Content uploaded by Denis Dumas
Author content
All content in this area was uploaded by Denis Dumas on Mar 15, 2021
Content may be subject to copyright.
CREATIVE ANALOGIES 1
THE INFLUENCE OF CREATIVE EXPERTISE ON THE SENSITIVITY AND
SELECTIVITY OF ANALOGICAL REASONING
Denis G. Dumasa
Yixiao Donga
Michael Dohertyb
aDepartment of Research Methods and Information Science, University of Denver
bActors Equity Association
Reference:
Dumas, D., Dong, Y., Doherty, M. (in press). The influence of creative expertise on the
sensitivity and selectivity of analogical reasoning. Mind, Brain, and Education.
Author Note
Correspondence concerning this article should be addressed to Denis Dumas, Department of
Research Methods and Information Science, University of Denver, Denver, CO, 80208. Email:
Denis.Dumas@du.edu. This research has followed all ethical guidelines of the American
Psychological Association and was approved by the Institutional Review Board at the University
of Denver.
CREATIVE ANALOGIES 2
Abstract
This study compared the analogical reasoning of three groups that differed in their creative
expertise: professional actors, undergraduate acting majors, and non-actors. Using an Analogy
Finding Task, in which participants identified valid and non-valid verbal analogies, three aspects
of participants’ analogical reasoning were measured: the number of analogies participants
selected as valid (Quantity), the rate of true-positive analogical identification (Sensitivity), and
the rate of true-negative identification of non-valid analogies (Selectivity). The Analogy Finding
Task was administered under both a baseline and a “think creatively” prompt. Results showed
that actors (professional or student) were significantly more Sensitive to valid analogies than
non-actors, and these creative experts were significantly more influenced by the “think
creatively” prompt, which increased the Quantity, and decreased the Selectivity, of actors’
analogical reasoning. To explain these results, we forward the general hypothesis that creative
experts may be more flexible in response to creativity-relevant contextual cues than non-experts.
Keywords: Analogical reasoning; creativity; expertise; actors
Lay Abstract
Recent research has suggested that individuals’ ability to think creatively is closely associated
with their ability to map analogies. Here we showed that individuals with demonstrated expertise
in a profession that requires creative thinking (i.e., stage and screen acting) perform differently
on analogical reasoning tasks than individuals who are not professionals in a creative area. We
interpret these data to mean that actors may be more flexible in their thinking than are non-
actors.
CREATIVE ANALOGIES 3
The Influence of Creative Expertise on
The Sensitivity and Selectivity of Analogical Reasoning
Analogical reasoning is a relational cognitive process whereby higher-order structural
similarities among pieces of information are mapped (Dumas et al., 2013; Goswami, 2013;
Richland & Simms, 2015). Psychologists and educational researchers have empirically linked
analogical reasoning to important educationally relevant outcomes across nearly every level of
schooling, from preschool (Walker et al., 2018) to medical school (Dumas et al., 2014), and
across domains of learning including science (Murphy et al., 2016), engineering (Dumas et al.,
2016), and history (Van Straaten et al., 2019), among many others. One other fundamentally
important mental attribute to which analogical and relational reasoning have been perennially
linked is creativity (Gentner et al., 1997; Green, 2018). Creativity—defined as the process by
which individuals generate ideas that are simultaneously novel and useful (Runco & Jaeger,
2012)—has long been theorized to require the mapping of structural patterns among ideas across
domains and contexts (Mednick, 1962), and therefore the role of analogical reasoning in the
creative process seems highly pertinent theoretically and pragmatically (Goel, 1997).
Relatively recently, a number of empirical studies have appeared that document the link
between creative thinking attributes and analogical reasoning across contexts including
naturalistic in vivo studies (Chan & Schunn, 2015), more controlled laboratory experiments
(Dumas, 2018), as well as neurological investigations using brain imaging (Green et al., 2012a;
Vartanian, 2011). In the current work, we take a related but distinct approach to studying the
relation between creativity and analogical reasoning: through the recruitment of individuals who
are trained to be experts at creative work (i.e., professional actors), individuals who are
acclimating within a creative domain (i.e., undergraduate acting majors), and individuals who
CREATIVE ANALOGIES 4
have not developed expertise in that domain (i.e., non-actors), we generally aim to understand
the influence of creative expertise on analogical reasoning.
What is Creative Expertise?
Typically, within the literature on expertise, experts are defined by their domain-specific
learning (e.g., Ericsson & Smith, 1991), and for this reason, it is less common to describe
expertise as existing on a domain-general mental attribute such as creativity. However, it is also
well-demonstrated that domain-specific expertise requires the development of a variety of more
domain-general abilities that are relevant to the domain (Schunn & Anderson, 2010). For
example, within educational psychology research (e.g., Alexander 2003; Fox, 2009; Wagner &
Stanovich, 1996) the argument has been built that experts in a number of academic domains can
also be accurately described as experts in the domain-general ability of reading comprehension.
In this way, an expert in a reading-heavy domain such as history might demonstrate their
historical expertise via a domain-specific practice (e.g., publishing a history book) but their
expertise would also be expected to manifest on a domain-general reading comprehension
measure.
Analogically, we posit the argument that experts in a domain that regularly requires
creative and divergent thinking—such as acting has been repeatedly shown to (e.g., Dumas,
Doherty, & Organisciak, 2020; Kogan, 2002; Noice & Noice, 1994; Stacey & Goldberg, 1953)—
would be expected to be able to demonstrate their expertise both in a domain-specific activity
(e.g., the interpretation and performance of a script) and on a domain-general measure designed
to relate to creative thought. In this way, just as domain-experts in a variety of academic domains
can be identified as expert readers, we suggest that creative experts be defined as those
individuals who have developed deep expertise in a domain that regularly requires creative
CREATIVE ANALOGIES 5
thinking, and therefore are also likely to be adept at creative thinking in a domain-general
setting. That said, it is important to note that the concept of creative expertise, although certainly
discussed by some in the literature (e.g., Simonton, 2014), is relatively under-studied, and as
such we see one purpose of the current study to be the empirical observation of this phenomenon
with a sample of expert performing artists to better define creative expertise itself, and to suggest
avenues of future study in this area.
