Effects of association interventions on students’ creative thinking, aptitude, empathy, and design scheme in a STEAM course: considering remote and close association

Abstract

One of the primary goals of STEAM education is to equip students with the capability of creativity to solve problems. Creativity is believed to be closely related to the ability of making remote associations and combining unrelated concepts. This article explored the effects of three kinds of association interventions (i.e., remote association, close association, free association) on students’ creative thinking, creative aptitude, empathy, and design scheme. A total of 94 middle school students participated in the study and were assigned to three groups: the experiment group 1 (n = 30) used remote association and experiment group 2 (n = 32) used close association, the control group (n = 32) used free association (brainstorming) to complete the STEAM integration design projects respectively. Creative Thinking Test, Williams Creativity Aptitude Test (WCAT), Basic Empathy Scale (BES), design ideas, and interviews of students become source data for analysis. Results indicated that both remote and close association were effective strategies in promoting creativity in the STEAM course. However, students in the remote association group achieved a significantly higher degree of creative thinking. While students in the close association group significantly outperformed the remote association group on creativity aptitude and quality of design ideas. No significant difference was found among the three association conditions in students’ degree of empathy. The findings highlight the different effects of remote and close association for creativity cultivation in STEAM education.

Introduction

As one of the official 21st century skills, creative thinking is central to the arts, sciences, and daily life (Beaty, Benedek, Silvia & Schacter, 2016), and it has aroused tremendous attention in recent years (Lin et al., 2022). Taking the form of originality and usefulness (Mednick, 1962), creativity is the fountainhead of human civilizations, as all progress and innovation depend on our ability to change existing thinking patterns, break with the present, and build something new (Dietrich & Kanso, 2010; Zhan et al., 2022). Kaufman (2016) also claimed that people who value creativity may point to its role in most things that define our civilization.

In the field of psychology, associative ability has an important impact on creativity (Mednick, 1962). Association intervention can be an effective method to foster creative idea generation (Acar, Runco & Park, 2020; Benedek et al., 2020; Green et al., 2015; Prabhakaran et al., 2014; Said-Metwaly et al., 2020). The association stimulus that is not closely related to the target object may be an effective enabler of successful creative idea generation as it contributes to promoting uncommon ideas by reducing the tendency to produce high-frequency ideas (Gupta et al., 2012).

However, although great importance nowadays has been attached to fostering creative thinkers through STEAM education, there has been limited research on how to inspire and evaluate students’ creative thinking during the STEAM learning process. Studies generally emphasized that cross-disciplinary, or interdisciplinary/multidisciplinary education, is of great value in promoting the cultivation of students’ creativity ability (Bevan, 2017; Harris & De Bruin, 2018; Li & Zhan, 2022; Tan, 2014; Wilson et al., 2021), there is only little evidence based on empirical studies to confirm the practical effects (Conradty & Bogner, 2019; Perignat & Katz-Buonincontro, 2019; Shen et al., 2021; Timotheou & Ioannou, 2021; Wilson et al., 2021). In addition, it tends to focus on the hands-on works or tools application but lacks attention to students’ thinking process and creative idea generation (Blikstein, 2013; Conradty & Bogner, 2019). Without targeted creative thinking training and guidance, it is difficult for students to generate innovative ideas and put them into practice (Wu et al., 2022). The methods for inspiring and evaluating students’ creativity in STEAM courses are in urgent need of development. Empirical exploration is also needed to explore the effect of association intervention targeting at creative idea generation.

Literature Review

Creativity and association

Creativity is frequently associated with originality, novelty, invention, appropriateness, imagination, etc. (Kaufman, 2016; Runco, 2014). According to existing research, there’s no universally agreed-upon definition for creativity. Representative researchers have studied the core concepts related to creativity, which focused mainly on two key determinants: originality and usefulness (Kaufman, 2016; Runco & Jaeger, 2012). In other words, the uniqueness and value orientation of creativity was emphasized respectively (Kaufman 2016; Simonton, 2012). For example, creative products must be both novel and appropriate (or useful), and not simply random responses (Prabhakaran, Green, & Gray, 2014). Except for originality and usefulness, ‘high quality’ was also suggested to be a third component of creativity(Sternberg et al., 2002).

The essence of creative thinking was claimed to be closely related to association, which is a mental process that occurs when imagination is triggered by an indirectly-related target. Correspondingly, association intervention is a pedagogical approach adopting associations to exert influence on students’ mental association activity and imagination process in a planned way step by step towards the expected goal.

According to the approach of launching associations, the association interventions can be classified into three categories (i.e., remote association, close association, and free association). Remote Association Intervention refers to the approach of using far-distance or irrelevant concepts as stimuli to trigger meaningful associations. Correspondingly, Close Association Intervention refers to the approach of using closely-related concepts as association stimuli, which can be an object that is obviously related to the target object or has similar or extended attributes for transfer. Both remote and close associations belong to compulsory associations that require students to think with a certain stimulus, and they are set up according to the semantic correlation distance between the associated source and target objects (Mednick, 1962). Whereas, free association allows students to think freely, and does not require compulsory stimuli.

According to Mednick’s (1962) remote association theory, the creative thinking process was explained as the forming of associated elements into new combinations which either meet specified requirements or are in some way useful. The ability to use associations lies at the heart of creativity, and the structure and strength of participants’ associative hierarchies impact individual responses in making remote associations and fruitfully combining unrelated concepts (Benedek et al., 2020; Prabhakaran, Green, & Gray, 2014). Less creative individuals were regarded as having steep associative hierarchies in that the activation of one concept (e.g., milk) prompts the activation of mainly closely associated concepts (e.g., white, tea), while highly creative people have flat associative hierarchies also activate weakly associated concepts (e.g., exploit) (Abraham, Rutter, Bantin, & Hermann, 2018). Mednick (1962) also suggested the remote associates test (RAT) as a way of testing for individual differences in creativity, which laid the foundation for subsequent studies related to creativity.

