Introduction  

Learning is described as the act, process, or experience of gaining knowledge or skill, while style is a quality of imagination and individuality expressed in one’s actions and tastes. Learning style is a term used to express individual differences in the processes of learning, including the way by which a learner begins to focus on the process to absorb and retain information [1]. A person’s learning style is a relatively stable pattern of behavior and is based on their background and experiences [2]. Learning styles have attracted the attention of researchers, educators, students, and parents of students over the past decades [3]. Deep and systematic research has been conducted into problems related to learning, and both theoretical and empirical findings have been widely used in the field of education. Coffield et al. (2004) conducted a literature review citing findings ranging from the early 20th century to 2004 and found 71 models of learning styles [4]. These models can be divided into five categories: constitutionally based learning styles and preferences, cognitive structure, stable personality types, “flexibly stable” learning preferences, and learning approaches and strategies [5]. There are 13 highly influential models, including Kolb’s Learning Style Inventory (LSI), Honey and Mumford’s Learning Styles Questionnaire, Cognitive style analysis, and the Myers-Briggs type indicator. Among these, arguably the most influential is the LSI. Statistical reports indicate that 4,400 articles relating to the LSI (including Experiential Learning Theory, ELT) were published between 1971 and 2019 [6], with an average of 92 research outputs published on LSI each year. The use of LSI is widespread and the instrument itself has been translated into Arabic, Chinese, French, Italian, Russian, Spanish, and Swedish [4].

Kolb describes learning as the process whereby knowledge is created through the transformation of experience, resulting from the combination of grasping and transforming experience. He asserts that learning is also a holistic process of adapting to the world, and integrating thoughts, feelings, perceptions, and behaviors [7]. In the summer of 1966, Kolb devised ELT based on the existing theory created by John Dewey, Kurt Lewin, Jean Piaget, Carl Jung, and Carl Rogers. In particular, he was inspired by Carl Jung's theory of psychological types [4]. ELT outlines learning to be the process whereby knowledge is created through the transformation of experience. ELT helps to explain how experience is transformed into learning and reliable knowledge. A key feature of ELT is the mode of the experiential learning cycle, comprising four nodes: Concrete Experience (CE), Reflective Observation (RO), Abstract Conceptualization (AC), and Active Experimentation (AE). Kolb characterizes these four nodes thus: CE relates to learning from specific experiences, relating to people, and being sensitive to feelings and people; RO relates to carefully observing before making judgments, viewing issues from different perspectives, and looking for the meaning of things; AC relates to logically analyzing ideas, planning systematically and acting on an intellectual understanding of the situation; AE relates to showing an ability to get things done, taking risks, and influencing people and events through action. The four nodes are assigned to the points of a compass, such that CE is “north”, RO is “east”, AC is “south”, and AE is “west”. Kolb describes the process of experiential learning in terms of this four-stage cycle, where these four nodes comprise the four learning modes and specifies a label for each mode, as follows: CE is feeling, RO is watching, AC is thinking, and AE is doing. These four labels reflect an individual’s preferred direction for personal learning (Figure 1).

Based on this four-cycle model, Kolb proposed a method of learning style evaluation based on an individual’s unique combination of preferences for CE, RO, AC, and AE. This forms a profile in the form of a “kite” shape and facilitates the identification of the individual to one of four types of LSI learning styles, namely diverging (CE/RO), assimilating (AC/RO), converging (AC/AE) and accommodating (AE/CE). Divergence signifies a reliance on apprehension transformed by intention; assimilation represents comprehension transformed by intention; convergence is the extensive transformation of comprehension; and accommodation is the extensive transformation of apprehension [4]. Similarly, Kolb also carries forwards the above labels to the four learning styles, such that diverging is feeling and watching, assimilating is watching and thinking, converging is doing and thinking, and accommodating is doing and feeling. The LSI is an inventory of learning styles, designed and revised by Kolb with a theoretical basis in ELT. The first version of the LSI appeared in 1976, the second in 1985, the third in 1999, and the fourth in 2012 [7]. The first version of the LSI comprised a brief questionnaire consisting of nine rows of four words which are to be ranked in order. The words in each row correspond to Kolb’s four nodes (or learning modes): CE, RO, AC, and AE. The second version consisted of 12 sentence-completion items in which respondents attempt to self-describe their learning styles. In 1982, Mumford and Hone, based on Kolb's ELT, combined their own research and practical experience to develop the Learning Styles Questionnaire (LSQ) to match and improve the application of LSI. The LSQ is a questionnaire comprising 80 items, with the outcome of this questionnaire corresponding to the four nodes. Thereafter, the LSI produced outcomes using the established concepts of CE, RO, AC, and AE in the format CE/RO, AC/RO, AC/AE, and AE/CE [7], to produce the final "learning style" model (Figure 1). In the 2005 revision, Kolb's learning model classification increased from four nodes to six, with the addition of AC/CE and AE/RO, and finally to nine. These final nine classifications are: initiating style, experiencing style, imagining style, reflecting style, analyzing style, thinking style, deciding style, acting style, and balancing style [7]. Subsequently, the four basic learning styles devised in Kolb’s LSI have been referred to in the literature using three different sets of terms: (1) diverger, assimilator, converger, and accommodator; (2) divergent, assimilative, convergent, and accommodative; (3) diverging, assimilating, converging, and accommodating. However, these terms only differ in semantics. For consistency, this article will use set (1), see Figure 1.