Of course, the creative expertise we begin to describe in this investigation would
necessarily develop through formal and informal educational experiences within a domain that
requires creative thinking. Therefore, expertise can be conceptualized as inherently tied to the
educational contexts in which that expertise is developed (Alexander, 2003). For this reason, we
suggest that, when the influence of creative expertise is under investigation, it would be
elucidating to include participants not only who have already demonstrated expertise, but also
participants who are acclimating to expertise in the same domain, and who have formally
dedicated themselves to the development of competence and eventual expertise (e.g.,
undergraduate students majoring in a creativity-heavy discipline). Following this line of thought,
we contend that research on expertise effects is fundamentally educationally-relevant in that
differences in psychological attributes and processes among experts and students can often be
interpreted in light of pedagogy and education.
State Augmentation of Creative Analogical Reasoning
Within the literature on analogical reasoning and creativity, specific evidence has been
presented that participants’ capacity to map semantically distant analogies can be augmented
within a short time-frame by the use of explicit creativity prompts (e.g., Green et al., 2012b;
Weinberger et al., 2016). What this means is that, when participants are cued to “think
CREATIVE ANALOGIES 6
creatively” while reasoning with analogies, they map more and more semantically distant
analogies, improving their analogical performance. And, at least within the samples of
participants previously used in this line of research, creativity prompts did not increase the
erroneous mapping of non-valid analogies (i.e., false alarms; Green et al., 2017), but appeared to
specifically augment valid analogical reasoning.
However, up to this point in the literature, no sample of creative experts has been utilized
in testing the state-augmentation of creative analogical reasoning, and therefore open questions
remain about how creative expertise may moderate this previously observed effect. It is well
documented that expertise has a strong effect on cognitive processing within the domain of
expertise as well as potentially on domain-general cognitive processes that play a role in expert
thinking (Ericsson et al., 2018). Therefore, we might predict that individuals who have
developed expertise in creativity may more readily map semantically distant analogies than non-
experts, and experts may also be influenced by a creativity prompt in a different way than are
non-experts.
To Map or Not to Map: Analogies and Incompatibilities
When an individual is thinking creatively, they may attend solely or mostly to valid
instances of relational similarity (i.e., analogies), or they may also specifically focus on instances
of relational incompatibility, where analogical relations cannot be mapped. For example,
Simonton (1999; 2015) has described a process akin to natural selection by which creative
thinkers parse their potential ideas, and Alexander (2012; 2016) has coined the term antinomous
reasoning to describe the process by which individuals identify relations of incompatibility
among ideas: an ability that was shown to be predictive of creative and educational outcomes,
especially in engineers (Dumas et al., 2016).
CREATIVE ANALOGIES 7
As a possible counter-argument, creative experts may identify analogical relations across
contexts in a way that other individuals do not see. This argument would shift the
conceptualization of analogical validity from one of true or not-true analogical relations to one
where the validity of analogical relations is socially determined. Following this line of thought,
creative individuals may work, in part, by identifying analogs that are currently not considered
valid, but can be meaningfully connected in a novel and useful way. Sternberg and Lubart’s
(1991) investment model of creativity, in which creative thinkers “buy low and sell high” by
identifying ideas that are not highly valued now, but may be valued in the future, would seem to
support this argument. In addition, Okada and colleague’s (2009) retrospective interview
research with visual artists uncovered a similar pattern, in which creative thinking involved the
constant re-formulation of what constituted socially-valid analogical relations.
Goals and Expectations of Current Study
The current study was conducted in the context of some literature-based expectations and
hypotheses related to the effect of creative expertise on analogical reasoning. For example, based
on literature just reviewed, we hypothesized that creative experts would display some observable
heightening of their analogical reasoning ability. However, how that facility with analogical
reasoning may manifest itself was the basis of various counter-hypotheses. For instance, whether
creative experts would be affected by a state-augmentation condition to a greater or lesser extent
than non-experts was difficult to specifically hypothesize. On the one hand, creative experts
might display heighted analogical ability across all conditions and therefore the state
augmentation may have little effect. On the other hand, creative experts may be more flexible in
response to creativity-related cues, and in that case the state-augmentation prompt may have a
larger effect on them. Moreover, it was not known how creative experts may differ from non-
CREATIVE ANALOGIES 8
experts in their mapping of technically non-valid analogies: it might be hypothesized that
creative thinkers would be better able to weed-out non-valid analogies and not map them, or the
inverse could be true, that creative individuals may be capable of mapping analogies across
greater semantic distance, even beyond what might be considered valid by the creators of an
analogical reasoning task. With these goals and expectations in mind, we now introduce the
methodology of this study.
Methodology
Participants
A total of 296 individuals (62.9% female) aged from 16 to 68 (M = 30.81, SD = 11.54).
Sufficient English language proficiency was essential to the verbal analogy tasks utilized, and all
participants reported themselves fluent in English. In terms of race/ethnicity, most participants (n
= 224, 75.7%) reported their ethnicity as White/European-American, while smaller proportions
of the sample reported their ethnicity as Black/African-American (n = 16, 5.4%), Asian/Pacific
Islander (n = 14, 4.7%), Latinx (n = 15, 5.1%) or multiple ethnicities/other (n = 24, 8.1%).
