Education is the most powerful weapon which you can use to change the world.

                                       Nelson Mandela

WELCOME

This is a Going Inductive learning educational strategy changing paper DRAFT roadmap¹, with two major goals:

1. Conclude about the impact on students learning of a strategic learning methodological major change, from direct instruction to inductive;
2. to share the work done about the impact of a educational strategic change: from Direct Instruction to Four Components Instructional Design (4C/ID).

This knowledge base source contains the results, conclusions, procedures, exploratory data analysis (EDA), data, statistical methods, educational methodologies, learning flow and bLearning practice. Hopefully, there will be, at least, one paper based on this.

TO DO LIST

• EDIT EDIT EDIT: final reading: revisit goals , hyphotesis, research questions, discussion and conclusion;
• Verify the conclusion text
• This work is prepared to be edited in a specialized journal;

Going Inductive: what is the impact on students learning?

Author: António Carlos Gonçalves, Teacher at the Portuguese Education Ministery (during this research), https://www.linkedin.com/in/fqantonio/

ABSTRACT

This study investigates the impact of shifting from deductive to inductive teaching methods within a blended learning (bLearning) environment for young science and technology students. While research on specific educational methodologies is extensive, this analysis focuses on a strategic change.

The findings suggest that, within the constraints of our research, a 4C/ID bLearning approach doesn't consistently outperform Direct Instruction (DI) across all outcomes. While we observed a significant positive impact on Lab Practice skills, effects on school success and learning transfer were less pronounced. Interestingly, we found a negative impact on social skills. These differential effects were observed across the student sample, with female students, ninth-graders, and those better adapted to school environments demonstrating greater benefits.

This research highlights the need for careful consideration when implementing such a pedagogical shift. It discusses the specific implications for teachers, policymakers, and school administrators, acknowledging the potential challenges and benefits associated with this change. The analysis is based on 12 years of teacher data (2003-2017), providing insights into the impact on learning transfer, lab practice, social skills, and academic results.

Keywords Inductive learning strategy; instructional Design; 4C/ID; Direct Instruction; Learning Flow; Learning Transfer, Lab Practice, Social Skills, Academic Results; Cognitive Load Theory; MultiIntelligence Theory; Brain Based Learning; bLearning, Statistical Non-parametric Inference.

GOALS, RESEARCH QUESTIONS and HYPOTHESIS

The main goal of this study aims, specifically, to investigate the impact of transitioning from Direct Instruction (DI) (Merrill, 2007) to Van Merriënboer's Inductive Four Component Instructional Design (4C/ID) model (Van Merriënboer, Kirschner, 2007) within a bLearning environment using Moodle as a Learning Management System (LMS) (Cole, 2005; Rice, 2006).

The research examines the impact on four key learning outcomes: Learning Transfer, Lab Practice, Social Skills, and Academic Achievement. To guide this investigation, the following research questions were addressed:

What is the impact on student learning outcomes?
Which student groups benefit most?
What are the advantages and disadvantages of this methodological shift?
What are the implications for stakeholders involved in the educational process?

The hypothesis posits that this transition will have an overall positive impact on student learning outcomes. Specifically, we expect to see improvements in learning transfer, lab practice, social skills, and academic achievement, regardless of the student's gender, grade level, or degree of school adaptation (Prince, Felder 2006). We believe that using the 4C/ID model can create an effective learning design that facilitates schema development and optimizes learning outcomes. This approach can also provide valuable insights for stakeholders, including teachers, policymakers, and school administrators, helping them enhance teaching practices and effectively implement instructional strategies (Kirschner, 2002).

THEORY

As point out by Sweller, van Merriënboer and Paas, "Cognitive load theory emphasised that all novel information first is processed by a capacity and duration limited working memory and then stored in an unlimited long-term memory for later use. Once information is stored in long-term memory, the capacity and duration limits of working memory disappear transforming our ability to function".

This study utilizes Van Merriënboer's 4C/ID instructional design which is grounded in an inductive strategy, CLT and in several robust educational theories, effectively bridging together Cognitive Architecture and Instructional Design. It integrates principles from:

* Neuroscience to Brain-Based Learning (Jensen, 2005; Sylvester, 1995);

* Cognitive Load Theory (CLT) (Plass, Moreno, Brünken, 2010);

* Multiple Intelligences Theory (Gardner, 2011);

* Multimedia Learning Principles (Mayer, 2005);

* Evolutionary Educational Psychology (Sweller, 1998);

* Inductive Methodologies as advocated by Felder (Felder, 1993).

