Cognitive Load and Energy Balance Knowledge in High-School Physical Education

Abstract

Purpose:

Guided by the cognitive load theory, the purpose of this study was to determine the impacts of cognitive load and school socioeconomic status-related environmental factors on ninth-graders’ energy-balanced living knowledge gain.

Methods:

A stratified random sample of high-school students (N = 150) participated in this study. Data were collected on students’ knowledge gain, cognitive load, free and reduced-price meal rates, and student-to-teacher ratio.

Results:

The path analysis results revealed that the reasoning learning tasks had direct significant effects on students’ knowledge gain (β_{i-Diet and i-Exercise} = 0.34, p < .01). The free and reduced-price meal rates and student-to-teacher ratio did not have significant effects on students’ knowledge gain (p > .05).

Discussion:

These findings advance our understanding of the role cognitive learning tasks play in enhancing student learning in the subjects of energy-balanced knowledge and healthy lifestyle.

Learning health-related fitness knowledge has been the center of physical education (PE) intervention for achieving physical literacy (Ennis, 2015). The Society of Health and Physical Educators, or SHAPE America (2014), has identified physical literacy as the ultimate learning goal for PE, which is defined as “the motivation, confidence, physical competence, knowledge and understanding of maintaining physical activity throughout the lifecourse” (Whitehead, 2010, pp. 11–12). SHAPE America (2014) also includes learning scientific knowledge in two of the five national learning standards: “Standard 2: the physically literate individual applies knowledge of concepts, principles, strategies, and tactics related to movement and performance” and “Standard 3: the physically literate individual demonstrates the knowledge and skills to achieve and maintain a health-enhancing level of physical activity and fitness” (p. 1). Knowledge related to health-related fitness and physical activity participation principles is regarded as the most relevant knowledge that PE teachers should teach (Santiago & Morrow, 2020). One critical component of health-related fitness knowledge includes the concepts associated with energy-balanced living. Teaching youth to live an energy-balanced lifestyle directly helps them make sound daily decisions about food intake and exercise. Energy-balance knowledge refers to the concepts related to energy intake through eating, drinking, and energy expenditure through physical activity (Dunford, 2010). It focuses on the principles and strategies that guide behaviors to balance energy intake and energy expenditure (Chen & Chen, 2012).

The Knowledge Learning Focus in PE

Youth fitness and physical activity behavior are strongly correlated with health-related fitness knowledge. In the academic literature, there is a well-documented link between fitness knowledge and physical activity (Chen & Nam, 2017; Liu & Chen, 2020; Thompson & Hannon, 2012). Specifically, Thompson and Hannon (2012) reported a moderate positive relationship between health-related fitness knowledge and self-reported physical activity in high-school students (r = .438, p < .001). Liu and Chen (2020) reported that the lack of health-related fitness knowledge predicted sedentary behavior for middle-school boys (β = 0.14, t = 2.16, p = .03, ${\mathit{R}}_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}}^{\mathrm{2}}\mathrm{=}.02$) and girls (β = 0.16, t = 2.56, p = .01, ${\mathit{R}}_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}}^2\mathrm{=}.02$) and for high-school boys (β = 0.25, t = 2.38, p = .02, ${\mathit{R}}_{\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}}^2\mathrm{=}.05$). Wang and Chen’s (2020) structural equation model revealed a direct predictive relation in middle-school students between health-related fitness knowledge and outside of school physical activity participation with their motivation as the moderator.

Cognitive Load Theory: A Primer

Learning is defined as “a multidimensional process that results in a relatively enduring change in a person or persons, and consequently how that person or persons will perceive the world and reciprocally respond to its affordances physically, psychologically, and socially” (Alexander et al., 2009, p. 186). In other words, learning is a process for students to acquire and apply knowledge relevant to addressing changes in life. Modern strategies to teach knowledge are mostly based on the constructivist learning theory (Ennis, 2015). The theory is considered one important guide to the design of cognitive and physical engagement tasks in PE interventions (Ennis, 2015).

Three major tenets from the constructivist theory family have guided the understanding of learning processes: zone of proximal development (ZPD), cognitive constructivism, and conceptual change. The ZPD is defined as “the distance between the actual development level as determined by independent problem-solving and the level of potential development as determined through problem-solving under guidance or in collaboration with more capable peers” (Vygotsky, 1978, p. 86). The ZPD postulates that learning is most effective when taking place in a supportive environment with well-defined ZPD for the learner. The cognitive constructivism perspective emphasizes learning as a personal meaning-making process through concrete cognitive experiences in each individual learner (Piaget, 1971). It signifies the mechanisms that take place in the confinement of individual experiences in and outside the learning context such as classrooms or schools. The conceptual change theory specifies learning as conception development from a naïve mental representation (naïve models) to a scientific representation. It delineates specific ways a concept is learned in relation to other concepts either existing in the knowledge repertoire (prior knowledge) or being processed simultaneously. Taken the three tenets together, a constructivist learning experience dictates that the learner be involved in specific conception development (changing of mental models) with personal meaning making in a socially supportive environment. Because the learning environment of PE is centered on physical activity participation, the knowledge learning process can be drastically different from that in the classroom. When the goal of a PE curriculum is to foster knowledge learning, the relationship between the cognitive process and physical engagement must be carefully designed to balance knowledge construction and reception of benefits from physical activity. In this circumstance, the cognitive load theory (CLT), a unique extension of the constructivist theories, can serve as a platform for task design.