The participants were sampled from three different populations with balanced group
sizes: (a) n = 92 non-acting adults, (b) n = 100 undergraduate acting majors, and (c) n = 104
professional actors. Non-acting adults were recruited online via Amazon Mechanical Turk, a
crowdsourcing platform widely used in psychological research (McKay, Karwowski &
Kaufman, 2017), and compensated $3.00 for participation. Amazon Mechanical Turk has been
shown to produce data that is similar in quality to more conventional data collection approaches
(e.g., undergraduate student participation pool; Buhrmester, Kwang, & Gosling, 2016). In this
data collection, we assumed that the MTurk participants were not experts in acting, acting
students, or experts in another creative discipline: this assumption follows in line with the bulk of
CREATIVE ANALOGIES 9
previous work using MTurk, in which MTurk participants are perceived as not likely to hold
demonstrated expertise in any particular domain (see Buhrmester et al., 2016, for a comparison
between MTurk participants and undergraduate psychology students).
Recruitment for undergraduate acting majors was accomplished through existing social
media listservs that connected undergraduate theater and acting students, and they were
compensated $10 for participation. All participating undergraduate students were currently
majoring in theater, although we allowed for diversity in terms of their specific concentrations
within that setting: of the 78 student actors who reported their concentrations 35 (44.87) were in
an acting concentration, 27 (34.61%) were in musical theater, and 16 (20.51%) were in a
directing, playwriting, or production concentration. In addition, the sample was relatively evenly
split among students in terms of the years they had spent in their undergraduate program: of the
87 student actors who reported their year-in-program, 24 (27.59%) were in their first year, 15
(17.24%) were in their second year, 18 (20.69%) were in their third year, 19 (21.84%) were in
their fourth year, and 11 (12.64) were in their fifth year as undergraduates.
Professional actors were recruited via listservs that connected members of two
professional actors’ labor unions: either Actors Equity (which focuses on the representation of
stage actors) or SAG-AFTRA (which focuses on the representation of screen actors), and they
were compensated $20 for their participation. To justify their designation as professional, the
professional actors needed to report either being members of one or both of these acting labor
unions: which exist to protect actors’ livelihood in terms of salary, benefits, and working
conditions. Because of this, union affiliated actors tend to be the only actors in the United States
who are able to make a living wage on acting work and therefore focus on acting principally
rather than other employment. If participants were not a member of either of these unions, we
CREATIVE ANALOGIES 10
also identified individuals as expert actors if they had previously booked or produced 10
previous paid theater contracts. Actors who did not report these attributes were not retained in
the sample of professionals. In addition, professional actors reported their past formal training in
acting at a university setting, with 56.38% (n = 53) reported holding a Bachelor’s degree in
acting, while 29.79% (n = 28) reported holding a Master’s degree, and 3.19% (n = 3) reported
holding a doctoral degree. 5 participants (5.32%) reported having no university training, and the
same proportion reported having a non-degree certification from a university. 72.34% (n = 68)
also reported having engaged in additional acting training at a studio apart from their university
training.
Please see Table 2 for specific demographic information within each of the three samples
of participants. Although sample sizes were balanced among the groups, the distributions of
participants in the three groups were significantly related to ethnicity [χ2 (8) = 18.99, p =.015],
gender [χ2 (4) = 24.62, p <.001], and age [F (2, 293) = 108.26, p < .001]. Therefore, the
influences of ethnicity, gender, and age on outcome measures were considered in the following
analyses.
Measure and Scoring Procedures
In this study, the Analogy Finding Task that consisted of two matrices of word-pairs was
utilized (see Weinberger, Iyer, & Green, 2016 for full measure development details). Each
matrix contained 25 word-pairs: 5 stem pairs on the left side and 20 completion pairs across the
top. Participants were instructed to identify analogies by combining stem pairs with completion
pairs, and they had five minutes to complete each matrix of word-pairs. Following the standard
and validated scoring procedures for this task, each stem pair could be combined with 3 or 4
valid completion pairs, and a maximum of 17 valid analogies (i.e., combinations of stem and
CREATIVE ANALOGIES 11
completion pairs) could be identified in each matrix. These analogies varied on how semantically
distant the word pairs were from one another (semantic distance calculated via Latent Semantic
Distance using the Touchstone Applied Sciences corpus; Dumais, 2003; http://lsa.colorado.edu/),
and the scoring procedure weighted each correct analogy that participants identify by the
semantic distance associated with that analogy. For example, the analogical pair
Watermelon:Rind::Orange:Peel is highly semantically similar, and was therefore given a lower
weight (x75 based on the scoring guide for this task) than the much more semantically distant but
also valid Watermelon:Rind::Cigarette:Butt (weighted at x269 according to the scoring guide).
In this way, participants received more “points” for mapping analogies that were valid but also
semantically distant in this task. As another example, the stem pair Kitten:Cat was validly
mappable to Puppy:Dog (weighted at x61 based on semantic distance), and also validly
mappable to the more semantically distant Seed:Tree (weighted at x79) and the still more distant
Spark:Fire (weighted at x93).
In the task directions, participants were informed that one stem pair could be combined
with multiple completion pairs, but they did not know the exact number of correct completion
pairs for any stem pair. For the first matrix, participants were given specific task directions to
identify all valid analogical combinations they could, but no prompts or cues related to their
cognitive processes. For the second matrix, the cue “please think creatively” was added to the
instructions. In previous work related to this Analogy Finding Task (i.e., Weinberger et al.,
2016), this creativity cue significantly increased the number of correct analogical pairs identified
(i.e., the sensitivity of analogical reasoning) for an Mturk sample. It should be noted that the two
matrices that make up the Analogy Finding Task were designed to be balanced in terms of the
semantic distance among the valid analogies available in each matrix, with the first matrix
CREATIVE ANALOGIES 12
having 1299 total units of semantic distance on its valid analogies, and the second matrix having
1328 total units of semantic distance available. Because the two matrices are close-to but not
perfectly balanced, and more generally do contain different word pairings, some past work has
counterbalanced the order of presentation of these matrices to participants. However, in the
current study, this counterbalancing was not done. For this reason—as is further described in the
measurement sections below—all the outcome measures related to semantic distance (i.e.,
Sensitivity and Selectivity) were divided by the total available semantic distance in that matrix,
so the proportion of the available semantic distance relevant to the outcome became the score for
analysis. Although it is important to note that the non-counterbalanced design is a limitation of
this study, the proportional scoring technique we used likely mitigated some of the issues
associated with that limitation.