As a reliable framework, the 4C/ID model provides a clear procedure for designing and planning a blended learning (bLearning) environment. Its components are explicitly defined and integrated into a cohesive structure, offering a robust theoretical foundation for instructional designers. This guides the substantial work of developing concrete Learning Management System (LMS) and classroom activities without losing the connection to cognitive and pedagogical knowledge. Ultimately, the model helps to build and develop activities that incrementally guide students in integrating new information into existing knowledge structures (schemas) within their long-term memory. It achieves this by minimizing working memory load through techniques like fading guidance, worked examples, and completion problems, all of which reduce the split-attention effect.

continue reading...

STATE OF THE ART

This work offers a strategic perspective that fills a gap in the existing research. While much of the literature focuses on specific effects (Lim, Reiser, Olina, 2009), which has certainly strengthened the model, it often overlooks the broader practical application and the real-world constraints that teachers face in their day-to-day professional lives.

This work motivation lies on the need to understand the impact of the professional strategic change decision within the educational work developed (flow charts, ²):

Was it worth it? Did students benefit? What can be done better? What conclusions can be drawned?

Most of the time the professional, specific, day-to-day teacher work data (see figure 2) is rarely or never investigated, at least, in the Portuguese educational environment. For the teachers, there is little or no time to look back and work through the data results, the qualitative remarks, or through the statistics. And, as a consequence, no robust conclusions are devised and worse, it's normal not to share it.

Inductive strategy is not widely spread but, in the last years professionals start to look at it as a strategy that could give to the educational environment, no matter which grades or subjects, a different perspective, and research also showed positive effects on learning (Klauer, Phye, 2008)(Kaur, Niwas., 2016). Some, also claim that it can be applyed along with DI. Its important to point out that former metodologies can be adapted to an inductive strategy (Prince, Felder, 2007), simplifying adopting it.

The 4C/ID model itself has being scrutinized in the last 40 years (van Merriënboer, Kirschner, Frèrejean, 2024), as well as the CLT theory's underlying, showing consistency, strategic application, falsifiability, and ability to generate new research. The same to Direct instruction, a much more older and wider model, applying around the world extensively.

Since the 1980s, CLT gave support to several instructional design models. CLT is based in uncontroversial aspects of human cognitive architecture. In the late 1990s, Sweller et al published extensive data using this theory that warranted an extended analysis (Sweller, van Merrienboer, Paas 1998). In 2019, the same authors (Sweller, van Merriënboer, Paas 2019), reviewed the theoretical and empirical work done so far, concluding that CLT, "(1) is firmly based in our—expanding—knowledge of human cognitive architecture; (2) it is under continuous development as our knowledge of human cognition advances; (3) it leads to testable hypotheses with possible negative results leading to modifications of the theory; (4) the vast bulk of the data generated by the theory is based on randomised, controlled trials; and (5) those randomised, controlled trials provide evidence for the effectiveness of instructional procedures that can be used in a wide range of educational contexts from conventional classrooms to e-learning, teaching all age groups from very young to adult learners, with an enormous range of subject matter from medical education to English literature."

continue reading...

PROCEDURE

Implementing the 4C/ID model, as described by Van Merriënboer, Clark, and Croock (2002) as akin to steering a "huge, slow, ponderous ocean liner" required significant effort. The process involved extensive research, testing, and iterative refinement of pedagogical planning. While the 4C/ID model presents a steep learning curve, it provides a valuable framework for consistently managing instructional sequences prior to LMS implementation.

The model emphasizes the crucial link between its underlying theories and its concrete structure, comprising Task Classes (TC), Learning Tasks (LT), Part-Task Practice (PTP), Supportive Information (SI), Procedural Information (PI), and Just-in-Time (JIT) information. This framework prioritizes whole-task learning within real-world problem contexts, facilitating gradual skill acquisition while minimizing cognitive load on working memory.

The 4C/ID model advocates for a progressive approach, moving from simple to complex task classes within each LT, with gradually diminishing levels of support (scaffolding). Sufficient SI is provided for general learning tasks, while JIT information guides students toward automaticity in achieving learning goals. PTP activities are incorporated to train constituent skills to a high level of automaticity, crucial for overcoming working memory limitations in complex learning scenarios (Kalyuga, Paas, et al., 2010).

continue reading...