Some studies focused on the ability of people with different levels of creativity to recognize remote associations and close associations. Gruszka & Necka (2002) carried out an empirical study on examining the acceptance of remote and close association, in which participants were supposed to decide whether pair words presented to them were either close associations or remote associations. It appeared that participants were more inclined to accept close rather than remote associations, and they also needed more time to make their decisions in the remote rather than in the close condition. Furthermore, creative people acknowledged the connection between two words more frequently, particularly in the remote condition, when remote associations were preceded by neutral primes and, to some extent, when close associations were preceded by positive primes. Wu et al., (2016) studied the Mandarin speakers’ performance between remote and close association through the Chinese Remote Associates Test (CRAT). Results indicated the correct rate of close association was significantly higher than that of the remote association, and remote and close associations work differently as patterns in the brain network. Others studied the individual difference of associative ability while completing word-generation association tasks. In the verbal fluency tasks, participants’ responses were analyzed by means of latent semantic analysis (LSA) and scored for semantic distance as a measure of associative ability, and results provided support for the notion that both associative and executive processes underlie the production of novel ideas (Beaty et al., 2014).

Some studies looked at the deliberate training of associations targeted at creative idea generation. Creativity is no longer being portrayed as a mysterious, elusive force (Runco, 2012). The generation of associations can occur either via spontaneous, free-associative or goal-directed, controlled mechanisms (Beaty et al., 2014; Benedek & Jauk, 2018; Sowden, Pringle, & Gabora, 2015). Usually, the spontaneous association occurs as an implicit, unconscious, bottom-up driven cognition during real everyday problem solving (Abraham et al., 2012). Free association reflects the basic spreading of activation in semantic networks, whereas controlled association generation reflects a goal-directed process of constrained recall considering specific search cues (Benedek et al., 2020). Supported by analogies between ideas generation and atoms combination, Martindale (2007) argued that the probability of having a creative idea was related in an inverted-U fashion to the degree that cognition was primordial in nature. Having a flat associative hierarchy, defocused attention or cognitive disinhibition allows access to more remotely associated concepts, thus it increases the likelihood of the generation of more unusual and original ideas (Abraham et al., 2018).

Appropriate association intervention may affect the stimulation of creativity. In one study, participants were instructed to say a verb upon seeing a noun displayed on a computer screen and were cued to respond creatively to half of the nouns (Prabhakaran et al., 2014). Results (from latent semantic analysis, LSA) showed that semantic distance was higher in the cued than in the uncued condition, suggesting that untrained participants were able to modulate their word production effectively, and do so on demand. Evidence of this strategy can also be gleaned from the findings of another similar fMRI study (Green et al., 2015). In the verb-generation task, participants generated verbs that were more semantically distant from noun prompts when they were cued to think creatively. The large effect size elicited by this creativity cue extended previous evidence that creativity cues can enhance creative state.

Benedek et al. (2020) designed three adjectives stimuli association tasks, requiring the generation of either (a) common associations (highly related concept to a given adjective, e.g., red: blood), (b) original associations (remotely related concept to a given adjective, e.g., red: ketchup stain), or (c) bi-associations ( two adjectives were presented and participants should find a concept that is semantically related to both cues and links them in an original way, e.g., red - round: clown nose). The concept generated needed to be a semantically related noun (either a single or multi-word term). Compared with the common association task, results showed that the original association task elicited more creative association responses. Their findings confirmed that the explicit instruction to “be creative” increases the creativity of responses.

In one meta-analytic review, Acar, Runco & Park, 2020 reviewed 31 studies that compared the explicit instructions emphasizing creativity, originality, and quality to quantity. Results indicated that creativity and quality instructions increased performance on divergent thinking when added to quantity instructions, more than quantity instructions alone, while the originality instructions did not change divergent thinking performance. Therefore, explicit instructions may increase or decrease divergent thinking performance, depending on which alternative explicit instructions were used and how they were presented. Another instruction meta-analysis also revealed that explicit (vs. standard) instructions significantly enhanced creative performance (Said-Metwaly et al., 2020).

This kind of “thinking cap” (i.e., to try and succeed at thinking more creatively) phenomenon is of broad importance for education and rich mental life (Green et al., 2015). It is also commonly experienced in some analogical reasoning activities, for example, creativity cues were used for open-ended analogical reasoning in a novel “Analogy Finding Task” and achieved positive effects (Weinberger et al., 2016). In the online test, participants were asked to seek valid analogical connections among the word-pairs matrix, assessed by the semantic distance and number of analogies identified. Findings showed that when explicitly instructed to think creatively or receiving proper creativity cues, participants could successfully elicit better creative state and perform more creative analogies.

Together, these studies provided evidence that humans have an impressive potential to improve their creative state through deliberate effort. Among these studies, associations played a significant role in triggering transfer (Mednick, 1962), analogical reasoning (Green et al., 2012), divergent thinking (Nusbaum, Silvia & Beaty, 2014), dynamic conscious augmentation of creative state (Green et al., 2015), creative cognition (Beaty et al., 2014; Weinberger et al., 2016), and the production of novel ideas and solutions (Vendetti, Wu & Holyoak, 2014; Benedek et al., 2020). However, there is limited research further examining the effect of different types of association interventions, and that is one of the goals of this study.