As with previous work with this measure, all data were collected via the Internet and
Qualtrics Survey Software. However, unlike previous work with the Analogy Finding task, we
scored both matrices of the task using three different scoring procedures in order to capture three
theoretically differing aspects of participants’ analogical reasoning: (a) Quantity, (b) Sensitivity
and (c) Selectivity. Because this task allowed for participants to identify analogies that were
considered valid as well as non-valid, we conceptualized participant responses to the Analogy
Finding task as a confusion matrix. This confusion matrix cross-tabulated the analogies that were
considered valid or non-valid by researchers who created the administered analogical reasoning
stimuli, and the analogies that were identified as valid and non-valid by study participants, in
order to allow for the quantification of the three outcome variables being investigated in this
study. See Table 1 for a confusion-matrix based delineation of how each of these scores were
operationalized, and details related to the scoring are presented below.
CREATIVE ANALOGIES 13
Quantity of Possible Analogical Relations
The responses were first scored by the total number of completion pairs (both valid and
non-valid analogies) that participants selected to combine with stem pairs in each matrix. As
previously reviewed, this scoring procedure aimed to measure how “wide a net” participants cast
in order to catch the valid analogies. The five stem pairs in each matrix were conceptualized as
five items, and participants selected different numbers of completion pairs for each item. The
Cronbach’s α coefficients were .89 for the five items in the first matrix and .84 for the five items
in the second matrix, which indicated these item-level counts were sufficiently reliable to sum
across the 5 items, producing total quantity scores for each matrix for each participant.
Sensitivity to Analogical Mappings
Considering only the valid analogies identified by participants, Sensitivity was quantified
by calculating the proportion of available valid analogical mappings correctly identified by each
participant for each stem pair, weighting those correct identifications by their semantic distance,
and then dividing by the total semantic distance of all valid analogies available for that stem pair.
Therefore, this score can be conceptualized as the proportion of total validly analogically
mappable semantic distance mapped by each participant for each stem pair, or a rate of true
positive identification of valid analogies. With 5 stem pairs per matrix, these Sensitivity scores
showed adequate reliability (α =.81 in both matrices) and were averaged across the five stem
pairs, producing one Sensitivity score per matrix for each participant.
Selectivity of Analogical Reasoning
Then, considering only non-valid analogies in the Finding Task, Selectivity was
calculated by taking the proportion of non-valid analogies that participants correctly identified as
non-valid (i.e., did not select as valid) to the total number of non-valid analogies present for that
CREATIVE ANALOGIES 14
stem pair. In this way, Selectivity can be conceptualized as participants’ rate of true negative
identification of non-valid analogies. Across the five stem pairs in each matrix, reliability was
sufficient (α =.89 for matrix 1 and α =.80 for matrix 2) for averaging item-level Selectivity
proportions, producing two Selectivity scores for each participant: one for each matrix of the
Analogy Finding Task.
Analysis Plan
To examine the effects of creative expertise group (i.e., professional actors, student
actors, non-actors), as well as the creativity prompt in the Analogy Finding Task on the three
outcome variables (i.e., Quantity, Sensitivity, and Selectivity) we conducted analyses of variance
with a mixed design: the creativity prompt was a within-subject factor and creative expertise
group was a between-subject factor. Three outcome measures were involved in this study, and
multiple univariate F tests may potentially inflate the operational alpha level (Gamst, Meyers, &
Guarino, 2008). Therefore, a multivariate analysis of variance (MANOVA) was performed to
reduce the type I error rate as well as account for intercorrelations among the outcome measures.
Based on the results of that MANOVA, univariate ANOVAs and post-hoc tests were also run in
order to investigate the effects of creative expertise group and the creativity prompt on each of
the three analogical reasoning outcome variables included in this study.
Results
Omnibus Multivariate Patterns
The analysis began with an omnibus examination of the effect of creative expertise group
and the creativity prompt on all three outcome variables, within a multivariate space. Ethnicity,
gender, and age of participants were included as covariates. Notably, the covariance matrices of
the outcome measures were not equal across all levels of the independent variables [Box’s M =
CREATIVE ANALOGIES 15
176.18, F (42, 244948) = 4.07, p < .001], and therefore Pillai’s Trace was examined to determine
the effects of the creativity prompt and creative expertise grouping in the MANOVA (Hahs-
Vaughn, 2017). There was a significant multivariate main effect of creative expertise group
[Pillai's Trace = .14, F (6, 564) = 6.98, p < .001, partial η2 = .07], and the interaction effect
between creative prompt and expertise group was also significant [Pillai's Trace = .07, F (6, 564)
= 3.40, p = .003, partial η2 = .04]. Taken together, these multivariate results indicated that the
creative expertise groups differed significantly on the multivariate linear combination of the
three analogical reasoning outcomes depending on whether they received the creativity prompt.
No significant main effect or interaction effect was detected on any of the demographic
covariates, and therefore they were not included in the later analyses.
Outcome-specific Univariate Examinations
To better understand the omnibus multivariate result, univariate repeated ANOVAs and
post-hoc tests were also checked for each of the three outcome measures included in this
investigation. Here, we present the specific univariate results for each of the three outcome
variables included in this investigation: (a) quantity, (b) sensitivity, (c) selectivity. Descriptive
statistics associated with these results are presented in Table 3, ANOVA coefficients are in Table
4, and the quantitative patterns for each outcome variable are illustrated in Figure 1.