The CLT focuses on providing relevant types and appropriate amounts of cognitive engagement to elicit optimal cognitive processes to enhance learning. Cognitive load refers to the burden that is imposed by a particular task on the learner’s cognitive system (Sweller et al., 1998). The CLT postulates that intrinsic cognitive load contributes to learning by influencing the learner’s existing conceptual structures (Sweller, 2010). Intrinsic cognitive load defines the magnitude of information based on the complexity of the relationship among knowledge elements required for the understanding of a concept or concepts to be learned. Element interactivity refers to the complexity of a concept in terms of its connections to and dependence on other concepts (Sweller, 2010). Learning tasks with high element interactivity are difficult to understand and yield a high cognitive load because the learner must consider and process several elements simultaneously when learning (Sweller, 2010). In contrast, learning tasks with low element interactivity are easy to understand because they allow individual elements to be learned with minimal reference to other elements (Sweller, 2010).

Therefore, depending on the different element interactivity, modifying content knowledge is required in teaching to manage intrinsic cognitive load for learners (Kalyuga, 2011). For example, learning aerobic and anaerobic concepts is found to be difficult in middle-school PE because of the high element interactivity. Merely teaching the concepts as associated with types of physical activities without other concepts such as intensity, oxygen, and duration is likely to lead to cognitive confusion and shallow understanding (Deng, Zhang, & Chen, 2021). In curriculum design and learning task preparation, teachers may avoid requiring simultaneous processing of many components (high element interactivity) before students are ready; instead, they should provide fewer well-sequenced components (low element interactivity) to enhance learning effectiveness. Teachers can also use careful scaffolding to control the amount and types of element interactivity to adjust cognitive load to facilitate learning effectiveness. Merely asking random and complex questions about fitness and exercise may not be helpful for effective learning because randomly searching answers that are related to multiple concepts can be a high element interactivity task for the learner. The question may sound “challenging,” but may not be helpful in facilitating a coherent understanding. On the other hand, teaching a concept such as aerobic and anaerobic exercise without connecting it to required elements such as intensity and oxygenation may not be productive either (Deng, Zhang, & Chen, 2021).

In summary, the intrinsic and germane (productive) and extraneous (unproductive) cognitive load are defined through the relevance or irrelevance, respectively, of the corresponding cognitive processes to achieving the learning goal. Moreover, the relevance is determined by the three implication factors: content, the learner and instructional strategies, and their interactions. The CLT further suggests that the curriculum and instruction designs should comply with the nature of the cognitive load imposed on the learner and control the load to provide the optimal element interactivity.

Factors Influencing Student Learning Achievement in PE

Research on knowledge learning in PE often regards socioeconomic status (SES) and school environment as confounding factors. The free and reduced-price meal (FARM) rates have been proposed as a widely used indicator of the SES at the school level. It has been argued that students’ learning achievement is influenced by SES due to insufficient resources in low SES (high FARM rate) schools, such as large student-to-teacher ratio, limited space and facilities, outdated equipment, insufficient curriculum materials, and lack of professional development opportunities (Zhang et al., 2021). Student-to-teacher ratio is generally associated with class size, and it is generally argued that smaller classes provide better teaching and learning (Koc & Celik, 2015).

Research findings have been mixed about the association between SES-related school environment factors and students’ knowledge learning outcomes in PE. For example, Zhu and Haegele (2018) found that students from low SES schools were more likely to have a lower growth rate in learning health-related fitness knowledge. Extra efforts may be required to address the SES barriers in disadvantaged areas to enhance students’ learning. However, Zhang et al. (2021) identified that school SES-related class environmental factors may not have an adverse effect on middle-school students’ learning in a constructivist PE context. The findings suggested that the power of the constructivist approach to PE enables students to overcome SES-related barriers to gain necessary knowledge in the PE environment.

The Present Study

Concept-based PE is characterized by a curriculum that relies on knowledge development to guide and achieve physical activity engagement (Corbin, 2021). Concept-based PE has always been challenged by the conflict between prioritizing cognitive or physical engagement in limited lesson time (Chen et al., 2007). One way to minimize the conflict is by following the principles in the CLT to teach both knowledge and physical activity by integrating them together (Ennis, 2015). These curricula use physical activity as a vehicle to teach health-related fitness knowledge. Although the concept of this curriculum design has been examined (Ennis, 2015), the specific role of cognitive load imposed on the learning experiences is yet to be empirically determined.

