View in EDS HTML Full Text PDF Full Text

No Role of Initial Problem Representation in Insight Problem Solving

Saved in:
Bibliographic Details
Title: No Role of Initial Problem Representation in Insight Problem Solving
Language: English
Authors: Chuderski, Adam, Jastrzebski, Jan
Source: Creativity Research Journal. 2018 30(4):428-438.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 11
Publication Date: 2018
Document Type: Journal Articles
Reports - Research
Descriptors: Creative Thinking, Problem Solving, Correlation, Fatigue (Biology), Volunteers, Foreign Countries, Accuracy, Intuition, Thinking Skills
Geographic Terms: Poland
DOI: 10.1080/10400419.2018.1531674
ISSN: 1040-0419
Abstract: The literature on insight problems--problems that supposedly can only be solved by rejection of an initial faulty problem representation and sudden comprehension of another, nonobvious representation (restructuring)--suggests that the size of initial representations affects the very process of problem solving. Large initial representations impose systematic, analytical search, whereas only small representations promote intuitive, associative processes assumed by some theorists to underpin insight. In a group of 353 young healthy participants, 6 previously validated insight problems were applied in either a small or large initial representation variant. Results demonstrated no reliable difference in performance between the problem variants with regard to (a) solution accuracy, (b) self-reported insight accompanying solutions, (c) effects of fatigue, (d) correlations with another 6 small representation-size problems, and (e) correlations with working memory capacity (which were notable). This outcome suggests that the size of initial faulty representation plays no role in insight problem solving process, supporting the account assuming its strong similarity to systematic, analytical problem solving.
Abstractor: As Provided
Number of References: 44
Entry Date: 2018
Accession Number: EJ1200015
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwGyeo5JRGw6JsowoCDD3Q1hAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDFjlkUrTGUPhm6ZXXQIBEICBm_ytPdAS-QFf5MUqQgNdsMDF0kMzluAq9Q9njZjz_3YG6hw_BIILs7-O9-VTxSvXXQWqRPEFCpcZyYFbxAdnbK5cUxIEIZtp-3QZG_2zxQ1EGjTWLo0Qc_MC_B3pJ0a544e20IGZafNydCgQGHrkhAzWiJcX3BKGDoUoa_xqscTc_yKHmJ35eZ-jw-O_qNWPW3k_SVmPnOp-Oabu
Text:
  Availability: 1
  Value: <anid>AN0133587424;7lo01oct.18;2018Dec18.14:06;v2.2.500</anid> <title id="AN0133587424-1">No Role of Initial Problem Representation in Insight Problem Solving </title> <p>The literature on insight problems—problems that supposedly can only be solved by rejection of an initial faulty problem representation and sudden comprehension of another, nonobvious representation (restructuring)—suggests that the size of initial representations affects the very process of problem solving. Large initial representations impose systematic, analytical search, whereas only small representations promote intuitive, associative processes assumed by some theorists to underpin insight. In a group of 353 young healthy participants, 6 previously validated insight problems were applied in either a small or large initial representation variant. Results demonstrated no reliable difference in performance between the problem variants with regard to (a) solution accuracy, (b) self-reported insight accompanying solutions, (c) effects of fatigue, (d) correlations with another 6 small representation-size problems, and (e) correlations with working memory capacity (which were notable). This outcome suggests that the size of initial faulty representation plays no role in insight problem solving process, supporting the account assuming its strong similarity to systematic, analytical problem solving.</p> <p>Keywords: Insight; problem representation; working memory; fatigue</p> <p>The literature on creative problem solving has devoted special attention to the so-called insight problems, as it is believed that such problems shed light on the dynamics of mental processes leading to great masterpieces, discoveries, and inventions (Ash, Cushen, & Wiley, 2009; Batchelder & Alexander, 2012; Chu & MacGregor, 2011; Chuderski & Jastrzębski, 2018a; Davidson, 1995; Dow & Mayer, 2004; DeYoung, Flanders, & Peterson, 2008; Duncker, 1945; Koffka, 1935; Maier, 1930; Ohlsson, 1992, 2011; Perkins, 1981; Wertheimer, 1945; Weisberg, 2006, 2013, 2015). In principle, insight problems are presented to solvers in a vague, or even misleading way that suggests a typical (but faulty) initial problem representation. For instance, when asked "how can you throw a ping-pong ball, so it always returns to you with no bouncing?", you might start imagining various horizontal throws that can potentially reverse the ball (e.g., rotating it, throwing it upwind, etc.), as the default meaning of the word "throw" suggests horizontal movement. However, no such throw guarantees a problem solution that always would work. Consequently, at some point of most insight problems, an impasse occurs (Beeftink, van Eerde, & Rutte, 2008). To succeed, one needs to come up with a completely novel problem representation; that is, one must restructure, or rerepresent, the problem (Ohlsson, 1992, 2011). In the case of the ball problem, a solver needs to discover the possibility of a vertical throw, so the ball can return as a result of gravitational force. Once the problem is properly restructured, the sudden comprehension of its solution usually occurs (Aha!). Besides verbal riddles such as the ball problem, insight problems can take the form of mathematical, figural, and spatial puzzles (Ansburg & Dominowski, 2000; Dow & Mayer, 2004). Some problems require minimal reflection ("There are seven sisters in the family, and each sister has exactly one brother; how many brothers does this family count?), but others ("How to connect nine equidistant, square-arranged dots with four straight lines drawn without lifting a hand") are rarely solved even if facilitating cues are provided to participants (Weisberg & Alba, 1981).</p> <p>The main goal of research on insight problem solving has been to establish whether the mental processes leading to the sudden, unexpected discovery of solutions resemble the processes driving more systematic, analytical thinking typical for problems that are solvable using either gradual, combinatorial heuristics (Newell & Simon, 1972) or existing knowledge and analogical transfer (Holyoak, 2005). In fact, some theorists have suggested that there is "nothing special" in insight problem solving, which can be carried out effectively by means of mechanistic chains of standard, well-defined (but not necessarily easily identifiable) mental operations such as attention shifts, memory retrieval, analogical mapping, logical inferences, and imagery (Davidson & Sternberg, 1984; MacGregor, Ormerod, & Chronicle, 2001; Perkins, 1981; Weisberg, 2006, 2013, 2015; Weisberg & Alba, 1981).</p> <p>In contrast, proponents of the <emph>special-process</emph> account of insight problem solving assume it involves holistic, associative, and intuitive thinking, driven by special processes that cannot be reduced to the sequence of standard, analytical cognitive operations. For insight to occur when coping with an insight problem, solvers need to disengage from the typical mental processing they normally would use (Ash et al., 2009) and develop a completely new way of dealing with the problem (Ohlsson, 1992, 2011). For example, initial unsuccessful attempts might lead to two major changes to the problem interpretation: (a) relaxation of unnecessary constraints that block access to the correct solution, and (b) decomposition of problem chunks into simpler pieces of information that can be recomposed adequately (Knöblich, Ohlsson, Haider, & Rhenius, 1999). According to this approach, the most effective insights result from holistic relaxations and decompositions. Another process posited to underpin insight is the unconstrained spread of activation in memory, leading spontaneously to remote associations between ideas that would not be activated under the tight focus and control of attention (Kounios & Beeman, 2014; Mednick, 1962).