem solving. In J. E. Davidson & R. J. Sternberg (Eds.), The Psychology of Problem Solving (pp. 176– 206). London: Cambridge Press. Hills, T. T., Todd, P. M., & Goldstone, R. G. (2010). The central executive as a search process: Priming exploration and exploitation across domains. Journal of Experimental Psychology: General, 139, 590–609. doi:10.1037/a0020666 Hitch, G. J. (1978). The role of short-term working memory in mental arithmetic. Cognitive Psychology, 10, 302–323. Ilkowska, M., & Engle, R. W. (2010). Trait and state differences in working memory capacity. In A. Gruszka, G. Matthews, & B. Szymura (Eds.), Handbook of Individual Differences in Cognition. Attention, Memory, and Executive Control (pp.295–320). Springer, NY. doi:10.1007/978-1-4419-1210-7_18 Jarosz, A. F., Colflesh, G. J. H., & Wiley, J. (2012). Uncorking the muse: Alcohol intoxication facilitates creative problem solving. Consciousness and Cognition, 21, 487–493. doi:10.1016/j.concog.2012.01.002 Jones, G. (2003). Testing two cognitive theories of insight. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29, 1017–1027. doi:10.1037/0278-7393.29.5.1017 Kane, M. J., & Engle, R. W. (2000). Working-memory capacity, proactive interference, and divided attention: Limits on long-term memory retrieval. Journal of Experimental Psychology: Learning, Memory, & Cognition, 26, 336-358. Kane, M. J., & Engle, R. W. (2002). The role of prefrontal cortex in working-memory capacity, executive attention, and general fluid intelligence: An individual-differences perspective. Psychonomic Bulletin & Review, 9, 637-671. Kane, M. J., & Engle, R. W. (2003). Working-memory capacity and the control of attention: The contributions of goal neglect, response competition, and task set to Stroop interference. Journal of Experimental Psychology: General, 132, 47-70. doi:10.1037/00963445.132.1.47 Kane, M. J., Hambrick, D. Z., Tuholski, S. W., Wilhelm, O., Payne, T. W., & Engle, R. W. (2004). The generality of working memory capacity: A latent variable approach to verbal and visuospatial memory span and reasoning. Journal of Experimental Psychology: General, 133, 189-217. Kaplan, C. A., & Simon, H. A. (1990). In search of insight. Cognitive Psychology, 22, 374–419. doi:10.1016/0010-0285(90)90008-R Kim, S., Hasher, L., & Zacks, R. T. (2007). Aging and a benefit of distractibility. Psychonomic Bulletin and Review, 14, 301-305. Kintsch, W. (1998) Comprehension: A paradigm for cognition. New York: Cambridge University Press. Kintsch, W., & Greeno, J. G. (1985). Understanding and solving word arithmetic problems. Psychological Review, 92, 109–129. doi:10.1037/0033-295X.92.1.109 Klahr, D., & Simon, H. A. (1999). Studies of scientific discovery: Complementary approaches and convergent findings. Psychological Bulletin, 125, 524–543. doi:10.1037/00332909.125.5.524 Klingberg, T. (2010). Training and plasticity of working memory. Trends in Cognitive Sciences, 14, 317–324. doi:10.1016/j.tics.2010.05.002