STEAM Education and Creativity

There is a growing body of studies that suggested the many ways in which STEAM can motivate and support learners’ purposeful activity, and facilitate their development of creative thinking (e.g., Harris & De Bruin, 2018; Runco, 2014; Wilson et al., 2021). Although creative thinking (or creativity) was mentioned frequently in articles as an outcome of STEAM, there is a lack of describing or further expansion upon the ways in which creativity is developed, practiced, or fostered through STEAM education (Perignat & Katz-Buonincontro, 2019). In fact, some articles only provided experiential or conceptual descriptions of creativity in STEAM education (e.g., Harris & De Bruin, 2018), few articles have reported creativity as a measured outcome.

For example, Wilson et al., (2021) conducted empirical comparison research in elementary and secondary schools, and found that students in the STEAM group reported a significant growth in their understanding of creative thinking, and also described more creative examples when solving problems. In the STEAM courses, Shen et al., (2021) examined the influence of teacher feedbacks on students’ creativity in 3D-printing performances, and found that brain-storming helped students to generate new ideas. According to the data collected from the Eugene Creativity Scale, the creativity score of the ‘teacher suggestions feedback group’ (teacher provided detailed instructions for students’ ideas) was significantly higher than the ‘teacher opinions feedback group’ (teacher only gave opinions feedback, such as your idea is good) and the ‘non-feedback group’ (no feedback was given to the students). Timotheou & Ioannou (2021) analyzed the video data and robot artifacts of primary school students, and their results showed that STEAM making activities can enact the development of collective creativity. A different result was also reported. Conradty & Bogner’s (2019) research showed that students’ creativity was not affected by a short-term STEAM intervention. In the four-hour science inquiry lesson, high school students participated in the workstation to explore the bird flight related scientific phenomena. Finally, results showed that the STEAM module produced long-term knowledge and built stable scores of intrinsic motivations, but not self-reported aspects of creativity.

There was an overall lack of measured learning outcomes in the areas of improved creativity (Perignat & Katz-Buonincontro, 2019), let alone specific strategies that focus on the stimulation of creative idea generation. Educators often favor well-structured lessons, recipe-style-learning laboratory work, with a controlled time frame for learning activities, yet without giving creativity space to develop (Conradty & Bogner, 2019). STEAM activities rely on kits with step-by-step instructions to constrain the creative design of works and seek multiple ways of solving problems. Besides, STEAM provides learning environments rich in tools, materials, and technologies, but an emphasis on tools causes the risks of blindly indulging in the tools and obscure fully participatory and interest-driven learning (Bevan, 2017). Some also cautioned about the danger of sheer ability to design and turn out, using tools, a vast number of ‘products’ risks (e.g., 3-D printing) becoming an end unto itself (Blikstein, 2013).

Few studies have shed light on the strategies related to the stimulation of creative ideas. Some works mentioned brainstorming in the context of creative idea generation (Zhan et al, 2022). Students were given a period of time for brainstorming and sharing ideas among group members to generate new ideas and design schemes (Khamhaengpol, Sriprom & Chuamchaitrakool, 2021; Conradty & Bogner, 2019; Shen et al., 2021). However, lacking in-depth guidance, students tend to generate ordinary ideas that lack creativity. Concrete and effective creativity training strategies are still missing.

In summary, while creative idea generation stands in the core place of innovative design or production, practices of current STEAM education tend to focus more on “doing” than “thinking”, and more on “repeating” than “creating”. There is a need to focus on the participation of mindsets in STEAM education (Martin, 2015), and extend available training strategies for creativity inspiration. Drawing upon research on association and creativity from the field of psychology, association intervention might be a potential way to foster creativity in STEAM education. Remote association and close association intervention are the most typical interventions that are used to trigger students’ knowledge transfer and creativity, thus we tried to investigate their different impacts and find out possible regulations for enhancing higher-order learning.

Research purpose and questions

The present study aimed to investigate the effects of creative thinking, aptitude, empathy, and design scheme underlying different association interventions (i.e., remote association, close association, and free association) in a STEAM course. Two research questions drove this inquiry:

RQ1

Will association intervention enhance students’ creative thinking, aptitude, empathy, and design scheme?

RQ2

Will students who participate in remote association intervention develop more positive reactions to creative thinking, aptitude, empathy, and design scheme than those who participate in close association intervention?

Method

Participants

The present study was conducted in a one-semester STEAM course of a public secondary school in Guangzhou, China. Altogether, ninety-four students aged from 13 to 16 years old participated in the study. They were randomly assigned to three conditions: remote association (class A), close association (class B), free association (class C). As Table 1 shows, class A had 30 students, class B had 32 students, Class C had 32 students. Class A and Class B consisted of the experimental groups, and Class C was the control group.

Table 1.

an overview of the research design

Class	Sample size	Duration	Project
A (remote association group)	30	8 weeks	Mask design challenge (4 weeks) 3D glasses design challenge (4 weeks)
B (close association group)	32
C (free association group)	32

The Association intervention

Students in all conditions participated in the STEAM course that included two design projects, one was to design a mask, another was to design a 3D glasses. Each design project followed four stages: Goal proposal, Creativity inspiration, Design challenge, Presentation and evaluation, as shown in Table 2. The curriculum was developed by the research team with backgrounds in STEAM education with guidance from information technology teachers from the middle school.

Table 2.

Comparison of intervention activities across conditions

Intervention activity	Remote association condition	Close association condition	Free association condition	Purpose
Goal proposal	Goal definition of the design task, construction of the association context	Goal definition of the design task, construction of the association context	Goal definition of the design task, construction of the association context	Stimulating learning interest
Creativity inspiration	Ideating by remote associative stimuli, such as apple and bulb	Ideating by close associative stimuli, such as Sichuan opera face and patch	Ideating by no associative stimulus-brainstorming	Fostering divergent thinking and convergent thinking
Design challenges	Completing the design plan	Completing the design plan	Completing the design plan	Cultivating empathy and creative thinking
Presentation and evaluation	Communicating and improving the design plan	Communicating and improving the design plan	Communicating and improving the design plan	Encouraging inter-group learning

In the stage of goal proposal, it focused on stimulating students’ interest in design and identifying the design task. By providing related stories, videos, and pictures, teachers guided students to identify the design theme, that is, to design creative masks and 3D glasses for these two projects respectively. Students needed to understand that the goal of associations was to generate creative ideas combined with the real problem situation.