</p> <p>It is also possible that both the nothing-special and the special-process account of insight can be valid at the same time. For instance, there may exist individual differences in how people cope with insight problems, and some participants (especially those with substantial cognitive resources) may be able to find solutions using analytical processes even though such problems do not promote this type of thinking, whereas others might achieve the same result with associative/intuitive processing. Indeed, analyzing verbal protocols, Fleck and Weisberg (2013) found that the majority of their participants rarely encountered impasse/restructuring and instead used their knowledge and general problem-solving heuristics to resolve a problem (see also Danek, Wiley, & Öllinger, 2016; Salvi, Bricolo, Bowden, Kounios, & Beeman, 2016; Webb, Little, & Cropper, 2016). However, even those few people who ended up at an impasse did not rely on spontaneous associations and intuitions, but rather approached the problem description and available data once again, potentially re-encoding them as well as relaxing unnecessary constraints. That is, these people <emph>restarted</emph> their analytical processing, and thus were able to restructure the problem successfully.</p> <p>A recent large-sample study (Chuderski & Jastrzębski, 2018b) tested the extent to which individual differences in insight problem solving can be related to systematic, analytical reasoning, as well as to working memory capacity (WMC)—the ability to encode, maintain, update, bind, and recall the task-relevant information, resulting from the underlying mechanisms of short-term storage and executive control. The study found a virtually isomorphic relationship between analytical reasoning and insight problem performance (84.6% shared variance) and the strong reliance of the latter on WMC (51.8% shared variance). Moreover, the strength of link between problem solving and reasoning/WMC was virtually identical for the problems solved via self-reported insight and the problems reported as solved analytically (without insight). These data suggested minimal (if any) involvement in insight problem solving of special processes beyond reasoning and working memory.</p> <p>However, another possibility for the heterogeneous nature of cognitive mechanisms underpinning insight problem solving may be rooted in highly different requirements of various insight problems. For some problems, despite their originator's intention, relatively simple analytical strategies may work well, so there is no need for special processes (and thus insight rarely occurs). Some other problems may, indeed, require restructuring, but the initial (faulty) representation of such problems may be so large that, to reach an impasse leading to insight, a solver must go through the large sequence of analytical operations before realizing that the adopted representation is wrong. Even if restructuring (via special processes) is required for solving the problem, the completion of complex analytical inferences is required for detection of the faultiness of the initial problem representation, as well as for the resulting restructuring. Thus, special processes can get <emph>dissolved</emph> in analytical processing, what may be responsible for strong correlations of this type of problem with reasoning/working memory (i.e., restructuring is available only for people having sufficient cognitive resources to complete the search of initial problem space). Moreover, as a solution of such problems would emerge gradually (first by analysis, then due to special processes), the subjective experience of suddenness and surprise related to solution may be relatively weak and/or rare. Only the last category of problems, whose initial representations are so small they can immediately be detected as faulty, and which possess unique characteristics making them solvable only via the special processes, might be valid <emph>insight problems</emph>. Uniquely, such problems would elicit the experience of insight and be independent from (or even negatively related to; see Wiley & Jarosz, 2012) analytic reasoning and working memory.</p> <p>If the size of initial representation of an insight problem influences how easily the person copes with a problem, five predictions can be drawn regarding the difference between problem variants that impose either large or small size of initial representation (henceforth, after Ash & Wiley, 2006, such variants are called <emph>many moves available</emph>, MMA, and <emph>few moves available</emph>, FMA, respectively). First, as the FMA variants might not require complex initial processing, a larger problem-solving accuracy can be expected in comparison to the MMA variants. Ormerod, MacGregor, and Chronicle (2002) used the eight-coin problem (see Figure 1 in Ash & Wiley, 2006). In this problem, eight coins are grouped on the 2-D plane so that some coins touch two other coins, and the other coins touch three. The task is to move exactly two coins so that each coin touches exactly three coins. No solution on the 2-D plane works, so a solver needs to restructure the problem to form two 3-D piles. In the MMA variant of the problem, 20 moves of a single coin originally touching two other coins were possible that made this coin touch three coins on the 2-D plane. However, both coins could never be moved in such a way. So, in this variant participants could carry out many combinatorial operations in search of a 2-D solution before realizing their fruitlessness. In contrast, the FMA variant allowed no such move, so an impasse quickly occurred. In two studies, Ormerod et al. found that more people solved the FMA variant than the MMA variant. Unfortunately, Ormerod et al.'s conclusions were based on a relatively small sample, as in each experiment they studied less than 30 people per variant.</p> <p>PHOTO (COLOR): FIGURE 1 Illustration of the stimuli sequence in the complex span and mental counters working memory tasks, as well as an example item of the Pattern Completion Test measuring reasoning ability.</p> <p>A related study (Ash & Wiley, 2006) used another five problems: one problem conceptually similar to the eight-coin problem (the glasses problem, Ashcraft, 1994), two matchstick arithmetic problems (Knöblich et al., 1999), and two Katona (1940) problems. It reported a slightly positive difference in accuracy (11%) observed between the FMA and MMA problem variants, but this difference was not statistically significant. Thus, it remains to be examined whether the small initial representation does or does not help in restructuring. Our study applied the six Ash and Wiley problems (for the problem descriptions, see their Figure 1) to a three-times-larger sample, to test whether the FMA variants yield higher accuracy.</p> <p>Second, some authors (Bowden, Jung-Beeman, Fleck, & Kounios, 2005; Danek et al., 2016; Salvi et al., 2016; Webb et al., 2016) have suggested that the valid marker of the impasse-restructuring-insight sequence during problem solving consists of self-reports rating how sudden, unexpected, obvious, and pleasant was the solution discovered, and how confident a solver was about the solution. If the FMA variants, indeed, yield more direct restructuring, then self-reported insight should be more frequently observed for these problem variants. To test this prediction, ratings of subjective experience accompanying the solution were collected using one-dimensional scale, which combined suddenness, expectancy, obviousness, and confidence.