CONTEXT and DATA

This work is based on real day-to-day teacher assessment data, for junior and junior high students of two different schools, from different regions and time periods, for the chemistry and physics subjects of the Portuguese curriculum.

The data was gathered over 12 years, in the period 2003 to 2017, with a break between 2009 and 2012, in order to study bLearning³, learn and prepare the aplication of the 4C/ID methodology (Van Merriënboer, Kirschner, 2007). Data includes students Learning Transfer, Lab Practice, Classroom Behavior (Social Skills), and Academic Results using four different random variables: TEST, LAB, BEHAV and CLASS. Data as well as R code and cmaptools planning examples are available online at the github https://github.com/fqantonio/GoingInductive.

VARIABLES DESCRIPTION

VARIABLE	Description
ID	Identification entry row data
DATE	first year of the raw sample: example: 2003/2004, lective year, is recorded as 2003
SCHOOL	there are two schools, in different regions, identified by 0 (till 2008) and 1 (after 2009)
GENDER	Female and Male, respectively, F and M
TEST	scale: 0-100: learning transfer, accessed by online/presencial written tests
LAB	scale: 0-100: laboratory practical skills assessed by teacher recording observational forms in the classroom
BEHAV	scale: 0-100: social skills assessed by teacher recording observational forms in the classroom
CLASS	scale: 0 to 100: Learning success, weight average formula [^average]
RANK	categories: 1 to 3: School adaptation 4
GRADE	categories 0 to 6, representing, respectively, grades 7,8,9,10,11, 10p (technical) and 11p (Technical) school grades
M4CID	category 0 and 1, respectively, without 4C/ID and with 4CID;

TIME SERIES

Figure 1 shows the variable longitudinal time patterns, with box plots for variables, TEST, LAB, BEHAV and CLASS: it includes junior and junior high grade students data for both schools.

The red vertical line shows the school change year and the blue one is the starting point for the implementation of 4C/ID strategic inductive methodology, the treatment group. Its clear that near each vertical line there is a change showed by the blue smooth line (polynomial local regression): around 2012, the first year of 4C/ID implementation, so something happened!

The box plots in figure 1 shows, not only that the samples are not normal distributed, but also they are skewed, has same outliers and are not symmetric. Further analysis confirms this: samples are independent, not symmetric, and don't have the same variances and shapes. However, some few sample for the CLASS and TEST variables where normal distributed.

Figure 1: Box plot time series

METHODS

The statistical analysis, made with R code software in the RStudio⁵ IDE release and the principal method is the non-parametric inferential group treatment effect with non-paired sign-rank Wilcoxon procedure with the respective assumptions analysis: samples independence, non-normality distribution, symmetry, shape and variances. Since the majority of the samples don't foloow this assumptions, they have the same variances, don't have the same shapes and symmetry is broken, and, in some cases, there are low observations, < 50, sample permutation test was also used. Sample independence was tested with Kendall procedure.

The same analysis, the same sequence and tests, was done using several samples (figure 2). So, the idea is to suppose that the correct result must be 'in the middle' of this extremes, or at least, to considered the all those results: in one end it was used "all sample" and in the other end, the same sample, but without school and behavior effect (this last one, for the years 2016 and 2017, see figure 1).

RESULTS

Figure 2 and 3 shows the median changes between control and treatment group, for each variables TEST, LAB, BEHAV and CLASS (see legend). Samples are presented on the y-axis. The left of the x-axis present the negative changes and on the right the positive: the horizontal lines means no changes detected.

The idea is to analyse the results considering both extremes results, all sample and school 1 - 2014 sample. The first shows a more unfavorable scenario while the former a better one. So, moving up on the y-axis of figure 2 graph, demonstrates that the results are turning to be more favorable to the educational 4C/ID methodological change. There are two clear results: BEHAV has an overall negative change and LAB and overall positive. Furthermore, CLASS and TEST, both has an overall positive change but in less degree (figure 3).

Introducing the school differences (school effect), the negative change persist in the BEHAV variable. If the sample is reduced until 2014 only, then BEHAV starts to have a "no change"" to a "positive change", and there is a positive change for all educational areas: this means get reed of school effect and behavior effect on the two last years, 2016 and 2017. In this sample, figure 3, the biggest positive change is for the LAB variable, CLASS and TEST are the next ones, while BEHAV has only a positive change in school 1 - 2014 sample.

Figure 2: 4C/ID group treatment effect from 2003 to 2017.