Next, students applied different association methods to promote the divergence and convergence of thinking and generate creative ideas. In the remote association condition, students were provided with an apple in the mask design project and a bulb in the 3D glasses design project as association stimuli. Both were not directly related to the design targets. In contrast, in the close association condition, students used Sichuan opera face and patch as association stimulus to help generate creative ideas. The stimuli are either similar to the target objects or have some identical or obvious attributes to the target design objects. At this stage, students analyzed the relevant characteristics of the associative stimulus as comprehensively as possible, and then compulsorily associated the target with these attributes to form novel and unique creative ideas.

In the design stage, students were guided to analyze the real needs of users based on the many ideas initially formed, that is, using “empathy” to help form the final idea.

In the last presentation and evaluation stage, each team displayed and reported their design plan, and summarized the problems and experiences encountered in the design process. This stage mainly aimed to facilitate the communication and learning between groups, and made the evaluation more fair and objective.

Measure instruments

Creative thinking, aptitude, and empathy were examined in this study. Measures included the Creative thinking test, Williams creativity aptitude test (WCAT), Basic Empathy Scale (BES), as presented in Table 3. All the measure instruments were translated into Chinese and double-checked by at least two domain experts and two English experts to ensure validity and reliability.

Table 3.

Instruments for data collection

Construct	Source	Number of items	Cronbach’s alpha
Creative thinking test (CTT)	Zheng & Xiao (1983)	5	0.93
Williams creativity aptitude test (WCAT)	Williams (1993)	50	0.81
Basic Empathy Scale (BES)	Jolliffe & Farrington (2006)	20	cognitive: 0.79 Affective: 0.85

Creative thinking test (CTT) items examined fluency, flexibility, and uniqueness. The test consists of 5 items, and the Cronbach’s alpha was 0.93 (Zheng & Xiao, 1983). The test content includes a language test and a graphics test. The “Consensual Assessment Technique” (Baer & McKool, 2009) was adopted to score the test: two authors from the research group scored separately, and then the average score was taken as the test score. If there was a big difference between the scores of the two people, one more researcher will be asked to score.

Williams creativity aptitude test (WCAT) (Williams, 1993) consists of 50 items, including four aptitude elements: imagination, risk-taking, curiosity and challenge. The present study took the Chinese revised version (Lin & Wang, 1994), and the Cronbach’s alpha was 0.81.

Basic Empathy Scale (BES) (Jolliffe & Farrington, 2006) is a 20-item scale (nine cognitive items and 11 affective items) that measures affective and cognitive empathy. Each item asked the participant to respond on a Likert scale from 1 representing “strongly disagree” to 5 representing “strongly agree”, depending on the degree to which the item described them.

For the design scheme analysis, we developed the creative idea evaluation form to score students’ designs. According to Torrance (1966), creative thinking has four characteristics: fluency, uniqueness, flexibility, and precision. Besides, it is believed that people with high creativity can quickly and timely judge and grasp unique and novel concepts, which means that such people have high sensitivity. Therefore, this study took the five dimensions: fluency, flexibility, uniqueness, precision, and sensitivity, to evaluate the ideas generated by students under different association interventions.

Finally, we invited one student from each group for an interview to collect students’ opinions on the association intervention strategies adopted in this study. Students are required to express their opinions about the learning experience, gains, and suggestions of this STEAM course.

Research process

Students participated in eight 1.5-hour class periods during the study (see Table 4). In week 1, students took the pretest. Later, students participated in two design tasks followed by instruction on either remote association, close association or free association: the mask design task during week 2 to 4, and the 3D glasses design task during week 5 to 7. On week 8, students completed the posttest. Overall, the study lasted for 2 months, spread across 8 weeks.

Table 4.

Schedule of the experiment

Design task	Week	Content
Mask designing	1st	Pretest of Creative thinking, creative disposition and empathy
	2nd	Mask design: idea generation
	3rd	Mask design: design
	4th	Mask design: presentation and evaluation
3D glasses designing	5th	3D glasses design: idea generation
	6th	3D glasses design: design
	7th	3D glasses: presentation and evaluation
	8th	Posttest of Creative thinking, creative disposition, empathy, creative idea analysis and student interview

Figure 1 demonstrates the research process of the study, and Fig. 2 presents the scenes of class activities when students were doing the design projects. All students and their parents were acknowledged that they were participating in an education research and their individual data would be collected only for research purposes. We ensured all procedures performed in this study were in accordance with the ethical standards of American Psychological Association.

Fig. 1.

Research process

Fig. 2.

Class activities of the design project

Data Analysis

All data in this study were analyzed using SPSS 26 to conduct descriptive statistics. The mean differences among three association conditions were tested by one-way analysis of variance (ANOVA), and the changes between the pre- and post-tests were tested by paired samples t-tests. Since there were multiple comparisons among three different groups for every dependent variable, one-way ANOVA Post Hoc analysis was adopted based on the Test of Homogeneity of Variances. Students’ design schemes were scored on five dimensions (i.e., fluency, flexibility, uniqueness, precision, and sensitivity) by adopting the consensus assessment technique. In order to validate and deepen the understanding of the statistical results, we randomly selected eight students for interviews, and obtained their feedback from three aspects (i.e., experience, gain, and suggestion), then summed up the main points related to the association interventions.