Guided by the CLT, this study was designed to address the following research questions: (a) Did the cognitive load impact ninth graders’ knowledge gain about caloric-balance and healthful living concepts? (b) Did school SES-related school environmental factors influence ninth graders’ knowledge gain in a concept-based PE context? Answering these questions will further our understanding about the ways to develop content knowledge and learning tasks to promote ninth graders’ systematic knowledge construction. The findings of this study also relate to the time constraint that PE faces now. The findings will assist curriculum designers and PE teachers in making curriculum decisions regarding what and how to apply pedagogical implication of CLT in PE lessons to provide maximum positive impact on ninth-graders’ knowledge gain during the limited PE class time.

The study is based on two primary assumptions. First, it is assumed that the knowledge about physical activity shares the same characteristics as the knowledge in other subject matters. The body of knowledge about physical activity and healthy lifestyles consists of various concepts that are characterized by the element interactivity. The cognitive content can be manipulated and organized for optimal processing. Second, physical activity is a unique component of knowledge learning in PE. Physical activity tasks can be understood as part of the knowledge repertoire, and the link between physical activity and the knowledge about it can also be understood as a unique element interactivity. This link makes it possible to examine the relationship between knowledge and physical activity.

Methods

Research Design

This study was a part of a large-scale high-school PE curriculum intervention project using a randomized placebo-controlled trial to determine the efficacy of the intervention curriculum. The intervention curriculum included 20 lessons that taught the concepts and principles of caloric-balanced living through knowledge and activity integrated learning tasks to ninth graders in five schools in the experimental condition. These five schools have different instruction schedules. Some schools had PE 3 days a week and health education 2 days a week; others ran an alternating schedule for PE and health by several weeks; yet others used alternating A-day and B-day or alternating A-week and B-week schedule to offer PE every other day or every other week. Most split the ninth-grade students into two cohorts, one received PE in one semester, the other cohort the next semester. The true randomization design afforded an opportunity to test the intervention curriculum in diverse contexts, by teachers with different experiences, with extremely diverse students. The participating schools were sampled with stratification on school SES and performances on state standardized science tests. Thus, the school sample represented those with high, average, and low SES and high, average, and low academic performance.

The Intervention Curriculum

The intervention curriculum includes two units. Unit 1, “i-Diet and i-Exercise” focuses on basic scientific concepts about energy sources and pathways in the human body, major concepts of nutrition and exercise, and the relationship between these concepts in sedentary or active states of the human body. These concepts are introduced and reinforced through many cognitive-physical integrated learning tasks. Unit 2, “myHealth, myWay,” focuses on behavioral science to demonstrate the connection between human physical activity and body biochemical reaction, which results in energy-balance/imbalance and health consequences. The table of contents of the curriculum is presented in Table 1.

Table 1.

Table of Contents of the Curriculum

Lesson	Units	Topic
1	i-Diet and i-Exercise	Introduction to the anabolism—glucose
2	i-Diet and i-Exercise	Introduction to the anabolism—glycerol
3	i-Diet and i-Exercise	Introduction to the anabolism—amino acids
4	i-Diet and i-Exercise	Measurement calorie
5	i-Diet and i-Exercise	Energy yielding pathways/releasing—carbohydrates
6	i-Diet and i-Exercise	Energy yielding pathways/releasing—fat
7	i-Diet and i-Exercise	Energy yielding pathways/releasing—protein
8	i-Diet and i-Exercise	PA intensity and energy supply relation of anaerobic exercise
9	i-Diet and i-Exercise	PA intensity and energy supply relation of aerobic exercise
10	i-Diet and i-Exercise	Introduction to caloric balance
11	myHealth, myWay	Introduction to the simple carbohydrates
12	myHealth, myWay	Introduction to the complex carbohydrates
13	myHealth, myWay	Excessive sugar and consequences
14	myHealth, myWay	High-fat diet-related health risks
15	myHealth, myWay	Applying exercise to reduce body fat
16	myHealth, myWay	Saturated and trans fat
17	myHealth, myWay	Thermal balance as part of energy balance
18	myHealth, myWay	Heat injury prevention
19	myHealth, myWay	Eat smart and move more
20	myHealth, myWay	Create an energy-balance living plan

Note. PA = physical activity.

Each lesson is implemented using a “5-E” instructional model: Engagement, Exploration, Explanation, Elaboration, and Evaluation—for students to assume the role of “Junior Scientists” (Bybee et al., 1989). In Engagement, the teacher involves students in some brief physical activities which help them to participate in learning a new concept or principle (Bybee et al., 1989). Students are asked to record their preactivity heart rate or other measurements in their workbook. During Exploration, the teacher or learning task provides the student with a variety of physical activities to collect postactivity physiological and psychological responses to compare with the preactivity measurements. During Explanation, the teacher guides students’ attention to a specific aspect of their Engagement and Exploration experiences and offers them the opportunity to demonstrate their conceptual understanding and behavior changes (Bybee et al., 1989). During Elaboration, the teacher elaborates on the relationship between different concepts and principles and then guides the students to discuss the implications of the concepts and principles in their daily lives. During Evaluation, the teacher leads students to assess their understanding and abilities and evaluates student progress toward achieving the educational objectives (Bybee et al., 1989).