Quantity of Possible Analogical Relations
As can be seen in Table 3, participants selected 14.04 (SE = .48) responses on average in
the first matrix. After receiving the creativity prompt, they selected more responses (M = 17.24,
SE = .53), indicating an overall increase in the Quantity of possible analogical relations in the
“think creatively” condition. The main effect of both the creative expertise groups and the
creativity prompt was statistically significant for the Quantity outcome variable, as was the
CREATIVE ANALOGIES 16
interaction effect between these independent variables. In Figure 1a it can be observed that,
although all groups showed an increasing trend in Quantity from the first to the second matrix of
the Analogy Finding Task, the magnitude of the increase in the non-acting group was very small,
and the increase in the two acting groups (i.e. professional and students) was much larger.
Statistically, the improvement of the outcome measure was larger for the two acting groups
[student actors: F (1, 99) = 42.75, p < .001, partial η2 = .30; professional actors: F (1, 103) =
43.00, p < .001, partial η2 = .30] than the non-acting group [F (1, 91) = .01, p = .916, partial η2 =
.001], implying that creative expertise may have facilitated the observed increase in analogical
quantity in response to the creativity prompt.
Sensitivity to Analogical Mappings
Table 3 also displays the proportions of available analogically valid semantic distance the
creative expertise groups successfully mapped before and after receiving the creativity prompt. A
significant main effect of creative expertise was observed on Sensitivity (see Table 4 for
ANOVA coefficients), but the main effect of the creativity prompt on Sensitivity was non-
significant. In response to the creativity prompt, Sensitivity improved a small (and statistically
non-significant) amount in the group of professional actors (Mmatrix1 = .58; Mmatrix2 = .60),
remained exactly the in the group of student actors (Mmatrix1 = .53; Mmatrix2 = .53), and decreased a
small and statistically non-significant amount in the non-acting group (Mmatrix1 = .41; Mmatrix2
= .38). See Figure 1b for a depiction of these patterns. Given that the interaction effect was also
not significant, we examined the effects of creative expertise group more closely by conducting
post-hoc tests. Specifically, over both matrices, levels of Sensitivity within the non-acting group
were significantly lower than either professional actors (Mdif = .20, p < .001, d = .94) or student
CREATIVE ANALOGIES 17
actors (Mdif = .13, p < .001, d = .68), and the difference between the two acting groups was not
significant (Mdif = .06, p =.07, d = 0.31).
Selectivity of Analogical Reasoning
Both the main effects of creative expertise group and creativity prompt, as well as the
interaction among them, were statistically significant (see Table 4). The Selectivity of all
participants decreased from the first (M = .95, SE = .01) to the second analogy finding task (M =
.91, SE = .01). As shown in Table 3, and illustrated in Figure 1c, the mean Selectivity of non-
actors was similar to the two acting groups in matrix 1. After receiving the creativity prompt, the
Selectivity of non-actors did not go down as steeply as the two actor groups, and the non-actors
ended with the highest levels of Selectivity. So, the effect of the creativity prompt was examined
specifically across the creative expertise groups: we found that the negative impact of the
creativity prompt on Selectivity was much smaller, and non-significant, for non-actors [F (1, 91)
= .26, p = .61, partial η2 = .003] than for either student actors [F (1, 99) = 65.65, p <
.001, partial η2 = .40] or professional actors [F (1, 103) = 58.36, p < .001, partial η2 = .36], for
whom the increase was larger and statistically significant.
Discussion
The link between analogical reasoning and creativity has long been theorized in the
psychological literature (Mednick, 1962; Gentner et al., 1997), and empirical evidence from the
last decade has begun to identify specific ways in which these two constructs share underlying
cognitive or neurological mechanisms (Dumas, 2018; Green et al., 2012a; Vartanian, 2012).
However, as far as we are aware, this study has been the first to examine the influence of creative
expertise on analogical reasoning within a relatively ill-structured problem space that allowed for
multiple aspects of participants’ analogical reasoning (i.e., Quantity, Sensitivity, and Selectivity)
CREATIVE ANALOGIES 18
to emerge. As such, this study forwards a number of key findings related to the influence of
creative expertise on analogical reasoning: key findings that we delineate here.
Creative Expertise was Associated with Higher Levels of Analogical Sensitivity
In this study, the acting groups (i.e. professional or undergraduate student actors)
displayed significantly more Sensitivity to analogical relations than did the non-actors. The
greater Sensitivity to analogies among the professional actors as compared to student actors did
not reach significance, in either the baseline or the “be creative” conditions. This finding appears
to converge in meaning with previous work that identified a greater capability in analogical
reasoning as associated with expert-level innovation in the visual arts (Okada et al., 2009),
medicine (Dumas et al., 2014), engineering (Chan et al., 2015) or the physical sciences (Gentner
et al., 1997), among other domains. Therefore, the findings of this study strongly suggest that
expertise in acting, as in other creatively demanding artistic or scientific domains, is linked to the
greater ability to identify and map valid analogies.
Of course, the methodology of this study was not able to disentangle whether experience
or education in acting supported the development of analogical Sensitivity, or whether those
individuals with a greater Sensitivity to analogical relations self-selected into training programs
and professional work in acting: this question would need to be addressed via a larger scale
longitudinal investigation that remains a future direction for this area of research. Future research
may be designed to answer specific questions about how variability in analogical reasoning
during schooling leads to expertise development and career selection later in the lifespan.
Actors Responded More Strongly to the Creativity Prompt
For two of the three outcome variables analyzed in this study (i.e., Quantity and
Selectivity), significant interaction effects were uncovered, indicating that creative expertise
CREATIVE ANALOGIES 19
significantly moderated the influence of the creativity prompt. This finding may provide a
window into the way in which creative expertise affects the way individuals experience and
respond to contextual cues related to creativity. In this discussion section, we further explain
both interaction effects identified in this study: first the interaction related to Quantity, and then
Selectivity.