Figure 3 shows that that impact on the median change is clearly positive for the LAB variable followed by the CLASS and TEST variable and negative for the first four samples for BEHAV.

Figure 3: 4C/ID group treatment effect from 2003 to 2017.

Figure 4 and 5 shows a resume of the results, with much more specific samples about GRADE, GENDER or RANK, in the time period 2003 to 2017. Clearly, there is an overall positive change for the LAB variable while a negative change for BEHAV (except for GRADE 9, RANK 3 and FEMALE samples). Similarly for the variable TEST, except for JUNIOR HIGH,GRADE 7 and 8, RANK 1 and 3, and FEMALE, no changes detected. Variable CLASS shows a mixture: in the samples JUNIOR and FEMALE there is a positive change, while negative for RANK 2 and MALE and no change for the other samples. BEHAV and LAB confirm the pattern already identified above. CLASS, has a overall positive change except for RANK 2 and MALE samples while TEST just has negative changes except for MALE, RANK 1 and 2, GRADE 9, JUNIOR and all sample.

didn't benefit	benefit more	Notes
the RANK 1 and males	Females and RANK 3
Grade 8	Grade 9
Junior High	junior	lower level of obs. for junior high sample

Figure 4: 4C/ID group treatment effect from 2003 to 2017.

Figure 5 presents the score change between the two groups, treatment and control. The LAB variables shows the biggest positive change while BEHAV and TEST shows the worst negative change. CLASS as a mixture behavior. Samples groups that present more negative changes are MALE and RANK 2. On the other, FEMALE and RANK 3 show positive change.

Figure 5: 4C/ID group treatment effect variables change for all period of time: 2003 to 2017.

On Figures 6 and 7 shows the analysis with a sample that doesn't include the school and behavior effects. So, there is a more favorable overall positive change between control and treatment group. For RANK 1, MALE there is a less benefit results, as well as for BEHAV variable that continues to show negative change. For GRADE 8 there is a positive overall, except for BEHAV variable. There is a clear positive change for FEMALE, RANK 3 and genders 7 and 9, with RANK2 sowing a positive change on teh LAB variable, and no change on the others variables.

benefit less	benefit more
the RANK 1,2 and males	Females, grade 9 (except on TEST) and RANK 3

Figure 6: Results resume graph, 4C/ID treatment effect for JUNIOR grade sample of school 1 until 2014.

Figure 7 presents the score median change between the groups treatment and control for school 1 junior grade sample until 2014: there is an overall positive change with the LAB variables showing the biggest while BEHAV shows a negative change for GRADE 8, RANK 1 and MALE samples. CLASS shows the second biggest change followed by TEST and BEHAV. So, the results are not drastically different.

Figure 7: Results resume variables change graph, 4C/ID treatment effect for JUNIOR grade sample of school 1 until 2014.

QUALITATIVE RESULTS

The nost importante results more important can be resumed like this:

1. Students show a lot of motivation while working online with a LMS MOODLE organized with game levels, anchors points, instantaneous feedback and activities for automation procedures;
2. The MOODLE environment provide moments of procedural automation;
3. During the online period of work the teacher as more time to support students specific needs;
4. The amount of work of planning, constructing and improving the online lessons is overwhelming for a teacher alone.

CONCLUSION

======= Is this in the conclusion?

The research examines the impact on four key learning outcomes: Learning Transfer, Lab Practice, Social Skills, and Academic Achievement. To guide this investigation, the following research questions were addressed:

What is the impact on student learning outcomes?
Which student groups benefit most?
What are the advantages and disadvantages of this methodological shift?
What are the implications for stakeholders involved in the educational process?

=======

NOTE 1: guidelines: what impact for the students learning outcomes (four)? Who benefits more? What are the pros and cons about this methodological decision? What are the implications for the stakeholders? Is it worth it? Do students benefit? What can be done better? What conclusions can be drawned?

NOTE 2: the conclusions presented here focus on the "mean" results between the less favorable results of figures 2 and 3 and the more favorable of figures 4 and 5.

NOTE 3: the Direct Instruction (Merrill, 2007) instructional design is a well established learning strategy with a overwhelming research background, some authors claim, with 200 years of history.

Before hand, we've to outline that the comparison is made to a well developed, well established and researched instructional design theory for over more then 200 years. Any slight improve is considered very strong.