Another salient characteristic of the curriculum is that students are equipped with a workbook in each lesson. The workbook contains content closely tied to the physical activities in each lesson and serves as a centerpiece of knowledge conceptualization tool to assist learning (Zhu et al., 2009). The cognitive learning tasks in the workbook are organized as descriptive, relational, and reasoning learning tasks. These learning tasks help curriculum designers and physical educators to identify the element interactivity levels associated with the three types of cognitive learning tasks as relevant to the physical activity tasks.

Teacher Professional Development

The intervention curriculum teachers received a 3-day of professional development training at the beginning of each semester. Additional 4-hour workshops were provided in the middle and at the end of instruction. The initial professional development was focused on introducing the intervention curriculum, discussing challenges related to cognitive attributes (values, beliefs, intentions, and attitudes) that had been identified in the previous research as potential barriers to teaching the lessons. First, teachers were given both lectures and engaged in discussions to understand the rationale and necessity for incorporating health-related fitness knowledge in PE. Then, the teachers were oriented to scripted lesson plans, studied learning tasks, and peer-taught the lessons. They followed the 5-E instructional cycle, prepared the equipment and space, peer taught the activities, completed the student workbooks using various management strategies, and reflected on the advantages and disadvantages of each teaching strategies. Finally, the teachers learned about how students’ learning achievement would be assessed and what they were allowed or not allowed to do to prepare students for knowledge and physical activity assessments. A previous study showed that the professional development format was most effective in enhancing student knowledge gain (Deng, Zhang, Wang, & Chen, 2021). Throughout the subsequent curriculum implementation and data collection, the research team verified implementation fidelity of the lessons taught using standardized lesson observation form. A check list was used to monitor fidelity throughout the intervention. Supplementary Material 1 (available online) shows an example of the observation guide developed specifically for this purpose. According to the purpose of this study, the data from the intervention condition were used because the comparison condition did not provide students with cognitive learning tasks.

Participants

A total of 385 students received the intervention curriculum and completed the entire workbook learning tasks. We estimated correlation size of similar studies to guide the determination of the sample size. The correlation coefficients between the descriptive, relational, reasoning tasks and knowledge gain from Zhu et al. (2009) and Wang et al. (2019) were averaged as the estimation of the correlation coefficient. The average of coefficients was .40. According to Gall et al.’s (2007) calculation and recommendation, the minimum sample size would be 42 observations in order to satisfy the potential correlation of .40 at a p value of .01. However, the range of the correlation coefficients from these studies is from .19 to .60. Taking this range and the fact that the coefficients were from only two studies, we determined that a sample size of 150 students was needed to accommodate the possibility of the lower-bound correlation (.19). A random sample 150 students from five high schools were drawn by using the random number tables.

These students received the intervention curriculum and completed the entire workbook learning tasks. The sample was selected to represent 6,560 students in these five schools. The sample consisted of 72 (48.0%) boys and 78 (52.0%) girls. The ethnicity composition of this sample was that 76 (50.7%) students were White, 38 (25.3%) Black, 20 (13.3%) Hispanic, six (4%) Asian/Pacific Islander, two (1.4%) Arabic American, and eight (5.3%) mixed races. The five schools represented a diverse student population based on four key variables: race/ethnicity, ratio of students eligible for the free and reduced lunch program, school size, and student-to-teacher ratio. Table 2 illustrates the detailed information of the five schools. The data collection protocols and data protection procedures, along with parent consent forms and student assent forms, were approved by the Internal Review Board of the University of North Carolina.

Table 2.

Basic Demographic Information of Five Selected Schools

	School A	School B	School C	School D	School E
Enrollment	719	1,588	1,299	2,124	830
Race/ethnicity (%)
White	434 (60.4%)	676 (42.6%)	3 (0.03%)	1,610 (75.8%)	622 (74.9%)
Hispanic	270 (37.6%)	173 (10.9%)	402 (31.0%)	149 (7%)	143 (17.2%)
African American	1 (0.01%)	510 (32.1%)	659 (50.8%)	149 (7%)	35 (4.2%)
Others	14 (1.99%)	229 (14.4%)	235 (18.17%)	216 (10.2%)	30 (3.7%)
FARM%	52.7%	33.6%	99.7%	0.09%	43.5%
Students per teacher	15.56	18.11	15.32	20.71	15.64

Note. FARM = free and reduced-price meal.

Variables and Measurements

To address the research questions, data were collected on students’ knowledge gain, cognitive load, FARM rates, and student-to-teacher ratio. The dependent variable was operationalized as students’ knowledge gain scores; the two independent variables were cognitive load and SES-related school environment variables.