</p> <p>Third, some scholars (Van Stockum & DeCaro, 2014; Wieth & Zacks, 2011; Wiley & Jarosz, 2012) have proposed that performance on <emph>real</emph> insight problems (supposedly the FMA-like ones) can be optimal during nonoptimal conditions in which such problems are dealt with. In principle, the tight focus and control of attention, typical for optimal conditions, may block special processes required for success in insight problems. Disrupting these conditions, for instance, via fatigue, workload, and/or sleepiness, can make room for the special processes. If solutions for the FMA variants really depend on the special mode of processing, one can expect that nonoptimal conditions will increase performance on FMA variants more than performance on the MMA variants (which to some extent depend on controlled attention). To test this prediction, half of our sample was tested just after the 3-hr session of demanding cognitive tasks, whereas the other half attempted insight problems just after entering the laboratory. Also of interest was whether the level of fatigue imposed could impact the self-reported experience of insight.</p> <p>Fourth, if the FMA problem variants really promote associative, intuitive, holistic thinking, instead of systematic, analytical processing needed on the MMA variants, then the Ash and Wiley (2006) set of FMA problem variants should more strongly correlate with other FMA-like problems, in comparison to their MMA variants. To test the specificity of their FMA problems, another six insight problems were identified that likely induce small initial representations, and the correlation between the total score on these new six problems and the score on the six original FMA variants was compared with the respective correlation for the original MMA variants. A significantly stronger FMA-FMA correlation than the FMA-MMA relationship will suggest that, indeed, problems yielding small initial representations consequently involve different cognitive processes than those supposedly imposing large initial representations.</p> <p>Fifth, it was tested whether the FMA variants, as compared to the MMA problem variants, would correlate more weakly with WMC. This prediction was directly tested by Ash and Wiley (2006), who assumed that WMC is related only to early stages of problem solving (initial representation evaluation typical for MMA), while special processes involved in restructuring are unrelated to individual differences in WMC. However, their results were inconclusive. Although they found a numerically stronger WMC correlation for the MMA (<emph>r</emph> = .42) than for the FMA variants (<emph>r</emph> = .17), they did not formally test the significance of this difference, just drawing their conclusions from the linear regression model including WMC and the significant interaction term that multiplied WMC and the dummy-coded group (problem type, either FMA or MMA). However, such a model yielded strong collinearity of predictors that could lead to statistical artifacts. When their difference in WMC correlation between the FMA and MMA variants was retested with the Fisher test, it appeared to be nonsignificant (<emph>p</emph> = .14)—a result that may be related to the small sample examined by Ash and Wiley (61/54 people in the FMA/MMA group). The large sample adopted here allows for a stronger and more conclusive test of our prediction. This study also examined whether the correlation between WMC and insight problem performance would vary between the optimal (no fatigue) and nonoptimal condition (fatigue). If fatigue indeed switches cognition from WMC-dependent, controlled processing to WMC-unrelated, associative thinking, weaker WMC correlations for the fatigue condition should be observed than for the no-fatigue condition. An exploratory analysis would also test whether this effect was somehow qualified by the insight problem variant.</p> <p>This study aimed at the comprehensive examination of potential impact of the initial (faulty) representation on cognitive processing involved in solving insight problems. The two influential theories of insight problem solving predict different roles of initial representation. The special-process theories claim the role is fundamental and qualitative (i.e., <emph>pure</emph> insight can be observed only in problems yielding minimal initial representation), whereas the nothing-special theories assume just a minor, quantitative role (i.e., large initial representations increase the complexity/difficulty of problem solving, but do not change its analytical nature). Thus, establishing the outcome of manipulating the representation size, related to multiple insight problem-solving indices (objective scores, subjective reports, correlations), potentially contributes to our understanding of the insight problem solving process.</p> <hd id="AN0133587424-2">METHOD</hd> <p></p> <hd id="AN0133587424-3">Participants</hd> <p>The sample size was defined as approximately three times larger than the sample examined by Ormerod et al. (2002) and Ash and Wiley (2006). As many as 353 volunteers (229 women) were tested, who to some extent represented the general population. They were recruited in Krakow (Poland) via advertisements on popular networking websites. Participants were paid the equivalent of 15 euros in Polish zloty. The mean age was 23.4 years (<emph>SD</emph> = 4.32, range 18-40). Two additional people did not attempt all insight problems applied and thus were not included in the analysis. All participants had normal or corrected-to-normal vision. All were informed that the study investigates human thinking, that their data would be anonymous, and that they could end their participation at will at any moment.</p> <hd id="AN0133587424-4">Materials</hd> <p></p> <hd id="AN0133587424-5">Insight problems</hd> <p>The exact set of six problems originally applied by Ash and Wiley (2006) was used, except that the problem descriptions were carefully translated into Polish. Each problem had the MMA and the FMA variant. Another six problems that were applied fulfilled the criteria for the FMA problems (see the problem descriptions in Table 1). Each solution that satisfied the description of a problem was counted as the correct solution. Two scores were calculated: the number of correct solutions for the original set of problems, and the number of correct solutions for the six new problems. Each problem also included a box that was to be checked by a participant if the solution for this very problem "came to mind suddenly and unexpectedly." Participants were reminded that they should monitor and report their subjective state when arriving at solutions. For each set of problems, a subjective measure (henceforth called <emph>insight solutions</emph>) reflected the number of problems with the box checked. The insight solutions measure included all solutions (both correct and incorrect), but the incorrect insight solutions were so rare that the total solutions and the correct insight solutions correlated at <emph>r</emph> = .97.</p> <p>Six new few moves available problems used in the study</p> <p> <ephtml> <table border="1" cellpadding="3"><tbody><tr><td>Problem Name</td><td align="center">Problem Description</td><td align="center">Sample Solution</td></tr></tbody><tbody><tr><td>Triangle</td><td>How to make the triangle below point downward by moving only three of the circles.</td><td><inline-graphic href="" /></td></tr><tr><td>Pig pen</td><td>Nine pigs are closed in a pen. How to separate each pig from all the others using only two additional square fences?</td><td><inline-graphic href="" /></td></tr><tr><td>Two people</td><td>How can one person stand behind a second person when at the same time the second person is standing behind the first one?</td><td>Two persons can stand with their backs turned toward each other.</td></tr><tr><td>Four dots</td><td>Connect all the 4 dots with two straight lines, without lines crossing each other, and without taking the pen off the paper.</td><td><inline-graphic href="" /></td></tr><tr><td>Matchsticks</td><td>The matchsticks in the following problems make roman numerals. Notice that both sides of the equation are not equal. You will need to make these into correct arithmetic equations by moving only a single matchstick in each problem. 