Increase in Quantity of Analogies
In this study, non-acting adults did not significantly change in their Quantity of selected
analogies in response to the “think creatively” prompt. In contrast, both the professional and
student actors did increase significantly in Quantity after being prompted to think creatively.
This finding is related to Green and colleagues’ (2017) finding, also with the Analogy Finding
Task, that the overall number of analogical pairs increased after a creativity prompt and
transcranial stimulation, but that increase was not significant. Because Green et al.’s (2017)
study did not utilize creative experts, or those developing creative expertise through education,
as participants, their findings are fully in convergence with the findings from this study, which
also did not find a significant increase in Quantity for the non-acting group.
Decrease in Analogical Selectivity
Similarly to previous work on state-augmentation of creative thinking (e.g., Green et al.,
2012b; Weinberger et al., 2016), the current study did not find an increase in the number of non-
valid analogies selected by the non-acting group under the “think creatively” condition.
However, the inclusion of individuals who were either creatively expert (professional actors) or
developing creative expertise (acting majors) in this study allowed us to observe that, for both
the professional and student actors, Selectivity of analogical reasoning (i.e., the rate of true-
negative identification of non-valid analogies) decreased, indicating that the actors were
CREATIVE ANALOGIES 20
selecting more non-valid analogies in the “think creatively” condition than they did in the
baseline condition. Taken together with the earlier finding of increased analogical Quantity
under the “think creatively” condition for the actors, this finding suggests that, when prompted to
think creatively, the actors began to identify analogical relations among word pairs that—at least
according to the creators of the task—were not-valid. In terms of their Selectivity, both acting
groups included in this study (professionals and students) followed similar patterns in response
to the “think creatively” prompt, and the non-actors differed from them. This finding implies
that, for a domain-general task such as Analogy Finding, those with established creative
expertise (the professionals) and those who are still acclimating to creative expertise (the
students) had similar patterns of performance in response to the prompt. A domain-specific
measurement paradigm may therefore be fruitful in the future to see more marked differences
between professional and student actors.
Also in the future, a think-aloud or retrospective interview methodology would be needed
to understand the way in which the actors themselves conceptualized the analogical relations
they identified. It could have been that, when explained, the technically non-valid analogies
mapped by the actors featured a highly original connection not identified during the formulation
of the Analogy Finding Task. On the other hand, the actors may have simply been casting a
wider net in their analogical reasoning, with the understanding that, in order to encompass the
maximum amount of valid analogies, some non-valid analogies would be included: a strategy
that coincides with classic brainstorming or divergent thinking methods (Acar, Runco, & Park,
2020). For this reason, although these data may be interpreted to indicate that professional
actors’ expertise in their domain positively supported their capacity to be Sensitive to possible
analogical relations even when those relations are semantically distant, a competing (and at this
CREATIVE ANALOGIES 21
point, also supported by the data) hypothesis would be that actors simply are more loose in their
associations, perhaps because of the way they are trained to take instructions by a director. At the
current time, future research is needed to disentangle these possibilities.
Conclusion
In general, the findings from this investigation suggest a potential mechanism in which
creative expertise moderates individuals’ experience of contextual cues related to creative work:
those individuals who are creatively expert, or developing competence in a creative domain, may
be more observant to contextual cues that support creative or divergent thinking, allowing them
to change their cognitive strategies to alter or augment their creative thinking more readily,
depending on context. To our knowledge, this type of pattern with experts in a creative discipline
has not been documented before in the psychological literature, and therefore we see this
inference: that creative experts (i.e., professional actors) and those acclimating to creative
expertise (i.e., undergraduate acting majors) may be more affected by a prompt to think
creatively than other individuals as a step towards a psychological understanding of what it may
mean for a person to develop expertise in creative thinking. In order to incorporate this finding
into the nascent conceptualization of creative expertise we presented in this article, we offer this
contextual-flexibility hypothesis of creative expertise as a possible focus for further research on
creativity, as well as the underlying role of relational reasoning in that critically important human
capacity.
CREATIVE ANALOGIES 22
References
Acar, S., Runco, M. A., & Park, H. (2020). What should people be told when they take a divergent
thinking test? A meta-analytic review of explicit instructions for divergent thinking. Psychology
of Aesthetics, Creativity, and the Arts, 14(1), 39–49. https://doi.org/10.1037/aca0000256
Alexander, P. A. (2003). The development of expertise: The journey from acclimation to proficiency.
Educational Researcher, 32(8), 10–14. https://doi.org/10.3102/0013189X032008010
Alexander, P. (2012). Reading into the future: Competence for the 21st century. Educational
Psychologist, 47(4), 259–280. https://doi.org/10.1080/00461520.2012.722511
Alexander, P. A., Dumas, D., Grossnickle, E. M., List, A., & Firetto, C. M. (2016). Measuring
relational reasoning. Journal of Experimental Education, 84(1), 119–151.
https://doi.org/10.1080/00220973.2014.963216
Buhrmester, M., Kwang, T., & Gosling, S. D. (2016). Amazon’s Mechanical Turk: A new source of
inexpensive, yet high-quality data? In A. E. Kazdin (Ed.), Methodological issues and
strategies in clinical research (p. 133–139). American Psychological
Association. https://doi.org/10.1037/14805-009
Chan, J., & Schunn, C. (2015). The impact of analogies on creative concept generation: Lessons from
an in vivo study in engineering design. Cognitive Science, 39(1), 126–155.
https://doi.org/10.1111/cogs.12127
Dumais, S.T., (2003). Latent Semantic Analysis. Annual Review of Information Science and
Technology, 38(1):188-230. DOI:10.1002/aris.1440380105
Dumas, D. (2018). Relational reasoning and divergent thinking: An examination of the threshold
hypothesis with quantile regression. Contemporary Educational Psychology, 53, 1–14.
https://doi.org/10.1016/j.cedpsych.2018.01.003
CREATIVE ANALOGIES 23
Dumas, D., Alexander, P. A., & Grossnickle, E. M. (2013). Relational reasoning and its
manifestations in the educational context: A systematic review of the literature. Educational
Psychology Review, 25(3), 391–427. https://doi.org/10.1007/s10648-013-9224-4
Dumas, D., Alexander, P. A., Baker, L. M., Jablansky, S., & Dunbar, K. N. (2014). Relational
reasoning in medical education: Patterns in discourse and diagnosis. Journal of Educational
Psychology, 106(4), 1021–1035. https://doi.org/10.1037/a0036777
Dumas, D., Doherty, M., Organisciak, P. (2020). The psychology of professional and student actors:
Creativity, personality, and motivation. PLoS: ONE.