Student Knowledge Gain

Student knowledge gain was measured separately in each of the curriculum units using the preinstruction and postinstruction knowledge tests. These items were selected from the question bank (109 questions) throughout the domains that were validated by researchers during the first year of the intervention. The content validity of the preinstruction and postinstruction knowledge tests were determined by experts (n = 5) with at least a master’s degree in exercise physiology and PE. The experts were all asked to rate each question on a 5-point scale for knowledge accuracy (1 = inaccurate, 5 = accurate) and language appropriateness for high-school students (1 = inappropriate, 5 = appropriate). Questions were discussed, revised, and re-rated if one or more experts rated below five on either scale. The content validation process continued until all the questions were rated five by all experts. Questions meet the standards of the difficulty index (.46–.60) and discrimination index (>.40) criteria (Morrow et al., 2005). Each knowledge test consisted of 32 questions; they were selected using the same randomized procedure with adjustments for domain and lesson representation. An example of a knowledge test question was: Which of the following foods contains healthy forms of carbohydrate (Lesson 2)? (a) Bananas, (b) Sweet Potatoes, (c) Blueberries, and (d) All of the above.

Cognitive Load

The cognitive load was measured using students’ responses (performance scores) to 65 cognitive learning tasks. Unit 1 included 11 descriptive learning tasks, 12 relational tasks, and 11 reasoning tasks; Unit 2 included 10 descriptive tasks, 12 relational tasks, and nine reasoning tasks. The learning tasks represented three cognitive levels (loads): descriptive tasks represent low cognitive load because of low element interactivity, relational tasks are of intermediate load with moderate element interactivity, and reasoning tasks are of high load with relatively high element interactivity. The descriptive learning tasks required students to work on factual information such as what physical activity they do in the class or what happens to their body following a physical activity. One example of a descriptive learning task was: “list 10 examples of carbohydrates (Lesson 2, Learning task 1, Unit 1).” The relational learning tasks asked students to relate their bodily responses to physical activities to their understanding of the nutrition and physical activity principles learned. One example of a relational task was: “choose a characteristic(s) compatible with the activity you did in Activity, Pathway and Byproduct [a game]. Circle at least one compatible activity in a similar activities box (Lesson 13, Learning task 1, Unit 2).” The reasoning learning tasks asked students to demonstrate conceptual understanding about why caloric intake and expenditure and physical activities will benefit health and well-being. The reasoning learning tasks led students to establish specific connections between their actions/behavior in the lessons and their applications to life. One example of a reasoning task was: “create 3 tips for avoiding saturated fat and trans-fat in your diet (Lesson 4, Learning task 4, Unit 1).” Supplementary Material 2 (available online) shows an example of a workbook assignment and the corresponding rubric.

SES and School Environment Predictors

School SES was the school average FARM rates obtained from the state education agency. School environment predictor was the student-to-teacher ratio, which was the average number of students a teacher instructs in a class at the school (Graue et al., 2009). Classes in the same school were assigned the same FARM rates and student-to-teacher ratio.

Data Collection

Data were collected in a planned sequence. First, the preinstruction knowledge test and two postinstruction knowledge tests were delivered through an online survey tool, Qualtrics (Provo, UT). The preinstruction knowledge test was delivered before the curriculum implementation, the postinstruction knowledge test for Unit 1 was delivered after the teacher finished Lesson 10, and the postinstruction knowledge test for Unit 2 was delivered after the teacher finished Lesson 20. Second, the cognitive learning tasks were used by students in each lesson and were collected by the research team for scoring when the instruction ended. In the subsequent months, the researchers scored all student workbooks using the validated grading criteria and scoring rubric.

Data Reduction

Trained researchers scored the cognitive learning tasks using the validated grading criteria and scoring rubrics. Using the group Adelphi method, seven content experts worked together to develop the grading criteria. They reviewed the learning objectives for each lesson, made a list of criteria for the most successful performance for each learning task, grouped and labeled them, created grading checklists, discussed, and eliminated any criteria that were not critical for evaluating student performance. They then discussed and revised the criteria several times until they reached a consensus on scoring. The scores were aggregated by the descriptive, relational, and reasoning categories by applying the following formula to each category to arrive at a performance score in the cognitive category: Performance score = Total scores earned ÷ The number of tasks. Subsequently, each student received an aggregated performance score for each level of cognitive learning task.

For the knowledge tests, each correct answer was coded as 1 and incorrect as 0. Students’ knowledge percent-correct total score was calculated for each test. The regression-residual adjustment method (Tracy & Rankin, 1967) was used to calculate the knowledge gain score. The calculation is based on regressing all students’ preinstruction knowledge test scores on the postinstruction knowledge test scores. Each student’s residual gain score was calculated as their actual postinstruction knowledge test score minus their predicted postinstruction knowledge test score (Tracy & Rankin, 1967).

Data Analysis

To answer the research questions, a path analysis was conducted to determine the effect of the cognitive load embedded in the learning tasks of different cognitive levels (descriptive, relational, and reasoning) that accounted for the knowledge gain. It also addressed the impact of the school SES-related school environmental factors on ninth-graders’ knowledge gain in a concept-based PE environment. The path analysis was used to determine the tenability of a hypothesized mode delineated in Figure 1. According to Kline (2011), to test the model fit, the following indices and standards were used: chi-squared (p > .05), root mean square error of approximation (RMSEA; <.08), comparative-fit index (CFI; >.90), and standardized root mean square residual (SRMR; <.08). The path analysis was conducted using IBM SPSS Amos (version 26.0; IBM Corp., Armonk, NY).