There were 4,400 officially published outputs from studies using the ELT and LSI numbered between 1971 and 2019 [6], spanning over 30 disciplines, including education, agriculture, computer science, engineering, biology, economics, geography, medicine, physics, and psychology [7]. One study used the LSI to classify 241 medical teachers, medical students, and residency doctors and found that the majority of the residency doctors to be of the assimilating style [8]. In a survey of 204 undergraduate medical students, 47% were found to be assimilating. Such students learn through inductive reasoning, focus on abstract ideas and concepts, and have the ability to create theoretical models by assimilating disparate observations into a coherent and cohesive explanation [9]. Amod & Brysiewicz (2019) reported that teachers adopt experiential teaching strategies to improve the quality of teaching by applying human body simulators in nursing teaching, with good results [10]. Also, researchers developed a tool using the concept of experiential learning that tracked the impact of the tool in 12 cases. Experiential learning has been introduced into innovative education [11].

In 1975, Kolb reported that students with humanities and liberal arts backgrounds were more likely to be divergers; while business administration students were often accommodators; students with applied science backgrounds were typically convergers, and students with backgrounds in pure science, mathematics, and law were typically found to be assimilators. Other studies have found statistically significant associations between students’ learning style preferences and academic achievement [12]. One study reported the adoption of experiential learning methods and applied human simulators to the teaching of midwives, which played a positive role in the interactions between students and teachers and the improvement of the learning effect [13]. Although some studies have found the LSI to have low reliability and construct validity. However, some studies have identified that in multiple learning theory and learning style models, Kolb’s experiential learning model and the LSI are found to be the most prominent theory and most commonly-used instrument. It has been reported to use a particular form of experiential learning teaching to stimulate students’ own involvement in the integration of their thoughts, feelings, and attitudes [14]. Annette & Vince (2019) discusses the value of learning from a psychodynamic approach to experiential learning [15]. Two new learning formats have been suggested, that allow students to systematically acquire active–motivational competencies: self-inquiry-based learning and self-experience-based learning. Research findings indicate that SIBL and SEBL are promising approaches for this purpose [16]. These findings taken together provide evidence for the validity and reliability of the ELT and Kolb’s LSI.

Despite extensive previous research investigating the theory and applications of LSI, no study has yet reported the distribution of Kolb’s Learning Styles across a population of college students from a variety of family backgrounds. This study aims to investigate the family backgrounds of a group of medical college students alongside their Learning Styles and investigate the association between Kolb’s Learning Style and family background in this cohort. The findings of this study will influence the design of future research in experimental learning.

Materials and Methods

Study design

College students were recruited to participate in the questionnaire survey. Participants gave informed consent before participation. Participants were asked to give relevant demographic information (gender, age, and level of education completed), family background (birth order, guardians, level of education completed by their father and mother, family financial situation, and place of residence), and to complete the LSQ.

Setting

The study participants were graduates from several nearby medical schools who sat the resident training entrance examination at Shanwei Second People's Hospital in China between February 1 and July 28, 2021. Only students who were willing to participate in the study were included. Any participant who made lack of data completing the questionnaire was excluded.

Ethics

Students were informed that their participation in the study was entirely voluntary and gave informed oral consent. Students were assured that their information would remain confidential. The study protocol was granted ethical approval (reference: 20210201) by the Ethical Review Board at Shanwei Second People’s Hospital.

Participants

A cohort of 108 participants completed the questionnaire study .

Instrument

The instruments used in this study were: (1) questionnaires capturing demographic factors and family background and (2) the LSQ [17].

Procedure

Data collection ran from February 1 to July 28, 2021, and analysis was conducted between August 1, 2021, and October 1, 2021. The research procedure was as follows: first, participants completed the demographic and family background questionnaires; second, participants completed the LSQ; third, the research team collated the data, removed invalid responses, confirmed that the responses were complete and valid, and input the data into the computer, whereupon the data were analyzed using Statistical Package for the Social Science (SPSS) version 19.0 (IBM) and finally reported.