Dumas, D., Schmidt, L. C., & Alexander, P. A. (2016). Predicting creative problem solving in
engineering design. Thinking Skills and Creativity, 21, 50–66.
https://doi.org/10.1016/j.tsc.2016.05.002
Ericsson, K. A., Hoffman, R. R., Kozbelt, A., & Williams, A. M. (2018). The Cambridge handbook of
expertise and expert performance, 2nd ed (K. A. Ericsson, R. R. Hoffman, A. Kozbelt, & A. M.
Williams, Eds.; 2018-27271-000). Cambridge University Press.
https://doi.org/10.1017/9781316480748
Ericsson, K. A., & Smith, J. (1991). Toward a General Theory of Expertise: Prospects and Limits.
Cambridge University Press.
Fox, E. (2009). The role of reader characteristics in processing and learning from informational text.
Review of Educational Research, 79(1), 197–261. https://doi.org/10.3102/0034654308324654
Gamst, G., Meyers, S., & Guarino, A.J. (2008). Analysis of variance designs. NY: Cambridge
University Press.
Goel, A. K. (1997). Design, analogy, and creativity. IEEE Expert, 12(3), 62–70.
https://doi.org/10.1109/64.590078
CREATIVE ANALOGIES 24
Goswami, U. (2013). The development of reasoning by analogy. In P. Barrouillet & C. Gauffroy
(Eds.), The development of thinking and reasoning. (2013-15453-004; pp. 49–70). Psychology
Press.
Green, A. E., Kraemer, D. J. M., Fugelsang, J. A., Gray, J. R., & Dunbar, K. N. (2012a). Neural
correlates of creativity in analogical reasoning. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 38(2), 264–272. https://doi.org/10.1037/a0025764
Green, A. E., Cohen, M. S., Kim, J. U., & Gray, J. R. (2012b). An explicit cue improves creative
analogical reasoning. Intelligence, 40(6), 598–603. https://doi.org/10.1016/j.intell.2012.08.005
Green, A. (2018). Creativity in the distance: The neurocognition of semantically distant relational
thinking and reasoning. In R. E. Jung & O. Vartanian (Eds.), The Cambridge handbook of the
neuroscience of creativity. (2018-03511-020; pp. 363–381). Cambridge University Press.
https://doi.org/10.1017/9781316556238.021
Green, A. E., Spiegel, K. A., Giangrande, E. J., Weinberger, A. B., Gallagher, N. M., & Turkeltaub,
P. E. (2017). Thinking cap plus thinking zap: TDCS of frontopolar cortex improves creative
analogical reasoning and facilitates conscious augmentation of state creativity in verb generation.
Cerebral Cortex, 27(4), 2628–2629.
Hahs-Vaughn, D. (2017). Applied multivariate statistical concepts. NY: Routledge.
Kogan, N. (2002). Careers in the performing arts: A psychological perspective. Creativity Research
Journal, 14(1), 1–16. https://doi.org/10.1207/S15326934CRJ1401_1
McKay, A. S., Karwowski, M., & Kaufman, J. C. (2017). Measuring the muses: Validating the
Kaufman Domains of Creativity Scale (K-DOCS). Psychology of Aesthetics, Creativity, and
the Arts, 11(2), 216–230. https://doi.org/10.1037/aca0000074
CREATIVE ANALOGIES 25
Mednick, S. (1962). The associative basis of the creative process. Psychological Review, 69(3), 220–
232. https://doi.org/10.1037/h0048850
Murphy, P. K., Firetto, C. M., & Greene, J. A. (2017). Enriching students’ scientific thinking through
relational reasoning: Seeking evidence in texts, tasks, and talk. Educational Psychology Review,
29(1), 105–117. https://doi.org/10.1007/s10648-016-9387-x
Noice, H., & Noice, T. (1994). An example of role preparation by a professional actor: A think-aloud
protocol. Discourse Processes, 18(3), 345–369. https://doi.org/10.1080/01638539409544899
Okada, T., Yokochi, S., Ishibashi, K., & Ueda, K. (2009). Analogical modification in the creation of
contemporary art. Cognitive Systems Research, 10(3), 189–203.
https://doi.org/10.1016/j.cogsys.2008.09.007
Richland, L. E., & Simms, N. (2015). Analogy, higher order thinking, and education. WIREs
Cognitive Science, 6(2), 177–192. https://doi.org/10.1002/wcs.1336
Runco, M. A., & Jaeger, G. J. (2012). The standard definition of creativity. Creativity Research
Journal, 24(1), 92–96. https://doi.org/10.1080/10400419.2012.650092
Schunn, C. D., & Anderson, J. R. (1999). The generality/specificity of expertise in scientific
reasoning. Cognitive Science, 23(3), 337–370. https://doi.org/10.1207/s15516709cog2303_3
Simonton, D. K. (1999). Creativity as blind variation and selective retention: Is the creative process
Darwinian? Psychological Inquiry, 10(4), 309–328. JSTOR.
