Surface and Deep Processing: Cognitive Behaviors of Aha! Moments (Part II)

This article is a continuation of a research entry from the May 24, 2021 edition:

The spectrum of early research on insight ranges from observing changes in behavior and understanding psychological patterns that influence learning (Bühler, 1907; Duncker & Lee, 1945; Wallas, 1926), to the present and how insight is a unique form of learning. There are a number of theories on insight; at present, no one theory dominates interpretation (Kounios & Beeman, 2015; Sternberg 1996). In spite of differences between theories, they share two principles: (a) sudden, conscious change in a person’s representation of a stimulus, situation, event, or problem (Davidson, 1995; Kaplan & Simon, 1990), and (b) the change occurs unexpectedly (Jung-Beeman, et al., 2004; Kounios & Beeman, 2014; Metcalfe, 1986). Further, a strong correlation has been demonstrated between moments of insight and increased engagement in learning, positive boost in mood, and greater likelihood of more moments of insight (Kizilirmak, Da Silva, Imamoglu, & Richardson-Klavehn, 2016; Kounios & Beeman, 2014). Aha! moments have been shown to increase and enhance memory performance (Ash, Jee, & Wiley, 2012; Auble, Franks, & Soraci, 1979; Danek, Fraps, von Müller, Grothe, & Öllinger, 2013; Dominowski & Buyer, 2000; Kizilirmak, Da Silva, Imamoglu, & Richardson- Klavehn, 2016), reliably grounded on insight’s proven ability to, “comprise associative novelty, schema, congruency, and intrinsic reward” (Kizilirmak et al., 2016, p. 1).

The observation and categorization of these moments can also be a source of valuable information for theorists and educators. Crocker and Algina (1986) demonstrate this operationalization in order to, “establish some rule of correspondence between the theoretical construct and observable behaviors that are legitimate indicators” (p. 4). The suddenness of Aha! moments makes observing behavioral changes (and subsequent changes in understanding) more dramatic and pronounced, as opposed to more gradual and deductively reasoned outcomes. Baker, Goldstein, and Heffernan (2010) have observed this distinction by studying the precise moment when understanding changes – graphing the precise moment of learning in humans. Baker et al. (2010) diagram the shift in surface to deep processing by showing the, “differences between gradual learning (such as strengthening of a memory association) and learning given to ‘eureka’ moments, where a knowledge component is understood suddenly” (p. 13).

Figure 5. A Single Student’s Performance on a Specific Knowledge Concept (Baker et al., 2010, p. 13)

Baker et al. explain that, “entering a common multiple” (left, Figure 5) results in a “spiky” graph, indicating eureka learning, while “identifying the converted value in the problem statement of a scaling problem” (right, Figure 5) results in a relatively smooth graph, indicating more gradual learning (p. 14).

Another important implication to consider is that deep processing seems to create greater investment in learning, along with more positive outcomes for students. Dolmans, Loyens, Marcq, and Gijbels (2016) have reviewed 21 different studies that reported on surface and deep processing strategies in relation to problem-based learning, and concluded that students using deep processing strategies use, “the freedom to select their own resources to answer the learning issues, which gives them ownership over their learning” (p. 1097). This ownership suggests a strong link between intrinsic and autonomous motivation, resulting in stronger and longer-lasting outcomes. Dolmans et al. also report that surface learning strategies with problem-based learning had a similar negative effect, stating:

a high perceived workload will more likely result in surface approaches to studying and might be detrimental for deep learning. Students who perceive the workload as high in their learning environment are more likely to display a lack of interest in their studies as well as exhaustion. This is particularly true for beginning [problem-based learning] students. (p. 1097)

““If we get the deep processing, we almost always get the surface, but with much richer and rewarding outcomes!””

— J. Littlejohn, Elementary School Math Instructor

The meta-analysis concluded by affirming these positive deep processing outcomes do not come at the cost of the various surface processing benefits (p. 1097). Deep processing strategies employed by learners have also been shown to boost long-term recall of information and wider conceptual understanding. Jensen, McDaniel, Woodard, and Kummer (2014) report that learners who utilized deep processing learning strategies while preparing for high-level assessments (i.e., problem solving, analysis, and evaluation) performed better than students that did not, and these students retained a, “deep conceptual understanding of the material and better memory for the course information” (p. 307). Jensen et al. (2014) have found that this higher level of cognitive processing and understanding also made transfer-appropriate processing more likely. This conclusion is supported by similar research conducted on learners using deep processing strategies and motivated by deeper conceptual understanding (Carpenter, 2012; Fisher & Craik, 1977; McDaniel, Friedman, & Bourne, 1978; McDaniel, Thomas, Agarwal, McDermott, & Roediger, 2013). Students using transfer-appropriate processing outcomes showed improved mastery and conceptual development greater than surface strategies and beyond the at-hand assessment; the gains were greater in current work and also in future assessments utilizing deep processing strategies. This developed processing strategy offers learners the greatest advantage in future outcomes. Studying Aha! moments in learning makes understanding surface processing and shifts into deep processing more probable, and the transfer-appropriate advantages more common, offering teachers a tremendous perspective into how to best develop pedagogy.