Result

Creative thinking

For the creative thinking test, we calculated ANOVA on pre- and post-tests by condition. Results showed that a significant difference existed among the three conditions in both pretest (F = 4.888; p = .010) and posttest (F = 4.311; p = .016) (see Table 5). Then paired samples t-test was conducted to compare the change of each condition. Results show that both remote association (p = .000, d = 2.0048) and close association (p = .000, d = 1.0891) fostered students’ overall performance on the creative thinking test.

Table 5.

Means and standard deviation of students’ creative thinking test

CTT items	Remote association(N = 30)		Close association(N = 32)		Free association(N = 32)
	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)
Fluency	18.43 (4.768)	24.87 (3.213)	19.50 (4.080)	25.69 (2.235)	17.88 (4.723)	23.94 (3.121)
Flexibility	11.70 (4.252)	19.23 (3.711)	15.19 (4.518)	18.06 (2.929)	12.44 (5.913)	18.88 (5.142)
Uniqueness	6.37 (2.906)	9.43 (3.059)	7.50 (3.370)	7.88 (3.220)	4.63 (2.612)	8.13 (3.998)
Total	36.50 (9.790)	53.53 (6.962)	42.19 (10.256)	51.63 (6.714)	34.94 (12.062)	50.94 (10.439)

For Fluency, no significant difference was found among the three groups either in the pretest (F = 1.468, p = .236) or the posttest (F = 0.680, p = .509). Compared with the pretest, students in the remote association condition (p = .000, d = 1.584) and close association condition (p = .000, d = 1.8817) both got significantly higher scores in the posttest.

For Flexibility, the one-way ANOVA shows a significant difference among the three conditions in the posttest (F = 8.469, p = .000), while not in the pretest (F = 1.920, p = .152). In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 7.686, p = .001), we run a one-way ANOVA Post Hoc analysis (Tamhane’s T2). Results show that compared with the free association condition, the remote association (p = .001, d = 0.9129) and close association (p = .044, d = 0. 591) performed differently on fostering students’ flexibility of creative thinking. In addition, there was no significant difference between remote association and close association (p = .490).

For Uniqueness, the one-way ANOVA shows that there is a significant difference among the three conditions in the pretest (F = 10.396, p = .000), while not in the posttest (F = 3.042, p = .053). According to the paired samples t-test, the remote association condition fostered students’ performance of Uniqueness significantly (p = .000, d = 1.0256). Students in the close association also performed significantly better in the posttest (p = .001, d = 0.1153).

Creativity aptitude

Imagination, risk-taking, curiosity, and challenge were analyzed for Creativity aptitude. We calculated ANOVA on the total score of the Creativity aptitude survey. In the pretest, results showed no significant difference among the three conditions (F = 2.486, p = .089). In the posttest, a significant difference existed among the three conditions (F = 20.189, p = .000) (see Table 6).

Table 6.

Means and standard deviation of students’ creative aptitude test

WCAT items	Remote association(N = 30)		Close association (N = 32)		Free association (N = 32)
	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)
Imagination	28.67 (2.657)	29.03 (2.871)	29.31 (3.021)	31.25 (2.396)	31.44 (2.409)	29.19 (2.191)
Risk-taking	24.40 (2.581)	26.20 (2.041)	25.25 (2.342)	26.81 (2.250)	25.00 (1.884)	28.31 (2.320)
Curiosity	34.27 (3.600)	34.30 (2.215)	35.00 (2.664)	36.94 (2.078)	35.88 (3.626)	37.50 (2.214)
Challenge	30.57 (2.096)	30.50 (1.925)	29.94 (2.699)	31.56 (1.831)	31.44 (2.409)	30.81 (1.908)
Total	119.40 (8.669)	120.03 (6.003)	121.00 (8.096)	126.56 (6.180)	124.06 (8.428)	125.81 (6.255)

In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 0.116, p = .890), we ran a one-way ANOVA Post Hoc analysis (Least Significant Difference, LSD). Results showed that the significant difference only existed between remote association and free association (p = .037, d = 0.5227). In addition, there was a significant difference between remote association and close association (p = .000, d = 1.0719). That was, different association interventions affected students’ creativity aptitude, and the effect of the close association condition was more obvious than that of the remote association condition.

For Imagination, the one-way ANOVA showed a significant difference among the three conditions in the posttest (F = 12.975, p = .000), while not in the pretest (F = 1.819, p = .168). In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 1.746, p = .180), we ran a one-way ANOVA Post Hoc analysis (Least Significant Difference, LSD). Results showed that there was a significant difference between remote association and free association condition (p = .000, d = 1.2374), while no difference between close association and free association condition (p = .136). In addition, there was also a significant difference between remote association and close association (p = .001, d = 0.8396).

For Risk-taking, the one-way ANOVA showed a significant difference among the three conditions in the posttest (F = 7.566, p = .001), while not in the pretest (F = 1.128, p = .328). In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 0.549, p = .579), we ran a one-way ANOVA Post Hoc analysis (Least Significant Difference, LSD). Results showed that there was a significant difference between remote association and free association (p = .000, d = 0.9657) as well as close association and free association (p = .008, d = 0.6564). In addition, there was no significant difference between remote association and close association (p = .278).

For Curiosity, the one-way ANOVA showed a significant difference among the three conditions in the posttest (F = 19.029, p = .000), while not in the pretest (F = 1.825, p = .167). In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 0.081, p = .922), we ran a one-way ANOVA Post Hoc analysis (Least Significant Difference, LSD). Results showed that there was a significant difference between remote association and free association condition (p = .000, d = 1.445), while no difference between close association and free association condition (p = .302). In addition, there was also a significant difference between remote association and close association (p = .000, d = 1.2293).