So, Overall, as a teacher of science and tech, you should (or at least think about it) move to implement the inductive strategy as your methodology for learning, because it has an overall positive impact on the learning outcomes analysed in this work, namely, Learning Transfer, Social Skills and academic results, when the sample is clear of the school and behavior effect.

Clearly, there is a strong positive impact for Lab Practice while some of the analysis show a negative impact for Social Skills. For Academic Results and Learning Transfer there is a less positive impact compared to lab practice.

Students that benefit more are females, grade 7, grade 9, junior high and the more adapted to the school system while, the one's that don't benefit are the less adapted to the school system and males. The other groups, grade 8, overall junior grade benefit less.

Bottom line, if you don't want to take the risks, change your methodology conditional to have a class group more adapted to school or if you want to increase the lab practices skills in some subject or school context, and if you have an engaged team for the bLearning implementation.

4C/ID is the methodology that works very well if the lab practice is a nuclear goal and/or your students are well adapted to school. However, in each moment there is a need for social skills classroom management. Same students claim about the inductive strategy because they like to have a manual to follow learning content in order to organize their work. So, its important to deliver a resume of the classroom content of the day.

On the other end, bLearning has a crucial benefit, it provides more individual time interaction between student and teacher which improves student engagement and motivation. While students are working online, during the classroom, the teacher can reply to specific student questions.

Implementing a bLearning system is a overwhelming task for a teacher alone because there is a huge amount of work to overcome: planning, designing, organizing and delivering the online course; afterwards there is a need to updating it, admnistrating, messaging with students, developing and correcting: team work is crucial.

Another thing that has to be acknowledge by the stakeholders is the fact that this change is overwhelming for one teacher alone: there is a crucial need to a commited, motivated and goal-oriented team work. However, the motivation cannot came from a external decision but must be driven by a inner group necessity.

Furthermore, of course one methodology is not a panacea for all of the learning process, but having an educational theory is very important to develop and adapt to your educational context. There is also a need for stimulating research and spread the knowledge about the specific impacts of each strategy.

DISCUSSION

Overall, inductive methodology has a positive impact on learning, as evidenced by various studies. Inductive methods, which involve learning through specific observations and problem-solving before introducing theories, have been shown to be more effective than traditional deductive methods in many educational contexts. Namely, better academic learning of classroom subject matter where observed. The review and meta-analysis summarizes the results of 74 training experiments with nearly 3,600 children. (Kauler and Phye, 2008). Furthermore, reports show that deductive methodologies contribute for students leaving the sciences and give a strong support for inductive teaching methods that encourage students to adopt a deep approach to learning and are precursors of intellectual development (Prince, Felder, 2007). A 4C/ID learning environment promotes the development of technical expertise in secondary technical education better than direct instruction (Sarfo, Elen, 2007).

This work aligns with research that shows, despite the strength of the evidence varies from one method to another, inductive methods are consistently found to be at least equal to, and in general more effective than, traditional deductive methods for achieving a broad range of learning outcomes (Prince, Felder, 2006). Other studies shows a high impact on performance (d = 0.79 standard deviations) for 4C/ID educational programs, suggesting that is use should be prioritized in college and university learning environments (Costa, Miranda, Melo, 2021). More detailed works show better results for groups that whole-task based in 4C/ID instructional desing on a skill acquisition test and a transfer test (Lim, Reiser, Olina, 2009). Event in computer programming course there is a description of benefits of using inductive (Khan et al., 2020) and also in grammar (Kaur, Niwas, 2016).

There was a big difference regards to gender, female getting more benefits of 4C/ID methodologies and bLearning then males.

Further research should include the develop of the same setting in other educational contexts, national and international, in order to confirm these findings and an analysis of the reason for the behavior declining over 2016. These work should be improved by including inference analysis that take in account samples with no symmetry, difference variances and different distribution shapes. Also, there should be done a regression and cluster analysis as a way to confirm the results and conclusions presented here, as a feedback strategy results confirmation.

Aside from the quantitative results, it was seen that bLearning systems provide the time to the teacher to interact more efectively during the process, tackling the individual problem for each student. So, one possibility for further research is to question the importance of this bLearning in the teaching-student interaction enhancement. Of course, adaptatively is the "holly Graal" of this bLearning systems. Further research on how to adjust the student prior-knowledge and is subject cognitve hability should be usefull to choose the best task class (see Plass et all, 2010, p.127) and cope with the expertise reversal effect. Another point, crucial for 4C/ID instructional, is the procedural automation process, identifyed and implemented in the MOODLE LMS system but not quantified, but it should be a line of research also.