Simonton, D. K. (2015). On praising convergent thinking: Creativity as blind variation and selective
retention. Creativity Research Journal, 27(3), 262–270.
https://doi.org/10.1080/10400419.2015.1063877
Stacey, C. L., & Goldberg, H. D. (1953). A personality study of professional and student actors.
Journal of Applied Psychology, 37(1), 24–25. https://doi.org/10.1037/h0058197
CREATIVE ANALOGIES 26
Sternberg, R. J., & Lubart, T. I. (1991). An investment theory of creativity and its development.
Human Development, 34(1), 1–31. https://doi.org/10.1159/000277029
Wagner, R. K., & Stanovich, K. E. (1996). Expertise in reading. In K. A. Ericsson (Ed.) The Road to
Excellence: The Acquisition of Expert Performance in the Arts and Sciences, Sports, and Games,
189–225.
Walker, C. M., Hubachek, S. Q., & Vendetti, M. S. (2018). Achieving abstraction: Generating far
analogies promotes relational reasoning in children. Developmental Psychology, 54(10), 1833–
1841. https://doi.org/10.1037/dev0000581
Weinberger, A. B., Iyer, H., & Green, A. E. (2016). Conscious augmentation of creative state
enhances “real” creativity in open-ended analogical reasoning. PLoS ONE, 11(3).
doi:10.1371/journal.pone.0150773
Van Straaten, D., Wilschut, A., Oostdam, R., & Fukkink, R. (2019). Fostering students’ appraisals of
the relevance of history by comparing analogous cases of an enduring human issue: A quasi-
experimental study. Cognition and Instruction. https://doi.org/10.1080/07370008.2019.1614590
Vartanian, O. (2012). Dissociable neural systems for analogy and metaphor: Implications for the
neuroscience of creativity. British Journal of Psychology, 103(3), 302–316.
https://doi.org/10.1111/j.2044-8295.2011.02073.x
CREATIVE ANALOGIES 27
Table 1.
Example Confusion Matrix for Analogical Reasoning
Participant
Identification
Researcher Designation
Valid
Non-Valid
Valid
True positive
False positive
Non-Valid
False negative
True negative
Note: Study outcome variables are operationalized in the following ways:
Quantity is the total number of analogies selected across both valid and non-
valid conditions; Sensitivity is the proportion of true positive identification;
and Selectivity is the proportion of true negative identification.
CREATIVE ANALOGIES 28
Table 2.
Sample Distributions: Ethnicity, Gender and Age across Creative Expertise Groups
Creative expertise groups
Non-acting
adults
Professional
actors
Student
actors
N by ethnicity
Caucasian/non-Hispanic
white
63
84
77
African American
7
3
6
Asian/pacific islanders
11
1
2
Hispanic
5
6
4
Multiple race/other
6
8
10
N by gender
Male
38
47
18
Female
54
54
77
Non-binary/third gender
0
1
5
Age Means (Std.)
36.98 (11.16)
35.43 (10.14)
20.33 (2.65)
CREATIVE ANALOGIES 29
Table 3.
Descriptive Statistics of Outcome Measures
Matrix 1
Matrix 2
Mean
S.E.
Mean
S.E.
Quantity of Possible Analogical Relations
All participants
14.04
.48
17.24
.53
Non-acting adults
12.17
1.12
12.29
.73
Professional actors
14.75
.64
19.40
.91
Student actors
15.03
.67
19.55
.92
Sensitivity to Analogical Mappings
All participants
.51
.01
.51
.01
Non-acting adults
.41
.02
.38
.02
Professional actors
.58
.02
.60
.02
Student actors
.53
.02
.53
.02
Selectivity of Analogical Reasoning
All participants
.95
.01
.91
.01
Non-acting adults
.95
.01
.94
.01
Professional actors
.95
.01
.90
.01
Student actors
.94
.01
.88
.01
CREATIVE ANALOGIES 30
Table 4.
ANOVA Coefficients for Three Outcome Measures on Creativity prompt and Creative Expertise
Groups
Sum of
Squares
df
Mean
Square
F p η2
Quantity of Possible Analogical Relations
Expertise group
3108.89
2
1554.45
14.91
<.001
.09
Creativity prompt
1416.56
1
1416.56
39.66
<.001
.12
Interaction effect
633.53
2
316.76
8.87
<.001
.06
Sensitivity to Analogical Mappings
Expertise group
3.88
2
1.94
24.09
<.001
.14
Creativity prompt
.01
1
.01
.27
.60
.001
Interaction effect
.07
2
.03
2.13
.12
.01
Selectivity of Analogical Reasoning
Expertise group
.12
2
.06
6.67
.001
.04
Creativity prompt
.22
1
.22
57.26
<.001
.16
Interaction effect
.07
2
.04
9.72
<.001
.06
CREATIVE ANALOGIES 31
(a) Quantity of Possible Analogical Relations
(b) Sensitivity to Analogical Mappings
10.00
12.00
14.00
16.00
18.00
20.00
22.00
Baseline (Matrix 1) Be Creative Prompt (Matrix 2)
Quantities
of possible analogical relations (counts)
Non-acting adults Professional actors Student Actors
.20
.25
.30
.35
.40
.45
.50
.55
.60
.65
Baseline (Matrix 1) Be Creative Prompt (Matrix 2)
Sensitivity levels to analogical mappings (proportions)
Non-acting adults Professional actors Student Actors
CREATIVE ANALOGIES 32
(c) Selectivity of Analogical Reasoning
Figure 1. Plots for means of outcome measures for the three creative expertise groups between
both matrices of the Analogy Finding Task. The second matrix included the specific prompt for
participants to think creatively. Error bars indicate standard errors of means.
.85
.87
.89
.91
.93
.95
.97
Baseline (Matrix 1) Be Creative Prompt (Matrix 2)
Selectivity levels of analogical reasoning (proportions)
Non-acting adults Professional actors Student Actors