For Challenge, the one-way ANOVA showed a significant difference among the three conditions in the pretest (F = 3.098, p = .050) and also in the posttest (F = 4.182, p = .018). Then paired samples t-test was conducted to compare the change of each condition. Results showed that the close association (p = .000, d = 0.7025) fostered students’ challenge reaction on the Creativity aptitude survey, but not in the remote association (p = .856).

Empathy

For the empathy survey, we calculated ANOVA on pre- and post-tests by condition. Results showed that no significant difference existed among the three conditions either in pretest (F = 0.496; p = .611) or posttest (F = 2.352; p = .101) (see Table 7). Then paired samples t-test was conducted to compare the change of each condition. Results showed that both remote association (p = .000, d = 1.1882) and close association (p = .000, d = 1.6606) fostered students’ overall reaction on empathy.

Table 7.

Means and standard deviation of students’ basic empathy test

BES items	Remote association (N = 30)		Close association (N = 32)		Free association (N = 32)
	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)	Pre M (SD)	Post M (SD)
Cognitive empathy	31.20 (2.497)	33.93 (3.443)	31.00 (3.132)	34.69 (2.177)	31.50 (2.759)	34.25 (2.817)
Affective empathy	31.63 (4.072)	35.10 (4.708)	30.81 (2.520)	33.19 (4.353)	31.06 (3.943)	32.69 (5.866)
Total	62.83 (4.178)	69.03 (6.083)	61.81 (2.788)	67.88 (4.353)	62.56 (5.267)	66.94 (6.470)

For cognitive empathy, the one-way ANOVA showed a significant difference among the three conditions in the posttest (F = 5.386, p = .006), while not in the pretest (F = 0.256, p = .775). In the posttest, according to the result of the Test of Homogeneity of Variances (Levene Statistic = 2.663, p = .075), we ran a one-way ANOVA Post Hoc analysis (Least Significant Difference, LSD). Results showed that there was a significant difference between remote association and free association condition (p = .002, d = 0.7375) as well as close association and free association condition (p = .031, d = 0.6197). In addition, there was no significant difference between remote association and close association (p = .300).

For affective empathy, the one-way ANOVA showed no significant difference among the three conditions in the pretest (F = 0.427, p = .654), and not in the posttest (F = 1.255, p = .290) either. Then paired samples t-test was conducted to compare the change of each condition. Results showed that both remote association (p = .000, d = 0.7884) and close association (p = .000, d = 0.6692) fostered students’ affective empathy.

Design scheme

Quantitative results of the Creative thinking test, Williams creativity aptitude test (WCAT) and Basic Empathy Scale (BES) showed that the association interventions had a certain effect in promoting the cultivation of students’ creative ability. To investigate the characteristics of students’ creative thinking, we analyzed the design scheme generated by students during the process of generating creative ideas through association. Some students’ design sketches under the three conditions were also shown in Figs. 3 and 4.

Fig. 3.

Examples of students’ Mask design sketches. (Note: Mask design: Inspired by the “Face-changing” technique in Sichuan Opera, students designed a reusable mask (with replaceable filters) that can show the changes in people’s emotions (left), and a school-specific multifunctional reusable mask (right).)

Fig. 4.

Example of students’ 3D glasses design sketches. (Note: 3D glasses design: Inspired by the patch, students designed auto-massaging 3D glasses (left), while inspired by the bulb, students designed 3D glasses that are with oval-shaped lenses, besides, the glasses are antivirus, automatic cooling, and shatter-resistant (right).)

Students’ design ideas were scored on five dimensions: fluency, flexibility, uniqueness, precision, and sensitivity. As shown in Fig. 5, compared with the control group, both remote and close association interventions have promoted the generation of students’ creative ideas. Besides, students under the close intervention condition performed best and achieved highest scores among the three groups in all five dimensions.

Fig. 5.

Results for design scheme analysis

Students interview

Through interviews, we found that students had a positive attitude in constructing problem situations, analyzing the attributes of associations, compelling related associations, collaborative group discussions, hands-on operations, group presentation and evaluation. Students reported that the association intervention not only provided them with ideas for diverging associations in the process of idea generating, but also allowed them to converge the ideas generated. At the same time, the association interventions promoted their learning by mobilizing enthusiasm for active learning and stimulating their learning interest. In addition, through compulsory association, students not only learned to use association clues to generate ideas that were more suitable for user needs, but also cultivated their creative thinking and creativity.

Students got practical experiences in creative problem-solving, empathy, and cooperation. Firstly, students reported that authentic problem situations could help inspire empathy. Everyone was faced with these problems that were closely related to life. For example, a student said: “Because we wear masks every day, such as going to school, taking the subway, we can understand the imagination of some special functions of masks during the COVID-19 pandemic” (Student 2). “When the cinema resumes business, audiences need to keep a safe distance when watching movies, and the 3D glasses used in public are not hygienic. These problems are real for each of us. So, when the teacher proposed the task of designing 3D glasses, everyone actively expressed their real experience and needs” (Student 4).

Secondly, when students were cued to make relations between the association stimulus and targets, their creative thinking and creativity were fostered. They actively participated in the association activities and team discussions. There were obviously more ideas generated in the second design project (3D glasses) than in the first one (Mask), and the speed was also faster. One student said: “I have never thought the connections between these things, such as apple and mask, light bulb and 3D glasses. In this course, during the compulsory association process, our group members propose kinds of interesting and well-founded opinions.” (Student 8).

Thirdly, students reported that through group collaboration, the members could listen to the opinions of others, and learned from each other. Through group discussion, they determined the theme of the design together. Students made joint effort and finally completed the design project. For example, student said: “When we discussed ‘apple’ in our group, from the external characteristics, some said that apples have red, green and yellow colors and associated the idea of discoloration; from the perspective of function, some put the point of view that as a kind of fruit, apples can provide energy for people etc. In this process, we are inspired from others, and our communication skills are also improved.” (Student 1). Student also said: “Our group combined the different design schemes suggested by each member, and through the discussion inside the group, we finally designed our own unique design, this feeling is very good” (Student 5).