24

WORKING MEMORY AND INSIGHT Knoblich, G., Ohlsson, S., Haider, H., & Rhenius, D. (1999). Constraint relaxation and chunk decomposition in insight problem solving. Journal of Experimental Psychology: Learning, Memory, and Cognition, 25, 1534–1555. doi:10.1037/0278-7393.25.6.1534 Knoblich, G., Ohlsson, S., & Raney, G. E. (2001). An eye movement study of insight problem solving. Memory & Cognition, 29, 1000–1009. doi:10.3758/BF03195762 Lavric, A., Forstmeier, S., & Rippon, G. (2000). Differences in working memory involvement in analytical and creative tasks: an ERP study. NeuroReport, 11, 1613–1618. doi:10.1097/00001756-200006050-00004 Lee, K., Ng, E. L., & Ng, S. F. (2009). The contributions of working memory and executive functioning to problem solving representation and solution generation in algebraic word problems. Journal of Educational Psychology, 101, 373–387. doi:10.1037/a0013843 Luchins, A. S. (1942). Mechanization in problem solving: The effect of Einstellung. Psychological Monographs, 54, 1–95. doi:10.1037/h0093502 MacGregor, J. N., Ormerod, T. C., & Chronicle, E. P. (2001). Information processing and insight: A process model of performance on the nine-dot and related problems. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 176-201. doi:10.1037/0278-7393.27.1.176 Martindale, C. (1995). Creativity and connectionism. In S. M. Smith, T. B. Ward, & R. A. Finke (Eds.), The creative cognition approach (pp. 249–268). Cambridge, MA: MIT Press. Mayer, R. E., & Hegarty, M. (1996). The process of understanding mathematical problem solving. In R. J. Sternberg & T. Ben-Zeev (Eds.), The nature of mathematical thinking (pp. 29–54). Mahwah, NJ: Lawrence Erlbaum. McNeil, N. M. (2008). Limitations to teaching children 2 + 2 = 4: Typical arithmetic problems can hinder learning of mathematical equivalence. Child Development, 79, 1524-1537. McNeil, N. M. (2014). A change-resistance account of children’s difficulties understanding mathematical equivalence. Child Development Perspectives, 8, 42-47. doi: 10.1111/cdep.12062 Melby-Lervåg, M., & Hulme, C. (2012). Is working memory training effective? A meta-analytic review. Developmental Psychology, 1-22. doi:10.1037/a0028228 Novick, L. R., and M. Bassock, 2005. Problem Solving. In K. J. Holyoak and R. G. Morrison (Eds.). The Cambridge Handbook of Thinking and Reasoning (pp. 321–350). New York: Cambridge University Press. Ohlsson, S. (1992). Information processing explanations of insight and related phenomenon. In M. Keane & K. Gilhooly (Eds.), Advances in the psychology of thinking (pp. 1–44). London: Harvester-Wheatsheaf. Ohlsson, S. (2011). Deep learning: How the mind overrides experience. Cambridge: Cambridge University Press. Passolunghi, M. C., Cornoldi, C., & De Liberto, S. (1999). Working memory and intrusions of irrelevant information in a group of specific poor problem solvers. Memory and Cognition, 27, 779–790. doi:10.3758/BF03198531 Perkins, D. (1981). The mind’s best work. Cambridge, MA: Harvard University Press. Raghubar, K. P., Barnes, M. A., & Hecht, S. A. (2010). Working memory and mathematics: A review of developmental, individual difference, and cognitive approaches. Learning and Individual Differences, 20, 110–122. doi:10.1016/j.lindif.2009.10.005 Redick, T. S., Broadway, J. M., Meier, M. E., Kuriakose, P. S., Unsworth, N., Kane, M. J., & Engle, R. W. (2012). Measuring working memory capacity with automated complex span 25

WORKING MEMORY AND INSIGHT tasks. European Journal of Psychological Assessment, 28, 164–171. doi:10.1027/10155759/a000123 Reverberi, C., Toraldo, A., D’Agostini S., & Skrap, M. (2005). Better without (lateral) frontal cortex? Insight problems solved by frontal patients. Brain, 128, 2882–2890. doi:10.1093/brain/awh577 Ricciardelli, L. A. (1992). Creativity and bilingualism. Journal of Creative Behavior, 26, 242−254. Ricks, T. R., Turley-Ames, K. J., & Wiley, J. (2007). Effects of working memory capacity on mental set due to domain knowledge. Memory & Cognition, 35, 1456–1462. doi:10.3758/BF03193615 Rosen, V. M., & Engle, R.W. (1997). The role of working memory capacity in retrieval. Journal of Experimental Psychology: General, 126, 211–227. Schooler, J. W. (2002). Verbalization produces a transfer inappropriate processing shift. Applied Cognitive Psychology, 16, 989-997. doi:10.1002/acp.930 Schooler, J. W., Ohlsson, S., & Brooks, K. (1993). Thoughts beyond words: When language overshadows insight. Journal of Experimental Psychology: General, 122, 166–183. doi: 10.1037/0096-3445.122.2.166 Shipstead, Z., Redick, T. S., & Engle, R. W. (2012). Is working memory training effective? Psychological Bulletin, 1-27. doi:10.1037/a0027473 Seifert, C. M., Meyer, D. E., Davidson, N., Patalano, A. L., & Yaniv, I. (1995). Demystification of cognitive insight: Opportunistic assimilation and the prepared-mind perspective. In R. J. Sternberg & J. E. Davidson (Eds.), The nature of insight (pp. 65–124). Cambridge, MA: MIT Press. Shah, P., & Miyake, A. (1996). The separability of working memory resources for spatial thinking and language processing: An individual differences approach. Journal of Experimental Psychology: General, 125, 4-27. Siegler, R. S. (2000). Unconscious insights. Current Directions in Psychological Science, 79-83. Simon, H. A. (1978). Information-Processing Theory of Human Problem Solving. In W.K. Estes (Ed.). Handbook of learning and cognitive process. Hillsdale, NJ: Lawrence Erlbaum. Simon, H. A., & Reed, S. K. (1976). Modeling strategy shifts in a problem-solving task. Cognitive Psychology, 8, 86–97. doi:10.1016/0010-0285(76)90005-0 Sternberg, R. J. (1982). Reasoning, problem-solving, and intelligence. In R. J. Sternberg (Ed.), Handbook of human intelligence (pp. 225–307). New York: Cambridge University Press. Thevenot, C. (2010). Arithmetic word problem solving: Evidence for the construction of a mental model. Acta psychologica, 133, 90–95.doi: 10.1016/j.actpsy.2009.10.004 Thevenot, C., & Oakhill, J. (2005). The strategic use of alternative representation in arithmetic word problem solving. Quarterly Journal of Experimental Psychology - A, 58, 13111323. doi:10.1080/02724980443000593 Thevenot, C., & Oakhill, J. (2006). Representations and strategies for solving dynamic and static arithmetic word problems: The role of working memory capacities. European Journal of Cognitive Psychology, 18, 756-775. doi:10.1080/09541440500412270 Thevenot, C., & Oakhill, J. (2008). A generalization of the representational change theory from insight to non-insight problems: The case of arithmetic word problems. Acta Psychologica, 129, 315-324. doi:10.1016/j.actpsy.2008.08.008 Thomas, J. C. (1974). An analysis of behavior in the Hobbits-Orcs problem. Cognitive Psychology, 6, 257-269. doi:10.1016/0010-0285(74)90013-9 26