1) VI + VI = VI 2) I - I = I</td><td>1) VI = VI = VI 2) - I = - I</td></tr></tbody></table> </ephtml> <emph>Note</emph>. Instruction for the matchstick problems included more detailed rules than are presented in the table.</p> <hd id="AN0133587424-6">Storage and processing task</hd> <p>The first task used to assess working memory capacity, specifically its storage and processing function of WM, was the letter complex span. The task required memorizing four, six, or eight (set size) letters, which were drawn from nine possible stimuli and were presented for 1.2 s apiece. After each letter presentation, participants indicated with a mouse button if a simple arithmetic equation (e.g., 2 × 3 - 1 = 5?) was or was not correct (see Figure 1), to prevent the chunking of letters. The participants were instructed to recall as many letters as they could (in proper order), and also to provide correct answers to the arithmetic task. Five trials for each set size (in increasing order) were presented. During the response procedure, in each task as many 3 × 3 matrices as was a particular set size were displayed. Each matrix contained all nine possible letters. The participants were required to select with the mouse those letters that had been presented in a sequence, in the correct order. There was no time limit for responding. The dependent variable for each complex span task was the proportion of correctly selected letters out of the 90 letters presented in the task.</p> <hd id="AN0133587424-7">Updating task</hd> <p>The other WM task, called mental counters (Larson, Merritt, & Williams, 1988), assessed the updating of WM. The task consisted of a sequence of 120 single letters (B, C, or D) presented in random order in the center of the screen (plus 12 training letters). The goal was to count how many times each letter had been presented, up to four (see Figure 1). Each time a letter was presented, participants, by pressing one of the two response keys, indicated whether they had seen the letter for the fourth time (target) or less than the fourth (nontarget). Thus, after each fourth occurrence of the letter, participants should have reset their mental counter of a particular letter to zero. To help in resetting the counters, the participants were presented with feedback after false alarms, as well as upon missed targets. Each letter, approximately 2 × 2 cm in size, was displayed for 2.5 s, followed by a mask displayed for 0.3 s. The final score was the proportion of correct hits in the target trials minus one third of false hit rate in the non-target trials.</p> <hd id="AN0133587424-8">Fluid reasoning test</hd> <p>The Pattern Completion Test, which was used to control for general cognitive ability, included 16 figural patterns (items) in various layouts. For each pattern, one fragment was missing, and the participant had to choose the fragment that correctly completed the pattern (according to a hidden rule) out of four to seven (depending on an item) response options provided in the test (see Figure 1). Twenty minutes were provided for solving this task. The score was the sum of correctly solved items.</p> <hd id="AN0133587424-9">Procedure</hd> <p>Participants were tested in a psychological laboratory in groups of four people on average. Each group received, randomly, either six FMA variants or six MMA variants of the original problems. Participants had 4 min to solve each problem (exactly as in Ash & Wiley, 2006). Problems were presented on the computer screen in a random order. Responses were given using the paper answer sheets. Then, participants attempted six new FMA problems, written on one sheet of paper, and had 20 min to solve them. The first 182 participants (the fatigue condition) were tested just after participation in a large psychometric study that lasted approximately 3 hr and consisted of five demanding cognitive tests that required processing of complex relations, as well as the letter complex span task, the mental counters task, and the Pattern Completion Test. The remaining 171 participants (the no-fatigue condition) attempted the original and the new insight problems first, then completed the letter complex span task, the mental counters task, and the Pattern Completion Test.</p> <hd id="AN0133587424-10">RESULTS</hd> <p></p> <hd id="AN0133587424-11">Descriptive statistics</hd> <p>Regarding the insight problems, the WM task, and the reasoning score, there were no missing data except for one complex span, which was substituted with the mean accuracy for that task. As the WMC measure the mean of <emph>z</emph>-scores from both WM tasks was used. Table 2 provides means and standard deviations for all 10 measures, presented separately for the fatigue and no-fatigue conditions. All scores fitted the widely accepted rule-of-thumb criteria of normality (skew < 2.0, kurtosis < 4.0). Reliabilities for both WM tasks were very good, whereas reliabilities pertaining to insight problems were as low as αs ≈ 6 (see Table 2). That was expected, given that each set included only six problems. When the original and new problems were combined into one set, alpha increased to an acceptable.68.</p> <p>Descriptive statistics for the measures analyzed in the study</p> <p> <ephtml> <table border="1" cellpadding="6"><tbody><tr><td /><td align="center">All</td><td align="center" colspan="2">Fatigue</td><td align="center" colspan="2">No Fatigue</td></tr><tr><td>Variable</td><td align="center">Alpha</td><td align="center">Mean</td><td align="center"><italic>SD</italic></td><td align="center">Mean</td><td align="center"><italic>SD</italic></td></tr></tbody><tbody><tr><td>Original FMA problems accuracy</td><td>0.60</td><td>3.90</td><td>1.45</td><td>3.76</td><td>1.55</td></tr><tr><td>Original MMA problems accuracy</td><td>0.57</td><td>3.56</td><td>1.56</td><td>3.62</td><td>1.47</td></tr><tr><td>New FMA problems accuracy</td><td>0.59</td><td>3.02</td><td>1.76</td><td>3.00</td><td>1.61</td></tr><tr><td>Original FMA insight solutions</td><td>0.65</td><td>1.32</td><td>1.51</td><td>0.70</td><td>1.09</td></tr><tr><td>Original MMA insight solutions</td><td>0.72</td><td>1.02</td><td>1.54</td><td>0.49</td><td>0.89</td></tr><tr><td>New FMA insight solutions</td><td>0.56</td><td>0.88</td><td>1.23</td><td>0.66</td><td>1.00</td></tr><tr><td>Letter complex span</td><td>0.87</td><td>0.66</td><td>0.17</td><td>0.61</td><td>0.19</td></tr><tr><td>Mental counters</td><td>0.91</td><td>0.56</td><td>0.12</td><td>0.53</td><td>0.13</td></tr><tr><td>WMC</td><td align="center">n/a</td><td>0.10</td><td>0.80</td><td>−0.12</td><td>0.89</td></tr><tr><td>Pattern Completion Test</td><td>0.74</td><td>9.18</td><td>3.07</td><td>8.86</td><td>3.23</td></tr></tbody></table> </ephtml> <emph>Note</emph>. FMA = few moves available. MMA = many moves available. WMC = working memory capacity. <emph>N</emph> for FMA/MMA problems was 89/93 in the fatigue condition (total <emph>N</emph> = 182), and 86/85 in the no-fatigue condition (total <emph>N</emph> = 171). WMC is calculated as the mean of z scores from the letter complex span and the mental counters.</p> <p>The level of general cognitive ability, indicated by the Pattern Completion Test, was comparable between the FMA and MMA groups, Δ<emph>M</emph><subs>ability</subs> = .04, <emph>t</emph>(<reflink idref="bib351" id="ref1">351</reflink>) = 1.84, <emph>p</emph> = .066, Cohen's <emph>d</emph> = .19, as well as between the fatigue and no-fatigue groups, Δ<emph>M</emph><subs>ability</subs> = .02, <emph>t</emph>(<reflink idref="bib351" id="ref2">351</reflink>) = 0.92, <emph>p</emph> = .356, Cohen's <emph>d</emph> = .10. Thus, any effects found could not be attributed to potential differences in general cognitive ability between groups.