Finally, students gave positive feedback on the group reporting. Through the group presentation and evaluation, they could find problems of their design and solve them in time, which was conducive to promoting reflection. Besides, students could learn from other groups, and make progress together. “This kind of evaluation can help us learn the advantages of other groups” (Student 7). “We score ourselves, so that teacher evaluation is not the only determinant of performance. This can make the evaluation more objective” (Student 6).

Students also put forward some suggestions. For example, some students suggested that in the process of compulsory association, teachers could provide them with more introductions related to the association stimulus to broaden their association range. Some students said that they encountered some difficulties in finding resources and hoped that more resources could be provided.

Totally, the association interventions have a certain effect on students’ creative thinking, creativity, empathy, collaborative ability, and creative problem-solving ability. Although there are still some problems, it gained overall satisfactory feedbacks from students.

Discussion

In respect of the two research questions stated in Sect. 3, we may conclude that the association interventions have an impact on the cultivation of students’ creativity in the present STEAM course. Overall, the remote association and close association interventions could both promote students’ creative thinking, creativity aptitude, and empathy (RQ1). The remote association intervention worked better to help students to achieve a higher degree of creative thinking. While the close association intervention was more effective in improving students’ creativity aptitude and quality of design ideas (RQ2). In addition, for some aspects, the free association condition also has its advantages. Further discussion is elaborated as follows.

Why did the association interventions foster students’ creativity?

With respect to RQ1, as expected, students who participated in the STEAM course under the two association interventions gained positive performance of creativity by their progress from pre- to post-test. Students in the remote and close association conditions generated more ideas, and the speed was also faster, especially in the second design project (3D glasses), indicating that the association clues enhanced students’ creativity on the design-focused creative problem-solving tasks. This is consistent with the former studies (Wilson et al., 2021; Shen et al., 2021; Timotheou & Ioannou, 2021). The possible explanation is that association interventions construct compulsory association situations, in which students are cued to generate uncommon and useful ideas. The remote and close association interventions stimulated students to generate multiple creative ideas through providing association clues. On one hand, these clues clearly pointed out the association directions and also provided students with scaffolds for idea generating. When being explicitly asked to generate creative ideas, students are likely to generate more creative ideas (Acar, Runco & Park, 2020; Benedek et al., 2020; Green et al., 2015; Prabhakaran et al., 2014; Said-Metwaly et al., 2020; Weinberger et al., 2016). Besides, association stimulus may help break through psychological stereotypes, therefore, students could generate more uncommon and distantly related design ideas.

On the other hand, association intervention relates to creativity because it enables persistence (i.e., sustained task-directed effort). Invoking the Dual Pathway to Creativity Model, De Dreu et al. (2012) hypothesized that working memory capacity (WMC) is related to creative performance because it enables persistent, focused, and systematic combining of elements and possibilities (persistence). It showed that the cognitive inhibition enabled the individual to maintain attention focused on the task and prevented undesirable mind wandering. Gupta et al. (2012) tested the hypothesis that highly creative individuals, as measured by the RAT, were able to access remote associations because they were not biased to consider only high-frequency words. In the experiment, for each RAT question, participants were presented with three cue words and asked to find the common associative link among them within the allotted time of 30-seconds. Result supported the claim that the tendency to consider high-frequency words impairs performance on the RAT. Individuals who performed poorly on the RAT should be biased to give high-frequency responses, while others who performed well in RAT were more inclined to use concepts that have low-frequency connections with cues as answers to questions. By clearly requiring students to break psychological stereotypes, the association stimulus used in the present study helped students to generate creative solutions that requiring avoidance of high-frequency candidate answers.

Another possible explanation is that the compulsory association activities were relatively new to the participants, so involved in the experimental course helped stimulate students’ learning motivation, increase engagement, and stimulate creativity (Rutland and Barlex, 2008). Generally, when engaging in a design process to solve a defined problem, students brainstorm possible solutions, evaluate and prioritize alternative solutions, and then decide among existing alternatives (Aranda, Lie, & Guzey, 2020). Without direction and restraint, the ideas produced may not be able to meet the expectations. But the compulsory association helps extend associations from familiar fields to unfamiliar fields, even unexpected fields. In other words, it not only makes full use of existing designs, but also benefits from their mutual combination.

Finally, empathy can also promote creativity. Empathy and the acquisition of empathy are considered essential components of adequate moral development (Jolliffe & Farrington, 2006), and empirical relations between many forms of prosocial behaviour and empathy have also been studied (e.g. Batson, Fultz, & Schoenrade, 1987). In line with this assertion, empathy helps raise feelings of human-centeredness and enhance thoughtfulness on design decisions. Therefore, while being asked to solve design problems that addressed the client’s demand, students showed consideration for others, thereby promoting the novelty and usefulness of creative thinking (Gong, 2018).

Why did the remote association and close association show different effects on creativity cultivation?

With respect to RQ2, results showed that the remote association intervention worked better than the close association intervention on improving students’ creative thinking and empathy. While in the creativity aptitude survey and the creative idea analysis, students in the close association condition outperformed the remote association condition.

First, empirical studies have confirmed that remote association intervention may promote creativity better. For instance, Howard-Jones and colleagues (2005) conducted an experiment that requiring participants to produce creative (generate a story that was as creative as possible) and uncreative (generate a story that was as uncreative as possible) stories from related (e.g., magician, trick, rabbit) and unrelated (e.g., flea, sing, sword) word sets. In particular, the strategy that incorporating a set of words that were unrelated to each other was used to encourage semantic divergence. Results showed that the inclusion of unrelated words in the stories improved the rated creativity of the outcomes. It was also evident that the rated creativity of the outcomes of the stories was influenced by providing a ‘‘Be creative’’ or ‘‘Be uncreative’’ objective. In brief, piecing together a story from unrelated words predictably led to more creative responses than doing the same for related words, and it would appear that the rated creativity of stories was influenced more by objective in the unrelated conditions than in the related conditions.