The Cognitive Neuroscience of Insight

Kounios and Beeman (2014) report on the variety of factors that influence and create insight moments. Their work represents the most comprehensive and provocative investigation on insight, focusing on changes in cognitive behaviors as a result of having experienced insight, whether through suddenly realizing a solution or suddenly becoming aware of one. Kounios and Beeman define insight occurring,

when a person suddenly reinterprets a stimulus, situation, or event to produce a nonobvious, nondominant interpretation. This can take the form of a solution to a problem (an “aha moment”), comprehension of a joke or metaphor, or recognition of an ambiguous percept. (p. 71)

Research shows that insight moments are distinct from other forms of learning, analytical thinking and processing in particular (Kounios & Beeman, 2014; Sternberg & Davidson, 1995). Kounios and Beeman (2015) report that, “except for a few limited and arguable counterexamples, only humans—most humans—have insights. It’s a basic human ability” (p.11).

Reliable production of insight moments has been accomplished through several scientific measures. Some early research made productive use of the Remote Associates Test (RAT), initially created to assess human creative potential (Mednick & Mednick, 1962/1967/1968), in order to induce moments of insight. A classic example from the original tests are the three words same/tennis/head, each associated in some fashion (i.e., synonymously, compound, or semantically) with the solution word: match. Same and match are associated as synonyms; match-head (or sometimes, matchhead) is a compound word; and tennis match is a semantic association. If and when a solution is accomplished or revealed, the test verifiably produces a change in thinking, often in the form of an insight. Bowden and Jung-Beeman (2003) modified the original RAT problems and developed them into a new subset of the original test, more commonly known as the Compound Remote Associates Test (CRAT). These CRAT problems are classified into two categories: (1) homogeneous, meaning the solution word is a prefix (or suffix) to each of the three challenge words in the triad; and (2) heterogeneous, meaning that the solution word is a prefix (or suffix) for at least one of the challenge words and a prefix (or suffix) to the other words in the triad. An example of an easy CRAT are the three words print/berry/bird, each associated with the solution word blue, whether as prefix or suffix to each of the words in the triad. Blue is the prefix to blueprint; blue is the prefix to blueberry; and blue is the prefix to the word bluebird. This is an example of a homogenous CRAT. Bowden and Jung-Beeman created this new hybrid because it fosters conditions that allow participants to solve challenges more quickly. Solutions require less abstract thinking and tests produce stronger reliability, and because participants can solve them more quickly, more of them can be observed to form a more cohesive and comprehensive understanding of insight and non-insight moments (Bowden & Jung-Beeman, 2003, p. 636).

Important preconditions exist with insight moments that have reported positive impact on the likelihood, frequency, and strength of Aha! moments. Mood has been studied and its effect on enhancing insight has been shown. Ashby, Isen, and Turken (1999) and Isen, Daubman, and Nowicki (1987) report on these effects and it appears that positive mood and affect, “enhances insight and other forms of creativity, both when the mood occurs naturally and when it is induced in the laboratory” (p. 83). Mood also impacts attention, positively increasing or negatively diminishing capacity based on naturally occurring or an induced emotional state. Fredrickson and Branigan (2005) show a distinct connection to positive mood and a broadening of novel and varied stimuli, creating a stronger opportunity for exploratory behavior. Subsequently, the variability of excitement and related phenomena of Aha! moments can fluctuate based on the context. Kounios and Beeman affirm that,

insights are often accompanied by surprise and a positive burst of conscious emotion, but we do not consider these to be defining features because individual insights in a sequence of insights, as occur in many experimental studies, don’t all elicit such conscious affective responses. (p. 74)

Related research draws upon Fredrickson and Branigan’s (2005) broaden-and-build theory:

The broaden hypothesis states that positive emotions broaden the scopes of attention, cognition, and action, widening the array of percepts, thoughts, and actions presently in mind. A corollary narrow hypothesis states that negative emotions shrink these same arrays. (p. 2)

Attention allows learners to narrow or broaden their focus on stimuli, which in the case of an Aha! moment can be most valuable. A person might choose to focus most of their energy on a singular problem, intending to solve it, at the expense of broader focus. The combination of mood and attention create an even stronger likelihood for insight to occur (Easterbrook, 1959; Rowe, Hirsh, & Anderson, 2007).

Kounios and Beeman (2014) conclude aspirationally, hoping that, “researchers may look back at the early twenty-first century as the beginning of a golden age of insight research!” (p. 88).