</p> <hd id="AN0133587424-12">Problem solving accuracy</hd> <p>Because in all the herein presented analyses the contrasting predictions of two models (theories) were validated, with one model (the special-process account) predicting the differences between the FMA and MMA variants, and the other model (the nothing-special account) predicting no differences, Bayes Factor (BF) was used for testing differences in means between the groups. Here, BF assessed the ratio of posterior likelihood in favor of the special-process account, given the evidence. Values of BF surpassing 2.0 would suggest marginal evidence for the FMA-MMA difference, values above 5.0 would reflect reliable evidence, and values above 10.0 would indicate strong evidence. Symmetrically, BF values below 0.50/.20/.10 would suggest weak/reliable/strong evidence in favor of the lack of any FMA-MMA difference (i.e., for the nothing-special theory). BF values between 0.50 and 2.0 would be inconclusive and favor no model. All calculations were carried out in JASP software version 0.8.4 for Windows (jasp-stats.org). The default value of prior parameter was adopted, assuming the equal a priori likelihood of both theories tested. Bayesian <emph>t</emph> test of the difference in accuracy on the six original FMA versus six original MMA problems, Δ<emph>M</emph> = 0.24 problems, <emph>d</emph> = 0.16, yielded evidence tending to support the lack of difference in accuracy between problem variants, BF<subs>FMA-MMA</subs> = 0.342. The huge value of BF<subs>ORIGINAL-NEW</subs> = 1.57 × 10<sups>11</sups> indicated that accuracy on the original problems (the FMA and MMA combined) was reliably higher than accuracy on the new problems, Δ<emph>M</emph> = 0.69 problems, <emph>d</emph> = 0.43.</p> <hd id="AN0133587424-13">Insight solutions</hd> <p>Evidence weakly supported the prediction that solving the FMA and MMA variants were equally frequently accompanied by subjective experience of insight, Δ<emph>M</emph> = 0.24 problems, <emph>d</emph> = 0.19, BF<subs>FMA-MMA</subs> = 0.516. The value of BF<subs>ORIGINAL-NEW</subs> = 0.235 indicated no difference in insight solutions between the original problems (the FMA and MMA combined) and the new problems, Δ<emph>M</emph> = 0.22 problems, <emph>d</emph> = 0.09.</p> <hd id="AN0133587424-14">Fatigue</hd> <p>Evidence strongly favored the lack of difference in accuracy on the six original problems between the fatigue versus no-fatigue conditions, Δ<emph>M</emph> = 0.03 problems, <emph>d</emph> = 0.02, BF<subs>FATIGUE</subs> = 0.120. Exactly the same conclusion pertained to the six new problems, Δ<emph>M</emph> = 0.03 problems, <emph>d</emph> = 0.02, BF<subs>FATIGUE</subs> = 0.119. However, there was strong support for the hypothesis that fatigue increased the insight solutions rate for the original problems, Δ<emph>M</emph> = 0.57 problems, <emph>d</emph> = 0.44, BF<subs>FATIGUE</subs> = 801.7. However, no reliable difference was found for the new problems, Δ<emph>M</emph> = 0.22 problems, <emph>d</emph> = 0.19, BF<subs>FATIGUE</subs> = 0.578. Bayesian Analysis of Variance (BANOVA), including the factor of insight problems variant (FMA vs. MMA) and the condition factor (fatigue vs. no-fatigue), strongly suggested the lack of difference in the FMA-MMA accuracy difference between the fatigue and no-fatigue condition, BF<subs>FMA-MMA×FATIGUE</subs> = 0.007. Similarly, no such interaction was found for the insight solutions rate, BF<subs>FMA-MMA×FATIGUE</subs> = 0.003.</p> <hd id="AN0133587424-15">Correlation with the new FMA problems</hd> <p>For assessing differences in Pearson correlations strength, their 95% confidence intervals were directly compared. The correlation between accuracy on the new FMA problems and the original FMA variants accuracy, <emph>r</emph><subs>FMA-FMA</subs> = .46 [.33,.57], was virtually identical to the correlation for the MMA variant accuracy, <emph>r</emph><subs>MMA-FMA</subs> = .47 [.35,.58]. Accounting for the condition (see Figure 2) did not yield any significant difference between the FMA and MMA correlations, as their confidence intervals strongly overlapped both under fatigue, <emph>r</emph><subs>FMA-FMA</subs> = .45 [.27,.60] and <emph>r</emph><subs>MMA-FMA</subs> = .56 [.40,.69], and no-fatigue, <emph>r</emph><subs>FMA-FMA</subs> = .47 [.29,.62] and <emph>r</emph><subs>MMA-FMA</subs> = .36 [.16,.53].</p> <p>PHOTO (COLOR): FIGURE 2 Scatterplots for the correlations of the new few moves available problems with the original few- versus many moves available problems (adapted from Ash & Wiley, 2006), and under fatigue and no-fatigue conditions.</p> <p>PHOTO (COLOR): FIGURE 3 Scatterplots for the working memory capacity correlations with the original few- versus many moves available problems (adapted from Ash & Wiley, 2006), and under fatigue and no-fatigue conditions.</p> <hd id="AN0133587424-16">Correlation with WMC</hd> <p>Again, the 95% confidence intervals of Pearson correlations were directly compared. The positive correlation between the WMC factor and the FMA variant accuracy, <emph>r</emph> = .36 [.22,.48], was virtually identical to the correlation for the MMA variant, <emph>r</emph> = .38 [.24,.50]. The condition did not affect this data pattern, as respective confidence intervals strongly overlapped both in the fatigue condition, <emph>r</emph><subs>FMA-WMC</subs> = .42 [.23,.58] and <emph>r</emph><subs>MMA-WMC</subs> = .47 [.29,.62], and no-fatigue, <emph>r</emph><subs>FMA-WMC</subs> = .31 [.11,.49] and <emph>r</emph><subs>MMA-WMC</subs> = .28 [.08,.46] conditions (see Figure 3). However, the overall effect of condition yielded significant difference in correlation between problem solving accuracy and WMC, with the fatigue condition yielding a significantly stronger positive correlation,<emph>r</emph><subs>WMC</subs> = .46 [.34,.57], than the no-fatigue condition, r<subs>WMC</subs> = .30 [.16,.43]. A similar difference between the fatigue condition, r<subs>WMC</subs> = .50 [.38,.60], and the no-fatigue condition, <emph>r</emph><subs>WMC</subs> = .34 [.20,.46], was observed for the new FMA problems. Overall, these correlations were positive and moderate to strong, replicating the previous finding on the substantial WMC-insight problem solving relationship (Chuderski & Jastrzębski, 2018b).</p> <p>The respective correlations between the problem-solving performance and single WM tasks closely matched the aforementioned pattern of results, except for the fact that correlation coefficients for the letter span task were, on average, lower by a negligible Δ<emph>r</emph> = -.03, and for the updating task by Δ<emph>r</emph> = -.07. This drop in effect size was naturally related to a lower reliability of a single WM task, in comparison to the WMC factor composed of both tasks.</p> <p>Constraining the analysis only to solutions accompanied by the self-reported insight, WMC was still significantly related to the number of insight solutions in the FMA variant, <emph>r</emph><subs>WMC</subs> = .17 [.02,.31], similarly as it was in the MMA variant, <emph>r</emph><subs>WMC</subs> = .21 [.06,.35]. The WMC correlation with insight solutions for the new FMA problems was also significant and comparable, <emph>r</emph><subs>WMC</subs> = .18 [.08,.28]. No reliable differences in correlations between WMC and self-reported insight were found between the fatigue and no-fatigue conditions.</p> <hd id="AN0133587424-17">DISCUSSION</hd> <p>This study used a large sample, 12 previously validated insight problems, and two established WM tasks to examine the role of the size of initial problem representation for insight problem solving. Such a size was assumed by Ormerod et al. (2002) and Ash and Wiley (2006) to be small for the FMA problems, but large for the MMA problems. To sum up the results, evidence was found that accuracy in the FMA and MMA problem variants was comparable. No reliable difference was found for the proportion of solutions reported as accompanied by insight. Fatigue strongly increased the proportion of self-reported insight for the original problems (but not for the new problems), but this fact did not yield a better insight problem performance under fatigue. Moreover, fatigue affected subjective experience in the same way in the FMA as in the MMA variants. The original FMA problems correlated with the new FMA problems with the same strength as did the original MMA problems. There was no reliable difference between the FMA and MMA variants in their correlation with WMC. Finally, even when only solutions accompanied by self-reported insight were taken into account, solution accuracy still correlated reliably with WMC, both for the original and the new insight problems.