Problem solving can either start in the restricted search space and then continue in the extended search space after restructuring, or could start in the extended search space right away. Examples plotted in green (#18 and #4) are from solvers, examples plotted in red (#79 and #7) are from non-solvers. The fourth stage could be very brief (depending on the problem) because problem solvers often find the solution shortly after representational change, if they have the right insight. Our main finding—that in most cases problem solving stages do not follow the simple stage model of insight—highlights the need for revising the model. The latter start to look for the solution in an extended search space, in which, eventually they find the solution (in the Five-Square problem, they start putting sticks to further positions, and decompose the initial cross shape). The dynamics of search, impasse, and representational change provide a coherent explanation of difficulty in the nine-dot problem. Another interesting direction would be to try to find the neural correlates of problem solving stages and thus to add the neural level of description to the narrative. It has automatically checked whether a correct solution was achieved. doi: 10.1016/0010-0285(90)90004-N, Danek, A., Fraps, T., and von Mueller, A. 25, 1534–1555. According to Öllinger et al. J. Exp. The frequency of grid positions used as target positions in the moves of solvers in the first quarter of their moves (left) and the last quarter of their moves (right). It is not a problem, if you cannot solve the task, but if you close the program before 15 min elapsed, your submission will be invalid.”. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). Their model emphasizes the difference between routes with and without incubation: although it is possible that one cannot solve the task even after incubation, but it is not possible to solve the task without it. According to the paired t-test, the difference was not significant, t(33) = 0.46, p = 0.65, d = 0.16. Bull. Since we did not find differences between the groups, we merged them for the rest of the analyses, except when we looked at the impasse monitoring group separately for the subjective impasse feeling analysis. The underlying cognitive process is supposed to be conscious search in the initial, constrained search space (e.g., MacGregor et al., 2001). Participants were allowed to quit the online experiment any time without providing a reason. We propose an objective behavioral measure to define problem solving stages individually. 111, 326–330. Restructuring leads to an extended search space, which includes more candidate solutions and among them, the solution, and thus repetitions and inactivity is less frequent than in the restricted search space. We did this, because it turned out that participants did not use all segments of the figure equivalently: the top left portion was used less often than the rest. It is also possible that the “Aha!” experience is not linked to representational change (Danek et al., 2014). Psychol. Alternatively, the program reset automatically after unsuccessful attempts of three moves and provided feedback (“This is not a correct solution. After solving the problem or reaching the upper time limit of 15 min, participants were asked to post the output log file to the experimenter via our website and to complete a short online questionnaire (about their age, gender, handedness, vision, educational background, nationality, mother tongue, and level of English). Psychol. We aimed to validate the stage models of insight by testing their predictions against this data. Comp. Psychol. An information processing theory of restructuring and insight. doi: 10.1162/NECO_a_00031, Fernando, C., Szathmáry, E., and Husbands, P. (2012). In the Five-Square problem (Katona, 1940), the task is to reduce the number of five squares seen on Figure 1 to four squares. Box-and-arrow type stage models explain the order of problem solving stages, and the underlying theory of representational change explains—to some extent—what drives the process through these stages. Other participants solved the problem by the application of heuristics that helped them to restrict the search space. As can be seen in Figure 1, the initial configuration of sticks was symmetric. The prepared mind. The third possibility also appears to be difficult because of the possibility of heavy rains. The third stage, insight, is named after its affective correlate, but its most important feature is the cognitive process of representational change, or restructuring (Knoblich et al., 1999). Ohlsson, S. (1992). doi: 10.1016/j.tics.2005.05.012, PubMed Abstract | CrossRef Full Text | Google Scholar, Bowers, K., Regehr, G., Balthazard, C., and Parker, K. (1990). We see a lack of clarity in this regard in the literature, so here we would like to disentangle the different levels of description, and identify which phenomena belong to the cognitive, behavioral and affective level (Table 1). In the impasse monitoring group 33% of the solvers (8 participants out of 24 solvers) and 79% of the non-solvers (34 participants out of 43 non-solvers) pressed the impasse button at least once. Öllinger et al. According to the analysis, the sample size for a matched-pairs t-test with these parameters should be 34, which we multiplied by 2.5 thinking that about half of the participants won't be able to solve the task and not all of these will press the impasse button. Psychol. Öllinger et al. (2008). We did this, because we could not monitor their behavior while the program was closed. He examined move durations and eye-movements during two versions of the car park problem. They received one dollar as a payment. 20, 358–364. This moment of enlightenment is usually—according to some, necessarily (Ohlsson, 1992; Knoblich et al., 1999, 2001; Jones, 2003; Öllinger et al., 2014a)—preceded by a longer period of impasse when the problem solver gets stuck and has no idea how to proceed. The overall solution rate was 51%; 55% in the control group and 49% in the impasse monitoring group (Table 2). The program recorded the movement of the sticks and the button presses (Impasse button and Restart button) along with the time that passed since starting the task. 1, eds D. H. Krantz and R. C. Atkinson, R. D. Luce, and P. Suppes (San Francisco, CA: W.H. The first possibility will not work as my neighbours have gone out. Repetitions and inactivity on the cognitive level are supposedly caused by the confined search space: when there is a finite number of candidate solutions, after trying enough of them, the problem solver inevitably bumps into repeating previous ones, unless the problem solver takes a break. Problem solvers who overcome this constraint expand their search space. doi: 10.1146/annurev-psych-010213-115154, Kounios, J., Frymiare, J. L., Bowden, E. M., Fleck, J. I., Subramaniam, K., Parrish, T. B., et al. Table 2. We introduced two experimental groups. Figure 5. Please, try again!”). J. Exp. Did you have this feeling before or when you solved the task?”. The link between the objectively defined impasse stage and the impasse reports of participants has several elements: the cognitive impasse state causes repetitions and inactivity on the behavioral level that we used to objectively define impasse. Neural Comput. Creat. In the above example of providing sitting accommodation for the large number of guests, I can think of various solutions. In the two groups, 74% of the solvers (29 participants out of 39 solvers) and 10% of the non-solvers (6 participants out of 62 non-solvers) reported insight feeling after the task. It involves imagining, thinking, trial-and-error and insight. In their model, fixation is an unconscious state, where people keep repeating previous problem solving attempts and incubation happens when the problem solver is in the conscious state of impasse: after unsuccessful attempts accumulate and reach a threshold, problem solvers realize that they are stuck and they take a break.