Figure 1 —

The a priori model for Question 1 (a’–h’ refer to path coefficients).

Results

In the preliminary analysis of implementation fidelity, percentages scores were summarized to determine possible deviation from the lesson plans. Each checklist item of the observation guide was cross-checked with the lesson plan to determine whether major learning activities were omitted by the teacher. Across all observed lessons at each school, the teachers consistently followed the lesson plans by using the materials in the teacher resources (100%), stating the essential question (100%), emphasizing the science vocabulary, following the timing of the lesson (100%), and utilizing the equipment prescribed (provided by the research team) for each lesson (98%). The preliminary analysis also included calculation of the descriptive statistics for all the variables and determination of data distribution patterns. The descriptive statistics are included in Table 3. The skewness indices ranged between −2 and 2 and the kurtosis indices were between −7 and 7, indicating no violation of the univariate normality assumption (Kline, 2011). Then, path analysis was conducted to address the research questions.

Table 3.

Descriptive Statistics for All Variables

Variable	M	SD	Skew	Kurt
Unit 1: Knowledge gain	0.00	1.00	−0.44	0.13
Unit 1: Descriptive learning tasks scores	3.52	0.40	−0.154	3.87
Unit 1: Relational learning tasks scores	3.03	0.51	−1.06	1.65
Unit 1: Reasoning learning task scores	2.84	0.62	−0.10	0.16
Unit 2: Knowledge gain	0.00	1.00	−0.64	1.36
Unit 2: Descriptive learning tasks scores	3.01	0.52	−0.99	3.21
Unit 2: Relational learning tasks scores	2.87	0.83	−0.72	−0.21
Unit 2: Reasoning learning task scores	3.21	0.72	−1.87	4.31

Note. Skew = skewness; Kurt = kurtosis; Unit 1 = i-Diet and i-Exercise unit; Unit 2 = myHealth, myWay unit.

Results for Learning Unit 1

The standardized path coefficients shown in Figure 2 indicate that the reasoning learning tasks had significant direct effects on students’ knowledge gain in the i-Diet and i-Exercise unit (path coefficient = .34, p < .01). The descriptive learning tasks and relational learning tasks did not have significant direct effects on students’ knowledge gain in the i-Diet and i-Exercise unit (path coefficient = .03, p = .87; path coefficient = .04, p = .90, respectively). However, the students’ performances in the descriptive learning tasks contributed to their performances in the relational learning tasks (path coefficient = .09, p < .01), and the students’ performances in the relational learning tasks contributed to their performances in the reasoning learning tasks (path coefficient = .08, p < .05). The FARM rates (path coefficient = .01, p = .28) and student-to-teacher ratio (path coefficient = .02, p = .80) did not have significant direct effects on students’ knowledge gain. The model fit results show that this model fits the data very well: χ² = 4.80, df = 4, p = .25; RMSEA = .034; CFI = .98; SRMR = .035.

Figure 2 —

The a priori model for Question 1 Unit 1. Note. The path analysis results of the a priori model. Unit 1 = i-Diet and i-Exercise unit. *p < .05. **p < .01.

Results for Learning Unit 2

The standardized path coefficients shown in Figure 3 also indicate that reasoning learning tasks had significant direct effects on students’ knowledge gain in the myHealth, myWay unit (path coefficient = .39, p < .01). Descriptive learning tasks and relational learning tasks did not have significant direct effects on students’ knowledge gain (path coefficient = .12, p = .46; path coefficient = −.07, p = .48, respectively). However, the students’ performances in the descriptive learning tasks (path coefficient = .14, p < .01) and relational learning tasks (path coefficient = .11, p < .05) significantly contributed to their performances in the reasoning learning tasks. The FARM rates (path coefficient = .01, p = .14) and student-to-teacher ratio (path coefficient = −.09, p = .45) did not have significant direct effects on students’ knowledge gain. The results also show that this model fits the data very well: χ² = 5.40, df = 4, p = .23; RMSEA = .043; CFI = .97; SRMR = .048.

Figure 3 —

The a priori model for Question 1 Unit 2. Note. The path analysis results of the a priori model. FARM = free and reduced-price meal; Unit 2 = myHealth, myWay unit. *p < .05. **p < .01.

Discussion

The purpose of this study was to address two research questions: (a) Did the cognitive load impact ninth-graders’ knowledge gain about energy-balance and healthful living concepts in a concept-based PE curriculum? and (b) Did school SES-related school environmental factors influence ninth graders’ knowledge gain in a concept-based PE context? The results show that the students’ performances in the descriptive learning tasks and relational learning tasks contributed to their performances in the reasoning tasks; the reasoning tasks had significant direct effects on the students’ knowledge gain in both the i-Diet and i-Exercise and myHealth, myWay units. Also, the FARM rate and student-to-teacher ratio factors were not significant predictors on students’ knowledge gain in both units.