</p> <p>Taken together, these results lead to a conclusion that the size of initial representation of an insight problem plays no role in the process of solving such a problem. Thus, this study failed to replicate Ormerod et al.'s (2002) observation: "When no ... move was available, ... an alternative operator was quickly discovered. However, when ... moves were available, the alternative successful move remained elusive" (p. 797). In contrast, it was found that blocking the initial faulty representation minimally affected the number of solutions, as well as their insightfulness. Especially, the proportion of participants who correctly solved the FMA variant of the original eight-coin problem (<emph>M</emph> = 17.4%, including 4.5% who reported insight) was very similar to the respective proportion for its MMA variant (<emph>M</emph> = 21.1%, including 4.0% reporting insight). One possible reason for the FMA-MMA difference observed by Ormerod et al. is that they presented hints (absent in our study), which could somehow interact with the initial representation. However, our data suggest that when the sheer variant of the problem is attempted (with no additional hints), the small initial problem representation does not help in restructuring the problem.</p> <p>The suggested conclusion is supported by the virtually identical correlation strength between WMC and the FMA versus MMA problem accuracy. Such a correlation was substantial (14% of FMA problem-solving variance was explained by WMC), so it can be concluded that, regardless of the size of initial representation, the insight problem-solving process strongly relies on the mechanisms of active storage and control of information that are typical for the WM tasks. In this regard, the Ash and Wiley (2006) claimed that "individual differences in WM span measure predicted success only on the MMA problems" (p. 72) was clearly refuted. In fact, as their claim was based on a nonsignificant difference in WMC correlation strength between the FMA and MMA variants, such a claim was unlikely to be replicated in the bigger sample that was examined here.</p> <p>No evidence was found for specificity of the FMA problems relative to the MMA problems. If the MMA problems involved initial processing stages that are absent in the FMA problems, and if individual differences in such processing (the systematic, local search of a large initial problem representation) were unrelated to individual differences in processing involved in restructuring (supposedly, associative spread of activation and similar intuitive processing), the variance in the MMA scores should have two distinct sources, whereas the FMA scores variance should be driven only by one source, namely by the processes specific to restructuring, given that general factors such as cognitive ability and task-engagement are balanced between both variants. As a result, a stronger correlation of the FMA problem scores with the scores on other FMA problems (also driven by the same source of variance), as compared to the correlation with the MMA problem scores (the latter affected also by analytical, systematic processing), should be expected. However, strong evidence was revealed showing that both these relationships had comparable strength.</p> <p>One surprising effect was a more frequent reporting of insight solutions in the fatigue condition. However, as more frequent experience of insight was not accompanied by any increase in problem-solving accuracy, the changes in the subjective experience during problem solving might be unrelated to the actual cognitive processes that give rise to that experience. This evidence is incompatible with the suggestions made by the special-process theorists (e.g., Wiley & Jarosz, 2012) who assume that less controlled, more fuzzy attention (very likely to appear after 3-hr intensive cognitive performance) may facilitate creative thinking in general, and insight problem solving in particular. It would also be problematic for the special-process theory to explain why under fatigue, that is, when participants could be tending toward associative, intuitive processing, instead of controlled, systematic thinking, the correlation between problem solving performance and WMC in fact was stronger than the correlation pertaining to problem solving with the <emph>fresh mind</emph> (less likely involving fuzzy attention). Actually, the special-processes theory predicts the opposite direction of this difference. This effect can be more easily explained on the grounds of the nothing-special account, given the assumption that sustained attention of low-WMC participants is vulnerable (see McVay & Kane, 2009; Unsworth, Redick, Lakey, & Young, 2010). Such participants might be able to catch up to high-WMC people in controlling attention early in the experiment, leveling the WMC-related differences in insight problem solving to some extent but, after intensive cognitive effort, their attention control might deteriorate, although that of high-WMC people might still be effective—a case that would amplify the WMC-related differences. Anyway, the effect of fatigue on both subjective experience and WMC were comparable across the FMA and MMA variants; this outcome, then, further supports the conclusion rejecting any visible role of the size of initial representation for performance on insight problems.</p> <p>This study has certain limitations. First, although the FMA problems have been intentionally designed to promote direct restructuring and block local, combinatorial strategies, such tasks might be still too artificial to induce the full-blown, holistic re-representation of the problem, as might occur in more realistic situations (see Beaty, Nusbaum, & Silvia, 2014). Thus, the current (but also Ormerod et al.'s, 2002; Ash & Wiley's, 2006) FMA/MMA manipulation might not have had a sufficient strength to induce the differences between the two problem variants, both in objective and subjective markers of problem solving. However, even if that was the case, and a more advanced design would be necessary, this study brought evidence that previous reports on the substantial differences between the FMA and MMA problems were premature. Second, the 3-hr session used in the fatigue condition might be too short or too easy (at least for some participants) to induce strong fatigue effects, and could even work as a warm-up. As neither objective measures of performance decrement nor subjective measures of fatigue were used, this option cannot be ruled out, although it seems unlikely. However, the effects pertaining to the FMA versus MMA distinction were not related to the fatigue manipulation, so any procedural shortcomings in inducing fatigue cannot affect conclusions pertaining to the null role of initial representation in insight problem solving. Finally, the one-dimensional measure of insight might not sufficiently capture subjective experience during problem solving, as some authors recommend probing up to four distinct aspects of insight (suddenness, expectancy, obviousness, and pleasure; see Salvi et al., 2016; Webb et al., 2016), or even promote recording complete verbal protocols related to problem solving (see Fleck & Weisberg, 2013). More-dimensional indices of subjective experience could have added further information; however, first, this study was primarily focused on objective measures of performance accuracy, and, second, too detailed probing of participants' internal states might have made them reflect too much on the goals of the study (the tasks themselves) and, as a result, could change the adopted strategy of coping with the insight problems. In this regard, verbal protocols would be even more invasive and barely applicable in the 353-people sample. Simple check-boxes minimally affected problem-solving performance, but might still have been effective in differentiating between the solutions worked out in a systematic, gradual way and the sudden, unexpected solutions.