Data analysis

Data were pre-processed using Microsoft Excel 2007 to determine the completeness and validity of the data before importing into SPSS. Descriptive statistics were used to characterize the distribution, variability, and range of the data. Associations between family background variables and LSQ outcomes were assessed, focusing on the taxonomic data of family background and the associated LSQ classification, using a Bivariate Correlation (Spearman) analysis.

Results and Discussion

Demographic information

A total of 108 subjects participated in this study, of whom, 58 (53.7%) were men and 50 (46.3%) were women. The average age was 23.9 years, and the median was 24 years. The level of education completed was as follows: 2 (1.8%) had Master’s degrees, 88 (81.5%) had undergraduate degrees and 18 (16.7%) had graduated from junior college.

Family background

Birth order

Following the methodology [18] used by Liu et al., we divided the participants by birth order into (1) only children (OC; n=9; 8%), (2) one sibling (OS; n=28; 26%), and (3) multiple siblings (MS; n=71; 66%).

Guardianship

We divided the participants by their guardianship during childhood, as follows: growing up with their parents (GP; n=87; 81%), growing up with their grandparents (GG; n=14; 12%), growing up in a single-parent family (GS; n=2; 2%), and other (OT; n=5; 5%).

Parents’ educational attainment

Paternal educational attainment was as follows: 2 (2%) had undergraduate degrees, 4 (4%) had graduated junior college, 32 (29%) had graduated high school, 52 (48%) had only completed junior high, 18 (17%) only primary school, and 0 were illiterate. Maternal educational attainment was as follows: 0 had undergraduate degrees, 4 (4%) had graduated junior college, 13 (12%) had graduated high school, 25 (23%) had only completed junior high, 64 (59%) only primary school, and 2 (2%) were illiterate.

Family financial situation

The family financial situation was divided into four levels: poor (an annual family income of less than $12520), low income (an annual income of $14085-$23475), middle income (an annual income of $25040-$78250), and high income (an annual income of more than $79815). Of the 108 participants, 22 (20%) were classified as poor, 72 (67%) low income, 12 (11%) middle income, and 2 (2%) high income.

Place of living

Participants' place of living was divided into (1) city (n=28; 26%) and (2) township or rural (n=80; 74%).

LSQ

LSQ distribution

The distribution of participants’ LSQ outcomes is given in Table 1. Of the 108 participants, 64 (59%) were assimilators, 24 (22%) were divergers, 12 (11%) were convergers, and 8 (7%) were accommodators.

Note. Honey and Mumford’s Learning Styles Questionnaire = LSQ

LSQ and family background

The family background variables assessed for association with LSQ were birth order, guardianship, paternal education, maternal education, family financial situation, and place of living. The associations between LSQ style and participant demographic characteristics are given in Table 2.

Note. Honey and Mumford’s Learning Styles Questionnaire = LSQ

Only Children = OC  One Sibling = OS  Multiple Siblings = MS  Growing with their Parents = GP 

Growing up with their Grandparents = GG  Growing in a Single-parent family = GS  Other = OT

Considering birth order, 67% of OC participants were assimilators, which is 9% higher than MS participants. Conversely, 25% of MS participants were divergers, while no OC participants were divergers. 62% of GP participants were assimilators, and 14% of GG participants were accommodators. Considering parental education, 48% (52) of participants' fathers’ had completed junior high, and among these, 33 participants were assimilators. Of the 64 (59%) of the participants whose mothers had completed primary school, 37 were assimilators. Of the 67% (72) of participants from low-income families, 44 were assimilators. Of the 80% (n=80) participants from rural households, 51 were assimilators.

The distribution of LSQ style across participants with different family background factors is described in Table 3 using the maximum, minimum, range, variance, standard deviation, and standard error.

Note.

Only Children = OC  One Sibling = OS  Multiple Siblings = MS  Growing with their Parents = GP 

Growing up with their Grandparents = GG  Growing in a Single-parent family = GS  Other = OT

The distribution of LSQ style with guardianship varied greatly, with GP participants showing the most highly represented style (n=54) and the least represented style (n=5) to differ by 49 participants. Figure 2 shows the LSQ distribution with factors characterizing the participant’s family background.

Figure 2 shows the greatest variation in the family backgrounds of assimilators, also with a wide range. The family backgrounds of accommodators vary the least with a relatively small range.

The associations between LSQ style and family background factors

Results of the bivariate correlation (Spearman) analysis are given in Table 4. These figures represent the different strengths of correlation between family background factors and LSQ style, identifying which are statistically significant (p<0.01).

Note.

Honey and Mumford’s Learning Styles Questionnaire = LSQ

Two-tailed tests.** p <.01.