???In my opinion, the ony struggle with this model is to make it more user friendly, for instance, to add a more strategic approach in order to make it moer user friendly. The elarning curve could be very steep..

Acronyms

DI - Direct Instruction

4C/ID - Four Components Instructional Design

LT - Learning Tasks

PTP - Part-task-Practice

JIT - Just in Time

SI - Suppotive Information

TC - Task Class

bLearning - blended learning

EDA - exploratory data analysis

LMS - Learning Management System

Funding

The author received no funding from private or national entities for this work

Intellectual property

I confirm that I have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that I have followed the regulations of my institutions concerning intellectual property.

Research ethics

I further confirm that any aspect of the work covered in this manuscript has been conducted with the ethical principles regarding students and teacher personal data that are never shown nor data is related to names, classes or dates.

Authorship

This text was originally written in English by the author, with revisions by Gemini AI and further review by the author. The author is solely responsible for all discussions, conclusions, procedures, and results presented.

REFERENCES

• Cole, Jason (2005), Using Moodle, O'relly Community Press, ISBN 0-596-00863-5;

• Collopy, Rachel & Arnold, Jackie. (2009). To Blend or Not to Blend: Online and Blended Learning Environments in Undergraduate Teacher Education. Issues in Teacher Education. 18;

• Costa, J., Miranda, G., & Melo, M. (2021). Four-component instructional design (4C/ID) model: a meta-analysis on use and effect. Learning Environments Research, 25, 445 - 463, https://link.springer.com/article/10.1007/s10984-021-09373-y#citeas;

• Coughlin, Cristy (2011); Research Syntheses of Direct Instruction Outcomes, published in Lloyd, J., Carnine D., Slocum, T., Watkins, C. (Eds.) Does Direct Instruction Deserve Status as an Evidence-Based Practice? ADI Press;

• Felder, Richard (1993); Reaching the Second Tier: Learning and Teaching Styles in College Science Education, Journal of College Science Teaching 23(5);

• Gardner, Howard (2011) Frames Of Mind: The Theory Of Multiple Intelligences, ISBN-13: 978-0465024339;

• Garrison, D.R.; Anderson, Terry (2003); E-Learning in the 21st Century, A framework for research and practice, RoutledgePalmer, ISBN: 978-0-415-26346-7;

• Jensen, Eric (2005), Teaching with the Brain in Mind, ASCD, ISBN-10: 146600302; ISBN-13:978-1416600305;

• Kalyuga, S. (2011). Cognitive Load Theory: How Many Types of Load Does It Really Need?. Educational Psychology Review, 23, 1-19. https://doi.org/10.1007/S10648-010-9150-7;

• Kaur, S., & Niwas., D. (2016). EFFECTIVENESS OF DEDUCTIVE AND INDUCTIVE METHODS IN TEACHING ENGLISH GRAMMAR AT ELEMENTARY LEVEL.. International Journal of Approximate Reasoning, 4, 1241-1247. https://doi.org/10.21474/IJAR01/1027;

• Khan, I., Iftikhar, M., Hussain, S., Rehman, A., Gul, N., Jadoon, W., & Nazir, B. (2020). Redesign and validation of a computer programming course using Inductive Teaching Method. PLoS ONE, 15. https://doi.org/10.1371/journal.pone.0233716;

• Kirschner, Paul (2002) Cognitive load theory: implications of cognitive load theory on the design of learning, Learning and Instruction, Volume 12, Issue 1, February 2002, Pages 1-10, https://doi.org/10.1016/S0959-4752(01)00014-7;

• Klauer, Karl Josef; Phye, Gary; (2008) Inductive Reasoning: A Training Approach, Review of Educational Research, Volume 78, Issue 1, https://doi.org/10.3102/0034654307313402;

• Lim, J., Reiser, R.A. & Olina, Z. (2009); The effects of part-task and whole-task instructional approaches on acquisition and transfer of a complex cognitive skill. Education Tech Research Dev 57, 61–77; https://doi.org/10.1007/s11423-007-9085-y;

• Melo, M. (2018). The 4C/ID-Model in Physics Education: Instructional Design of a Digital Learning Environment to Teach Electrical Circuits. International Journal of Instruction, 11(1), 103-122. https://doi.org/10.12973/iji.2018.1118a;