WORKING MEMORY AND INSIGHT Unsworth, N., Fukuda, K., Awh, E., & Vogel, E. K. (2014). Working memory and fluid intelligence: Capacity, attention control, and secondary memory retrieval. Cognitive Psychology, 71, 1-26. doi:10.1016/j.cogpsych.2014.01.003 Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated version of the operation span task. Behavior Research Methods, 37, 498–505. Van Stockum, C. A., & DeCaro, M. S. (2014a). Enclothed cognition and controlled attention during insight problem-solving. The Journal of Problem Solving, 7(1), 73-83. doi:10.7771/1932-6246.1164 Van Stockum, C. A., & DeCaro, M. S. (2014b). The Upside of Down-Regulated Self-Control: Depletion Enhances Insight. Poster presented at the Annual Meeting of the Psychonomic Society, Long Beach, CA. Weisberg, R. W. (2013). On the “demystification” of insight: A critique of neuroimaging studies of insight. Creativity Research Journal, 25, 1-14. doi:10.1080/10400419.2013.752178 Weisberg, R. W. (2015). Toward an integrated theory of insight in problem solving. Thinking and Reasoning, 21, 5-39. doi:10.1080/13546783.2014.886625 White, H. A., & Shah, P. (2006). Uninhibited imaginations: Creativity in adults with AttentionDeficit/Hyperactivity Disorder. Personality and Individual Differences, 40, 1121-1131. doi:10.1016/j.paid.2005.11.007 White, H. A., & Shah, P. (2011). Creative style and achievement in adults with attentiondeficit/hyperactivity disorder. Personality and Individual Differences, 50, 673-677. doi:10.1016/j.paid.2010.12.015 Wieth, M. B., & Zacks, R. T. (2011). Time of day effects on problem solving: When the nonoptimal is optimal. Thinking & Reasoning, 17, 387–401. doi:10.1080/13546783.2011.625663 Wiley, J. (1998). Expertise as mental set: The effects of domain knowledge in creative problem solving. Memory & Cognition, 26, 716–730. Wiley, J., & Jarosz, A. (2012a). Working memory capacity, attentional focus, and problem solving. Current Directions in Psychological Science, 21, 258–262. doi:10.1177/0963721412447622 Wiley, J., & Jarosz, A. (2012b). How working memory capacity affects problem solving. Psychology of Learning and Motivation, 56, 185–227. doi:10.1016/B978-0-12-3943934.00006-6 Wolford, G., Newman, S. E., Miller, M. B., & Wig, G. S. (2004). Searching for patterns in random sequences. Canadian Journal of Experimental Psychology, 58, 221–228.

27