On the basis of the remote association theory, Gupta et al. (2012) also supported the claim that a tendency to consider high-frequency words impairs performance on the RAT. Individuals who performed poorly responded with high-frequency incorrect words. Therefore, remote association may be confirmed as a key enabler of successful creative idea generation as it contributes to promote the uncommon and long-distance semantic ideas, and reduce the tendency to produce high-frequency ideas.

Second, the remote association condition did not outperform the close association condition in the creativity aptitude survey and the creative idea analysis as predicted. According to our study, different association interventions affected students’ creativity aptitude as well as the creative problem-solving ability, and judging from the total score, the effect of the close association condition was more obvious than that of the remote association condition. Specifically, for the creativity aptitude survey, students involved in the close association condition reported better response on the aspect of curiosity and imagination. Furthermore, for the challenge dimension, students in the close association condition performed significantly better in the posttest than the pretest, in contrast, students in the remote association showed a little declination. As for the risk-taking dimension, there was no significant difference between these two association intervention conditions. Possible explanations for the results may be that (a) the stimuli used in the close association intervention have a closer attribute relationship to the type of target object than those used in the remote association condition, thus it is easier to guide students to produce creative ideas within a certain range during the ideating process (Gruszka & Necka, 2002). The imagination was thus far more prompted. (b) the closer relationship between the object and the association stimulus may reduce fear of difficulty, then it is easy for students to have the feeling of ​​“the two are more easily related”, thus generating strong curiosity and challenge (Wu et al., 2016). (c) risk-taking may be susceptible to interference from the external environment. The provided association stimuli have shown the directions for the design project, so the risky nature has not changed much.

For the creative scheme analysis, students under the close intervention condition performed better than the remote association condition in both two design projects. It may be inferred that considering the original creativity level of the participants, the close association intervention may be more effective at promoting the near-transfer than the remote association for far-transfer. Also, as Abraham et al. (2012) indicated, it is cognitively far more demanding to generate uncommon uses than common uses, even when explicitly told to generate uncommon uses, it is highly likely that the subjects will also think about common uses while doing so, and vice versa. It is undeniably more cognitively demanding to generate novel design ideas from unrelated association stimulus compared to related association stimulus. In STEAM education, similar result was also reported. Conradty & Bogner (2019) reported that the Short-term STEAM module produced long-term knowledge and built stable scores of intrinsic motivations, but not self-reported aspects of creativity.

What are the differences among the three conditions for the measured variables?

For the posttest, significant differences among the three conditions were reported both in the creativity thinking test and the creativity aptitude survey. While for the empathy survey, there existed no significant difference. Besides, the design idea analysis also provided evidence of the effectiveness of the association interventions, that is, students participated in the association intervention conditions reported better learning outcomes than those under the brainstorming condition. As predicted, the compulsory association conditions showed effectiveness on improving students’ creativity.

But it is worth noting that free association also has its own advantages on creativity cultivation. Results show that for some dimensions, such as the risk-taking, curiosity and imagination of creativity aptitude, and also the cognitive empathy, free association condition can promote more positive outcomes. As for the reasons, one is that the design problems are complex and ill defined, and the creative problem-solving process is also extremely complicated. The creative thinking process may consist of four stages: preparation, incubation, illumination and verification) (Wallas, 1926), which suggests that the creativity does not happen overnight. Moreover, existing studies have discussed multiple elements possibly related to creative thinking, such as attention, working memory, executive function, intelligence, and emotion. (Bott et al., 2014; Moss, & Wilson, 2015; Lin, & Lien, 2013; Beaty et al., 2014; Benedek et al., 2014). In particular, another important reason for the results may be related to the conflicting views that the creative thinking is happened consciously or unconsciously (Beaty et al., 2014; Benedek & Jauk, 2018; Martindale, 2007; Segal, 2004; Smith, 1995; Sowden et al., 2015). From this perspective, it is understandable that the free association can promote part of the dimensions of creativity better.

In conclusion, the findings illustrate that students could raise the level of creativity by deliberate training under different association strategies through the STEAM course. From one side of the coin, association intervention helps guide the development direction of individual creativity, but it does not determine its level. From another side, to some extent, the compulsory association may limit students’ thinking within a certain range. Students’ original level may affect the effect of creativity training: students who are originally at a low level of creativity may be more suitable for close association or brainstorming training, and students who are originally at a high level of creativity may be more suitable for remote association (Clapham, 1997; Fleith, Renzulli, & Westberg, 2002; Ge & Bai, 2007). More elements, such as time pressure, creative thinking stages, task types and so on, may also be taken into consideration for nurturing students’ creativity (Abraham et al., 2018; Conradty & Bogner, 2019; Wallas, 1926).

Limitations and future studies

The present study reported that STEAM course integrated with association interventions is beneficial for nurturing creativity. However, as creative thinking is among the most complex of human abilities (Abraham et al., 2012), to further elucidate the effects of association intervention on creativity, and the differences among association interventions through which link to creative performances, more empirical studies are needed. For example, it could be interesting to explore the effect of the combination of different association interventions, the effect of different stimuli amounts and types, and the effect of long-term vs. short-term interventions, etc.

Funding

This research was financially supported by the National Natural Science Foundation in China (62277018; 62237001), Ministry of Education in China Project of Humanities and Social Sciences (22YJC880106), the Major Project of Social Science in South China Normal University (ZDPY2208).