</p> <p>Overall, this study contributes to the theoretical dispute between the two main approaches to the nature of insight in problem solving. The results demonstrated little support for the special-process theories, which predict the important role of initial problem representation size in the processing of insight problems. Such approaches claim that only problems yielding minimal initial representation can involve restructuring that results from associative, intuitive, holistic processes specific for insight, whereas large initial representations result in confounding such processes with more systematic, controlled cognitive processing (e.g., Ash & Wiley, 2006). The study showed that the size of initial problem representation has no effect on either the objective nor subjective markers of problem solving performance. This result implicates further support in favor of the nothing-special theories of insight (e.g., Weisberg & Alba, 1981), which predict that regardless of the nature of an insight problem, its solution relies on the systematic, controlled processes of attention, memory, imagery, and reasoning. Insight might be just a subjective phenomenon occurring when the mental representation yielded by such processes passes the neurocognitive level of activation, above which it starts to be consciously perceived by a solver.</p> <ref id="AN0133587424-18"> <title> REFERENCES </title> <blist> <bibl id="bib1" type="bt">1</bibl> <bibtext> Ansburg, P. I., & Dominowski, R. L. ( 2000 ). Promoting insightful problem solving. <emph>Journal of Creative Behavior</emph>, 34, 30 - 60. doi: 10.1002/j.2162-6057.2000.tb01201.x</bibtext> </blist> <blist> <bibl id="bib2" type="bt">2</bibl> <bibtext> Ash, I. K., Cushen, P. J., & Wiley, J. ( 2009 ). Obstacles in Investigating the role of restructuring in insightful problem solving. <emph>The Journal of Problem Solving</emph>, 2, 6 - 41. doi: 10.7771/1932-6246.1056</bibtext> </blist> <blist> <bibl id="bib3" type="bt">3</bibl> <bibtext> Ash, I. K., & Wiley, J. ( 2006 ). The nature of restructuring in insight: An individual differences approach. <emph>Psychonomic Bulletin & Review</emph>, 13, 66 - 73. doi: 10.3758/BF03193814</bibtext> </blist> <blist> <bibl id="bib4" type="bt">4</bibl> <bibtext> Ashcraft, M. H. ( 1994 ). <emph>Human memory and cognition</emph> ( 2nd ed.). New York, NY : HarperCollins.</bibtext> </blist> <blist> <bibl id="bib5" type="bt">5</bibl> <bibtext> Batchelder, W. H., & Alexander, G. E. ( 2012 ). Insight problem solving: A critical examination of the possibility of formal theory. <emph>The Journal of Problem Solving</emph>, 5, Article 6. doi: 10.7771/1932-6246.1143</bibtext> </blist> <blist> <bibl id="bib6" type="bt">6</bibl> <bibtext> Beaty, R. E., Nusbaum, E. C., & Silvia, P. J. ( 2014 ). Does insight problem solving predict real-world creativity? <emph>Psychology of Aesthetics, Creativity, and the Arts</emph>, 8, 287 - 292. doi: 10.1037/a0035727</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Beeftink, F., van Eerde, W., & Rutte, C. G. ( 2008 ). The effect of interruptions and breaks on insight and impasses: Do you need a break right now? <emph>Creativity Research Journal</emph>, 20, 358 - 364. doi: 10.1080/10400410802391314</bibtext> </blist> <blist> <bibl id="bib8" type="bt">8</bibl> <bibtext> Bowden, E. M., Jung-Beeman, M., Fleck, J., & Kounios, J. ( 2005 ). New approaches to demystifying insight. <emph>Trends in Cognitive Sciences</emph>, 9, 322 - 328. doi: 10.1016/j.tics.2005.05.012</bibtext> </blist> <blist> <bibl id="bib9" type="bt">9</bibl> <bibtext> Chu, Y., & MacGregor, J. N. ( 2011 ). Human performance on insight problem solving: A review. <emph>The Journal of Problem Solving</emph>, 3, 119 - 150. doi: 10.7771/1932-6246.1094</bibtext> </blist> <blist> <bibtext> Chuderski, A., & Jastrzębski, J. ( 2018a ). The relationship of insight problem solving to analytical thinking. To appear. In F. Vallée-Tourangeau (Ed.), <emph>Insight: On the origin of ideas</emph> (pp. 120-142). Abingdon, UK : Routledge.</bibtext> </blist> <blist> <bibtext> Chuderski, A., & Jastrzębski, J. ( 2018b ). Much ado about Aha! Insight problem solving is strongly related to working memory capacity and reasoning ability. <emph>Journal of Experimental Psychology: General</emph>, 147, 257 - 281. doi: 10.1037/xge0000378</bibtext> </blist> <blist> <bibtext> Danek, A. H., Wiley, J., & Öllinger, M. ( 2016 ). Solving classical insight problems without Aha! experience: 9 dot, 8 coin, and matchstick arithmetic problems. <emph>The Journal of Problem Solving</emph>, 9, Article 4. doi: 10.7771/1932-6246.1183</bibtext> </blist> <blist> <bibtext> Davidson, J. E. ( 1995 ). The suddenness of insight. In R. J. Sternberg & J. E. Davidson (Eds.), <emph>The nature of insight</emph> (pp. 125 - 155 ). New York, NY : Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Davidson, J. E., & Sternberg, R. A. ( 1984 ). The role of insight in intellectual giftedness. <emph>Gifted Child Quarterly</emph>, 28, 58 - 64. doi: 10.1177/001698628402800203</bibtext> </blist> <blist> <bibtext> DeYoung, C. G., Flanders, J. L., & Peterson, J. B. ( 2008 ). Cognitive abilities involved in insight problem solving: An individual differences model. <emph>Creativity Research Journal</emph>, 20, 278 - 290. doi: 10.1080/10400410802278719</bibtext> </blist> <blist> <bibtext> Dow, G. T., & Mayer, R. E. ( 2004 ). Teaching students to solve insight problems: Evidence for domain specificity in creativity training. <emph>Creativity Research Journal</emph>, 16, 389 - 398. doi: 10.1080/10400410409534550</bibtext> </blist> <blist> <bibtext> Duncker, K. ( 1945 ). On problem-solving. <emph>Psychological Monographs</emph>, 58, 270. doi: 10.1037/h0093599</bibtext> </blist> <blist> <bibtext> Fleck, J. I., & Weisberg, R. W. ( 2013 ). Insight versus analysis: Evidence for diverse methods in problem solving. <emph>Journal of Cognitive Psychology</emph>, 25, 436 - 463. doi: 10.1080/20445911.2013.779248</bibtext> </blist> <blist> <bibtext> Holyoak, K. ( 2005 ). Analogy. In K. J. Holyoak & R. Morrison (Eds.), <emph>Cambridge handbook of thinking and reasoning</emph> (pp. 117-142). Cambridge, UK : Cambridge Univ. Press.</bibtext> </blist> <blist> <bibtext> Katona, G. ( 1940 ). <emph>Organizing and memorizing</emph>. New York, NY : Columbia University Press.</bibtext> </blist> <blist> <bibtext> Knoblich, G., Ohlsson, S., Haider, H., & Rhenius, D. ( 1999 ). Constraint relaxation and chunk decomposition in insight problem solving. <emph>Journal of Experimental Psychology: Learning, Memory, & Cognition</emph>, 25, 1534 - 1555.</bibtext> </blist> <blist> <bibtext> Koffka, K. ( 1935 ). <emph>Principles of Gestalt psychology</emph>. New York, NY : Harcourt, Brace and Company.</bibtext> </blist> <blist> <bibtext> Kounios, J., & Beeman, M. ( 2014 ). The cognitive neuroscience of insight. <emph>Annual Review Of Psychology</emph>, 65, 71 - 93. doi: 10.1146/annurev-psych-010213-115154</bibtext> </blist> <blist> <bibtext> Larson, G. E., Merritt, C. R., & Williams, S. E. ( 1988 ). Information processing and intelligence: some implications of task complexity. <emph>Intelligence</emph>, 1, 131 - 147. doi: 10.1016/0160-2896(88)90012-8</bibtext> </blist> <blist> <bibtext> 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. <emph>Journal of Experimental Psychology: Learning, Memory, & Cognition</emph>, 27, 176 - 201.</bibtext> </blist> <blist> <bibtext> Maier, N. R. F. ( 1930 ). Reasoning in humans: I. On direction. <emph>Journal of Comparative Psychology</emph>, 10, 115 - 143. doi: 10.1037/h0073232</bibtext> </blist> <blist> <bibtext> McVay, J. C., & Kane, M. J. ( 2009 ). Conducting the train of thought: Working memory capacity, goal neglect, and mind wandering in an executive-control task. <emph>Journal of Experimental Psychology: Learning, Memory, and Cognition</emph>, 35, 196 - 204.