The Nature of Learning Tasks

The results indicate the low cognitive load tasks, such as descriptive learning tasks, serve as necessary building blocks for completing high cognitive load tasks (Wang et al., 2019). The level of cognitive learning tasks and student engagement in learning tasks interact with each other to influence students’ learning outcomes (Wang et al., 2019). The data showed that the students’ performances in the low and intermediate cognitive load tasks, such as descriptive learning tasks and relational learning tasks, did contribute to the students’ performances in the reasoning learning tasks. According to CLT, high cognitive load tasks (reasoning learning tasks), which include higher order thinking skills, have a higher inherent difficulty than low cognitive load tasks (descriptive learning tasks) and intermediate load tasks (relational learning tasks). Although the learning tasks with high element interactivity are difficult to understand, the only way to foster understanding is to develop cognitive schemata through lower level cognitive learning tasks that incorporate the interacting elements (Sweller, 2020). Therefore, the results seem to be consistent with previous findings (e.g., Wang et al., 2019) to suggest that curriculum designers or teachers should incorporate both low cognitive load tasks and high cognitive load tasks but should not overrate the value of low-level cognitive learning tasks.

High Cognitive Load Tasks and Knowledge Gain

The intrinsic cognitive load in reasoning learning tasks has been considered greater than in descriptive and relational learning tasks because reasoning learning tasks consist of high element interactivity (multiple components) that must be processed simultaneously in working memory in order to be understood (Sweller, 2020). When engaged in high cognitive load tasks, students seem to rely upon their organizational skills and the ability to pull knowledge components needed to make sense out of the problem and apply problem-solving and reasoning skills to address the assignments (Sweller, 2010). The role of high cognitive load tasks is not only to provide a guide to the scaffold of knowledge components with a simple-to-complex sequencing, but also to help the learner reorganize prior knowledge with those knowledge components to be learned in a problem-solving situation to elicit possible physical activity solutions. In other words, the essential information presented to students in high element interactivity tasks can be critical from a cognitive load perspective in that it is a determinant for successful learning (Plass et al., 2010). The results clearly lend support to the theoretical tenets by showing that the students’ performances in the high cognitive load tasks (reasoning learning tasks) directly contributed to their knowledge gain in both units. The students’ high cognitive load task performances may derive from the instructional component that is built into the intervention curriculum. The intervention curriculum was organized using spiral sequencing to help students acquire and store essential concepts in long-term memory (Harden, 1999). The spiral sequencing takes advantage of the positive effect of element interactivity to allow students to visit and revisit concepts in interactive tasks to connect knowledge elements (Harden, 1999). The 5-E instructional approaches might help scaffold the learning experiences and incorporate cognitive learning tasks in every lesson to impose intrinsic cognitive load to maximize students’ knowledge learning outcomes. The overall learning objective was presented first so that learners could construct a schema to be used throughout the tasks, whereas specific real-life problems were presented at particular points of time when it was required.

Instructional strategies and content knowledge based on CLT aim to manage intrinsic cognitive load to improve learning (Plass et al., 2010). First, the student must be supported in the meaning-making process in that they are making sense of the concepts, definitions, and principles that they is confronted with (Plass et al., 2010). The intervention curriculum design includes relevant cognitive load principles, such as the multiactivity principle, the split attention principle, and the simple-to-complex principle, to facilitate the meaning-making process (Plass et al., 2010). The multiactivity principle recommends replacing singular teaching sources (e.g., visual information; written explanatory text) with multiple teaching sources (e.g., visual and auditory tasks). The intervention curriculum teaches students to learn energy-balanced living through a combination of cognitive learning tasks and physical activity tasks. The split attention principle recommends teachers provide integrated information for students to visit and revisit. The intervention curriculum provides necessary knowledge in the posters and workbook information pages, as well as teacher instruction, so students can visit and revisit critical energy-balanced knowledge in each lesson. The simple-to-complex principle recommends replacing a series of complex tasks with tasks that first present only isolated elements (low element interactivity) and gradually work up to high element interactivity. The intervention curriculum was organized using spiral sequencing to follow the simple-to-complex principle.

Second, the student must be supported in their use of selected information for improving their own learning performance (Sweller, 2010). The high cognitive load tasks require students’ ability to monitor their own learning and the ability to control or regulate their own behavior based on interacting knowledge structures. The data appear to suggest that high cognitive load tasks should be emphasized in the high-school concept-based PE curriculum because of their significant contribution to knowledge gain. It has been argued that it is sometimes necessary to increase intrinsic cognitive load in order to promote conceptual change (Sweller, 2020). The high cognitive load tasks help learners to construct and develop high-quality cognitive schemata. The learners with high-quality schemata have better performances in the reasoning learning tasks and knowledge tests because well-learned or automated solution procedures chunk element interactivity together. This indicates that the quality of available problem-solving schema (high cognitive load tasks) facilitates conceptual change and knowledge learning. The design of the intervention curriculum that follows CLT principles enables the students to gain necessary knowledge in the PE experience.