• Mason L, Otero M. (2021); Just How Effective is Direct Instruction? Perspect Behav Sci. 2021 Jun 2;44(2-3):225-244. doi: 10.1007/s40614-021-00295-x. PMID: 34632277; PMCID: PMC8476697;

• Mayer, Richard (editor)(2005); The Cambridge Handbook of Multimedia Learning, Cambridge University Press; ISBN: 0-521-83873-8;

• Merrill, M.D. (2002) First principles of instruction. ETR&D 50, 43–59. https://doi.org/10.1007/BF02505024;

• Merrill, M. David (2007) A Task-Centered Instructional Strategy, Journal of Research on Technology in Education, 40:1, 5-22, DOI: 10.1080/15391523.2007.10782493;

• Plass, Jan; Moreno, Roxana; Brünken, Roland (editors) (2010), Cognitive Load Theory, ISBN: 978-0-521-67758-5, Cambridge university Press;

• Prince, M.; Felder, R. (2006). Inductive Teaching and Learning Methods: Definitions, Comparisons, and Research Bases. Journal of Engineering Education, 95. https://doi.org/10.1002/j.2168-9830.2006.tb00884.x;

• Prince, M.; & Felder, R. (2007); The Many Faces of Inductive Teaching and Learning. The journal of college science teaching, https://www.pfw.edu/offices/celt/pdfs/Inductive(JCST).pdf;

• Rice, William (2006); Moodle, E-Learning Course Development, Packt Publishing, ISBN: 1-904811-29-9;

• Sarfo, F.K., Elen, J. (2007); Developing technical expertise in secondary technical schools: The effect of 4C/ID learning environments. Learning Environ Res 10, 207–221. https://doi.org/10.1007/s10984-007-9031-2;

• Silvester, Robert (1995) "A celebration of the neurons", ASCD, Alexandria, Virginia;

• Stockard, Jean; W. Wood, Timothy; Coughlin, Cristy; Khoury, Rasplica (2018); The Effectiveness of Direct Instruction Curricula: A Meta-Analysis of a Half Century of Research, Review of Educational Research, Vol. 88, No. 4 (August 2018), pp. 479-507 (29 pages), Published By: American Educational Research Association, https://www.jstor.org/stable/45277261;

• Sweller, John; van Merrienboer, Jeroen; Paas, Fred (1998); F.G.W.C. Cognitive Architecture and Instructional Design. Educational Psychology Review 10, 251–296. https://doi.org/10.1023/A:1022193728205;

• Sweller, J.; van Merriënboer, Jeroen; Paas, Fred (2019); Cognitive Architecture and Instructional Design: 20 Years Later. Educ Psychol Rev 31, 261–292. https://doi.org/10.1007/s10648-019-09465-5;

• Sweller, J. (2023); Education needs to embrace the science of learning, Knowledge Society, https://www.youtube.com/watch?v=575i45cJLvM;

• Van Merriënboer, Jeroen J. G.; Clark, Richard; Croock, Marcel (2002) Blueprints for complex learning: The 4C/ID-model. Educational Technology Research and Development. 50. 39-61. 10.1007/BF02504993; https://www.researchgate.net/publication/225798787_Blueprints_for_complex_learning_The_4CID-model;

• Van Merriënboer, Jeroen J. G.; Kirschner, Paul A. (2007) Ten Steps to Complex Learning, Routledge; 1st edition, SBN-10:0805857931; ISBN-13:978-0805857931;

• van Merriënboer, J.J.G., Kirschner, P.A., & Frèrejean, J. (2024). Ten Steps to Complex Learning: A Systematic Approach to Four-Component Instructional Design (4th ed.). Routledge. https://doi.org/10.4324/9781003322481;

NOTES

Footnotes

1. This work, as well as the data, is also presented at the blog "https://4cidchange.edublogs.org/" and in the github repository https://github.com/fqantonio/GoingInductive. Currently, working at a non-profit association, Lab Aberto FAB LAB (https://lababerto.pt/), that promotes the implementation of FAB LAB in schools. ↩

2. cmap tools by ihmc https://cmap.ihmc.us/ ↩

3. EDX learning course, https://www.edx.org/learn/education/edx-blendedx-blended-learning-with-edx ↩

4. RANK1 if CLASS variable is less than 45%; RANK 3 if CLASS variable is more then 70% and RANK 2 for the rest. ↩

5. RStudio, https://posit.co/products/open-source/rstudio/ ↩