</bibtext> </blist> <blist> <bibtext> Mednick, S. A. ( 1962 ). The associative basis of the creative process. <emph>Psychological Review</emph>, 69, 220 - 232. doi: 10.1037/h0048850</bibtext> </blist> <blist> <bibtext> Newell, A., & Simon, H. A. ( 1972 ). <emph>Human problem solving</emph>. Englewood Cliffs, NJ : Prentice Hall.</bibtext> </blist> <blist> <bibtext> Ohlsson, S. ( 1992 ). Information-processing explanations of insight and related phenomena. In M. Keane & K. Gilhooly (Eds.), <emph>Advances in the psychology of thinking</emph> (pp. 1 - 44 ). London, England : Harvester- Wheatsheaf.</bibtext> </blist> <blist> <bibtext> Ohlsson, S. ( 2011 ). <emph>Deep learning. How the mind overrides experience</emph>. New York, NY: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Ormerod, T. C., MacGregor, J. N., & Chronicle, E. P. ( 2002 ). Dynamics and constraints in insight problem solving. <emph>Journal of Experimental Psychology: Learning, Memory, and Cognition</emph>, 28, 791 - 799.</bibtext> </blist> <blist> <bibtext> Perkins, D. ( 1981 ). <emph>The mind's best work</emph>. Cambridge, MA : Harvard University Press.</bibtext> </blist> <blist> <bibtext> Salvi, C., Bricolo, E., Kounios, J., Bowden, E., & Beeman, M. ( 2016 ). Insight solutions are correct more often than analytic solutions. <emph>Thinking & Reasoning</emph>, 22, 443 - 460. doi: 10.1080/13546783.2016.1141798</bibtext> </blist> <blist> <bibtext> Unsworth, N., Redick, T. S., Lakey, C. E., & Young, D. L. ( 2010 ). Lapses in sustained attention and their relation to executive control and fluid abilities: An individual differences investigation. <emph>Intelligence</emph>, 38, 111 - 122. doi: 10.1016/j.intell.2009.08.002</bibtext> </blist> <blist> <bibtext> Van Stockum, C. A., Jr., & DeCaro, M. S. ( 2014 ). Enclothed cognition and controlled attention during insight problem solving. <emph>The Journal of Problem Solving</emph>, 7, 73 - 83. doi: 10.7771/1932-6246.1164</bibtext> </blist> <blist> <bibtext> Webb, M., Little, D., & Cropper, S. ( 2016 ). Insight is not in the problem: Investigating insight in problem solving across task types. <emph>Frontiers in Psychology</emph>, 7, Article 1424. doi: 10.3389/fpsyg.2016.01424</bibtext> </blist> <blist> <bibtext> Weisberg, R. W. ( 2006 ). <emph>Creativity: Understanding innovation in problem solving, science, invention, and the arts</emph>. Hoboken, NJ : Wiley.</bibtext> </blist> <blist> <bibtext> Weisberg, R. W. ( 2013 ). On the "demystification" of insight. <emph>A Critique of Neuroimaging Studies of Insight Creativity Research Journal</emph>, 25, 1 - 14.</bibtext> </blist> <blist> <bibtext> Weisberg, R. W. ( 2015 ). Toward an integrated theory of insight in problem solving. <emph>Thinking & Reasoning</emph>, 21, 5 - 39. doi: 10.1080/13546783.2014.886625</bibtext> </blist> <blist> <bibtext> Weisberg, R. W., & Alba, J. W. ( 1981 ). An examination of the alleged role of "fixation" in the solution of several "insight" problems. <emph>Journal of Experimental Psychology: General</emph>, 110, 169 - 192. doi: 10.1037/0096-3445.110.2.169</bibtext> </blist> <blist> <bibtext> Wertheimer, M. ( 1945 ). <emph>Productive thinking</emph>. New York, NY : Harper.</bibtext> </blist> <blist> <bibtext> Wieth, M., & Zacks, R. ( 2011 ). Time of day effects on problem solving: When the non-optimal is optimal. <emph>Thinking & Reasoning</emph>, 17, 387 - 401. doi: 10.1080/13546783.2011.625663</bibtext> </blist> <blist> <bibtext> Wiley, J., & Jarosz, A. F. ( 2012 ). How working memory capacity affects problem solving. <emph>Psychology of Learning and Motivation</emph>, 56, 185 - 227.</bibtext> </blist> </ref> <aug> <p>By Adam Chuderski and Jan Jastrzębski</p> </aug> <nolink nlid="nl1" bibid="bib351" firstref="ref1"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1200015
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: No Role of Initial Problem Representation in Insight Problem Solving
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Chuderski%2C+Adam%22">Chuderski, Adam</searchLink><br /><searchLink fieldCode="AR" term="%22Jastrzebski%2C+Jan%22">Jastrzebski, Jan</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Creativity+Research+Journal%22"><i>Creativity Research Journal</i></searchLink>. 2018 30(4):428-438.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 11
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2018
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Creative+Thinking%22">Creative Thinking</searchLink><br /><searchLink fieldCode="DE" term="%22Problem+Solving%22">Problem Solving</searchLink><br /><searchLink fieldCode="DE" term="%22Correlation%22">Correlation</searchLink><br /><searchLink fieldCode="DE" term="%22Fatigue+%28Biology%29%22">Fatigue (Biology)</searchLink><br /><searchLink fieldCode="DE" term="%22Volunteers%22">Volunteers</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22Accuracy%22">Accuracy</searchLink><br /><searchLink fieldCode="DE" term="%22Intuition%22">Intuition</searchLink><br /><searchLink fieldCode="DE" term="%22Thinking+Skills%22">Thinking Skills</searchLink>
– Name: Subject
  Label: Geographic Terms
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Poland%22">Poland</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/10400419.2018.1531674
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1040-0419
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The literature on insight problems--problems that supposedly can only be solved by rejection of an initial faulty problem representation and sudden comprehension of another, nonobvious representation (restructuring)--suggests that the size of initial representations affects the very process of problem solving. Large initial representations impose systematic, analytical search, whereas only small representations promote intuitive, associative processes assumed by some theorists to underpin insight. In a group of 353 young healthy participants, 6 previously validated insight problems were applied in either a small or large initial representation variant. Results demonstrated no reliable difference in performance between the problem variants with regard to (a) solution accuracy, (b) self-reported insight accompanying solutions, (c) effects of fatigue, (d) correlations with another 6 small representation-size problems, and (e) correlations with working memory capacity (which were notable). This outcome suggests that the size of initial faulty representation plays no role in insight problem solving process, supporting the account assuming its strong similarity to systematic, analytical problem solving.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: Ref
  Label: Number of References
  Group: RefInfo
  Data: 44
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2018
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1200015
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1200015
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/10400419.2018.1531674
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 11
        StartPage: 428
    Subjects:
      – SubjectFull: Creative Thinking
        Type: general
      – SubjectFull: Problem Solving
        Type: general
      – SubjectFull: Correlation
        Type: general
      – SubjectFull: Fatigue (Biology)
        Type: general
      – SubjectFull: Volunteers
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: Accuracy
        Type: general
      – SubjectFull: Intuition
        Type: general
      – SubjectFull: Thinking Skills
        Type: general
      – SubjectFull: Poland
        Type: general
    Titles:
      – TitleFull: No Role of Initial Problem Representation in Insight Problem Solving
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Chuderski, Adam
      – PersonEntity:
          Name:
            NameFull: Jastrzebski, Jan
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2018
          Identifiers:
            – Type: issn-print
              Value: 1040-0419
          Numbering:
            – Type: volume
              Value: 30
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Creativity Research Journal
              Type: main
ResultId 1