The Power of the Curriculum

The powerful effect of family income and SES has been recognized in the education research literature as a dominant factor of student achievement (Alderson, 2020). Students from low-income and minority backgrounds are attending schools for underserved students because of low education spending, access to educational resources, residential housing patterns, and discrimination (Alderson, 2020). The schools for underserved students are facing additional challenges with various school barriers such as high teacher burnout, lacking facilities and equipment, and high student-to-teacher ratio (Zhang et al., 2021). These challenges in schools that serve low SES communities may drive the learning achievement gap between students from high and low SES schools (Alderson, 2020).

Research findings suggest that school curriculum plays a major role in that it can help reduce educational inequalities and the achievement gap (Cohen & Mckay, 2020). An effective curriculum should contain relevant knowledge for meaning making. The intervention curriculum was designed to translate complex scientific concepts, research procedures, and data analytical skills into personally meaningful, hands-on activity tasks to all ninth-graders regardless of their social demographics or SES status. The results of this study showed that FARM rates and number of students per teacher predictors did not significantly explain the amount of variance in the overall model of knowledge gain in a concept-based PE curriculum. The findings seem to suggest that the concept-based PE curriculum provided rigorous knowledge, used constructivist model of learning, and demonstrated immediate and meaningful outcomes to enhance learning for all students regardless of their interests, abilities, or SES (Cohen & Mckay, 2020).

The findings seem to be clear about the positive impact of the student workbook in PE. Although cognitive content has been taught in PE lessons frequently, using a systematically designed workbook to engage students constantly during a lesson is a rare pedagogical practice. Integrating cognitive tasks with physical tasks appeared to be effective in providing students with mind-body integrated embodiment experiences and helping them develop a sense of meaningfulness throughout the learning process. It is logical to speculate that the integrated, personally meaningful embodiment experiences might have overcome possible SES barriers and enabled all the students to learn energy-balanced knowledge. It can be reasoned that the concept-based PE curriculum has the potential to optimize students’ knowledge learning despite the SES-related school environment factors.

Limitation of the Study

Although the study addresses the important issue of cognitive learning in a physically active context, it only extended our knowledge in the perspective of cognitive load which is tied closely to constructivism, one of the three major tenets in the constructivist theory family. The findings are limited in terms of the influence of necessary social interaction on learning knowledge and the specific conception development with mental model changes. Examining these issues would also have profound significance in advancing the understanding of learning in PE and the relation between cognitive content and physical activity engagement.

In addition, the study was part of a large curriculum intervention research that provided rich resources in curriculum development, teacher training, equipment, and assessment assistance. These resources are severely lacking for PE in most schools (Zhang, 2021). The findings should be interpreted with an understanding that replicating the study in schools without such resources may generate different findings. In other words, the sustainability of the findings can be limited by available resources in individual schools, which limits the generalizability of the findings.

Conclusion

This study is based on a general theoretical framework of CLT to understand and study the effects of cognitive load on learning in PE. The findings provide pedagogical implications on future PE curriculum development. CLT informs curriculum development and task design in the concept-based PE curriculum as a specific platform based on which curricular and instructional decisions can be made to enhance the learner capacity of processing cognitive information in a physically active setting. The results show that knowledge and activity integrated learning tasks are an effective way to provide a “mind-body” integrated embodiment experience to high-school students. The findings indicate that the intervention curriculum may provide an optimized germane cognitive load and minimize extraneous cognitive load (Sweller, 2020), which affirms the positive function of the three-level scaffolding structure of the tasks.

The research findings would contribute to the refinement of the current approaches to cognitive learning and teaching in PE. First, it is crucial to manage intrinsic cognitive load by presenting information at a level of difficulty that is appropriate to high-school students’ level of understanding. More importantly, implementing well-designed high cognitive load tasks in high-school PE class can be a key element to facilitate learning outcomes. The findings are meaningful in that it shows the critical value of high cognitive load tasks to students’ knowledge gain.

Second, from the CLT perspective, instruction is a process to impose extrinsic cognitive load to the learner. A well-designed instructional model should minimize extraneous cognitive load and optimize germane cognitive load to improve students’ knowledge learning (Plass et al., 2010). One challenge for physical educators is to choose and prioritize the content: scientific concepts or physical activity (Deng, Zhang, Wang, & Chen, 2021). To address the challenge, it seems that teachers can use the “5-E” instructional model to incorporate cognitive learning tasks in every lesson to balance and integrate cognitive and physical demands. The evidence seems to suggest that the “5-E” instructional model is able to reduce the amount of extraneous cognitive load and promote germane cognitive load.

Third, the intervention curriculum is a powerful concept-based PE curriculum in overcoming potential barriers stemming out of a diverse instructional context. Consistent with the previous findings (Zhang et al., 2021), the findings continue to show the power of the well-structured concept-based PE curriculum in developing students’ health-related fitness knowledge. The findings will assist PE teachers in making curriculum decisions to provide both cognitive and physical benefits for all students in high-school PE class.