Have you ever felt stuck while solving a problem? That frustrating feeling of hitting a mental wall is called an impasse. But what if these moments of feeling stuck are more than just obstacles? In recent work published in Cognition, researchers Wendy Ross and Selene Arfini find that these impasses play a crucial role in problem-solving, especially when they lead to sudden moments of insight—those “aha!” breakthroughs.
They argue that understanding when and how people feel stuck can shed light on cognitive processes that are essential not only for solving puzzles but for learning and innovation. This research also addresses the paradox of how impasses, though seemingly detrimental, can sometimes enhance problem-solving abilities.
Experiment 1 involved 125 participants tackling a “stumper” riddle within 60 seconds to elicit impasses. See the riddle below.
“Dame Dora owns an Old Masters painting in a heavy gilded frame. The cord for hanging the painting, as old as the painting itself, is made of thick 3- ply hemp, and is somewhat frayed. Dame Dora was thinking of replacing it. But before she could, a couple of hungry little mice invaded her mansion. Sneaking behind the painting, they chewed right through the cord. For a while nobody noticed because the painting didn’t budge. Explain the painting’s stability briefly.”
Participants also completed phenomenological measures gauging their experiences (e.g., feeling stupid, awareness of time) via free-text responses.
The researchers found that 69% of participants reported experiencing an impasse during the problem-solving tasks. Interestingly, those who did not report an impasse had a problem-solving success rate of 32%, which exceeded the 15% success rate among those who felt stuck. Feeling stuck was negatively associated with successful problem resolution, and experiencing moments of insight. Only 23% of those who felt stuck reported an insight experience, compared to 83% among those who did not feel impasse.
While the average latency to arrive at a correct solution was higher for those who felt stuck (57.1 seconds) compared to those who did not (51.9 seconds), this difference was not statistically significant, suggesting that feeling stuck does not always lead to delays in finding solutions. There were varied emotional responses to feeling stuck, ranging from frustration and anxiety to cognitive looping, highlighting the multidimensional nature of impasse in problem-solving.
Experiment 2 introduced 247 participants to similar riddles within 90 seconds. They completed a scale measuring impasse as unresolved, resolved, or not experienced. A substantial 74% of participants experienced impasse during the task. Those who resolved their impasse showed significantly better results, with 53% solving the problems correctly, compared to only 11% who did not resolve their impasse, and 52% who never experienced it. Those resolving impasses took about 84 seconds, slightly longer than those who never experienced impasses at around 69 seconds.
Participants who resolved their impasses reported higher levels of insight (67%) compared to those who did not experience impasse (45%). Those unable to resolve impasse saw drastically lower insight rates (7%). Resolved impasse was associated with more active and potentially motivating emotions like anxiety or happiness, while unresolved impasse was associated with more passive emotions such as sadness or hopelessness.
Experiment 3 included 119 participants and assessed the impact of time pressure by manipulating problem-solving time across several groups with time constraints of 30, 60, 90, and 120 seconds. Two parallel experiments were conducted: one with easier stimuli to increase the likelihood of resolving impasses, and another with stimuli matching the difficulty of previous experiments.
In the easier set of stimuli, correct solution rates improved with increased time—from 56% at 30 seconds to 72% at 120 seconds—though the increase was not statistically significant. However, with harder problems, time significantly impacted performance, improving from 36% at 30 seconds to 66% at 90 seconds. Most participants experienced impasse, but those who resolved it matched the success rates of those who never felt stuck. Longer times generally led to higher latency but not necessarily better outcomes.
Experiment 4 broadened the scope to different types of cognitive tasks, including a verbal insight puzzle, a verbal analytical puzzle, a word problem from the UK GCSE mathematics curriculum, and the original stumper used in earlier experiments. Participants were given 4 minutes per task.
The researchers observed higher solution rates and levels of insight in trials where impasse was resolved compared to those where it was not, indicating that resolving impasses can significantly enhance problem-solving performance. However, unresolved impasses drastically reduced insight levels. Consistent with prior experiments, there were longer times in unresolved cases. Resolved impasses were associated with higher affect and motivation levels, reinforcing the positive effects of overcoming impasses.
Interestingly, trials characterized by fast impasse reported higher solution rates and insights, showing that the subjective experience of speediness was beneficial, although it did not necessarily correlate with actual faster solution times.
Further research in more naturalistic settings will be important to better understand the real-world dynamics of feeling stuck.
This work challenges the traditional view of impasses as mere obstacles, proposing instead that they can act as catalysts for deeper learning and innovation.
The research, “Impasse-Driven problem solving: The multidimensional nature of feeling stuck”, was authored by Wendy Rossa and Selene Arfini.