DESIGNING
LEAGILE METHODS FOR
IDENTIFICATION AND CLUSTERING DEMAND IN
PERSONALIZED LEARNING
Lean in secondary education
MASTER THESIS
Jon Hummel
s2354381June 23, 2017
Technology & Operations Management
Faculty of Economics & Business
1
Abstract
2
Table of contents
1. Introduction ... 4 1.1. Problem Definition ... 6 1.2. Research Questions ... 6 1.3. Methodology ... 71.4. Outline of this thesis ... 10
2. Educational context ... 11
2.1. Theoretical background about Learners’ demand and supply of activities ... 11
2.2. Empirical findings ... 13
2.3. Merging empirical findings & theory: Factors for demand and supply of activities ... 15
2.4. Design constraints for a design artifact ... 17
2.5. Main conclusion about the educational context ... 19
3. Make-to-order production context and leagile philosophy ... 20
3.1. Theoretical background ... 20
3.2. Observations in make-to-order production environments ... 25
3.3. Main conclusion about the make-to-order production context ... 26
4. Model ... 27
4.1. Overview of design constraints ... 27
4.2. Conceptualization Scrum towards the design artifact ... 27
4.3. Constructing a leagile based design artifact ... 29
4.4. Main conclusion about design artifact ... 33
5. Implementation ... 34
5.1. General overview ... 34
5.2. First phase: identification ... 36
5.3. Second phase: clustering ... 37
5.4. Third phase: planning ... 39
5.5. Conclusion... 41
6. Validation ... 42
6.1. Validation: Picasso Lyceum ... 42
6.2. Generalization: Het Lyceum Vos ... 43
6.3. Generalization: Het College Vos ... 44
6.4. Conclusion... 45
7. Conclusion ... 46
7.1. Answering the research questions ... 46
7.3. Limitations research ... 47
7.4. Further research ... 47
References ... 49
3
Preface
(English translation below)
Het schrijven van dit voorwoord kan niet op een betere plaats dan in Rotterdam, het logistieke hart van ons land. De afgelopen maanden heb ik het voorrecht gehad om in de omgeving van Rotterdam logistieke kennis te verbinden met de onderwijsinstellingen in deze regio. Het is een inspirerende afsluiting van mijn studietijd geweest. Graag bedank ik prof. dr. Iris F.A. Vis voor alle begeleiding, adviezen en feedback die ik op een zeer prettige en constructieve manier de afgelopen periode heb mogen ontvangen. In drukke agenda’s werd er altijd tijd vrij gemaakt en naast de al geplande momenten, werd er door haar de moeite genomen voor Skypegesprekken over deze scriptie vanuit allerlei plekken over de wereld. Daarnaast wil ik graag prof. dr. Jan Riezebos bedanken voor de hulp en inspiratie op het gebied van het onderwerp en de literatuur binnen deze scriptie. Het samen brainstormen over dit onderwerp heeft veel bijgedragen aan deze scriptie. Deze scriptie zou verder niet mogelijk zijn geweest zonder de hulp van drs. Wim Kokx, ik ben hem zeer dankbaar voor het delen van zijn kennis en adviezen met betrekking tot de wereld van de middelbare scholen. Dankzij het brede netwerk van de Zo.Leer.Ik! scholen is het mogelijk geweest om mijn onderzoek uit te voeren. Graag bedank ik daarom ook alle schoolleiders, leraren en leerlingen voor de gastvrijheid die ik de afgelopen maanden heb mogen ervaren. Ook heb ik de gastvrijheid ervaren bij de bedrijven Neopost, Norma Drachten en Lely: het heeft mij veel inspiratie opgeleverd voor deze scriptie. Als laatste bedank ik uiteraard mijn ouders, broer, familie, collega’s en vrienden. Zij hebben mij de afgelopen 25 jaar de kansen gegeven en mij gevormd tot wie ik vandaag de dag ben.
Ik wens u veel leesplezier tijdens het lezen van deze scriptie. Jon Hummel,
Groningen, juni 2016.
English translation
4
1. Introduction
Every Dutch learner will visit secondary education after finishing primary education, at an age of approximately 12 years. The secondary education institutions will prepare the learner for a study at undergraduate level at research universities, universities of applied sciences or a study at a senior vocational education level (EP-Nuffic, 2014). Fifty percent of the learners do repeat one year in the total time period in primary and secondary education (van Vuuren and van der Wiel, 2015). Next to that, there is a large number of learners leaving secondary education without diploma. The government in the Netherlands placed a target of maximal 20.000 early school leavers without diploma in secondary education (Ministerie van Onderwijs, Cultuur en Wetenschap, 2017). Furthermore, it became clear over the last years that it is difficult to remain motivated during all lessons for the learners in secondary education (Inspectie van het Onderwijs, 2016). One of the solutions is the possibility to graduate at various levels for different courses (VO-raad, 2015).
In order to cope with all those challenges, the schools are implementing more solutions with personal learning paths (Inspectie van het Onderwijs, 2016). The practice of personalized learning can be described as the practical form of education whereby high-quality teaching is based on a sound knowledge and understanding of each individual learner (Miliband, 2004). Miliband (2004) emphasizes that this means not that learners only should use an electronic device or that learners will sit alone during this form of education. It should be regarded as the practice of shaping the teaching activities according to the way individual learners learn.
In the study of Prain, Cox, Deed, Dorman, Edwards, Farrely, Keeffe, Lovejoy, Sellings, Waldrip and Yager (2013), the difficulties regarding the organization of the personalized learning are summarized: problems arise regarding the flexibility of teachers (adaptability), student grouping and how the needs of different learners can be incorporated in the organization of personalized learning.
5 There is the need for research regarding the organization of personalized learning: “Personalized learning demands a new organization of education” in accordance with Schouwenburg and Keyser in a podcast about the Kunskapsskolan model (Schouwenburg, 2017). The characteristics of the personalized learning environment of the Kunskapsskolan model is that the learning activities are predefined in a portal. A learner will compose a planning every day and combines the planning with the pre-planned activities. This process of supplying learning activities and composing the planning every day for every learner can be regarded as the process of matching supply and demand in education.
Matching supply and demand is also common business for production companies (Cachon & Terwiesch, 2009). Mason-Jones, Naylor and Towill (2000) mentions lean, agile and leagile as practices for matching supply and demand. They defined agile as a practice which incorporates the customer more often, which makes it possible to readapt to changes and is focused on the shortening of production times. Lean is defined as a focus on seven wastes and a focus on operational efficiency with special attention on repetitive tasks. Leagile is defined by Mason-Jones et al. (2000) as the combination of lean and agile, for instance in a supply chain with a decoupling point where the first part is an agile supply chain and the last part is a lean supply chain.
Mason-Jones et al. (2000) mentions also the importance of placing the decoupling point. This is the point where a customer order will start with having influence on the process. This point is also important for the personalized learning environment of the Kunskapsskolan system: it is the point where a forecast driven sub process will stop and order driven sub process will start. In accordance with Vermeulen (2016), the learning environment with a Kunskapsskolan model can be regarded as a make-to-order (MtO) environment, because the learner will only use the predefined learning activities when a learner planned to do.
Therefore, the choice for the leagile philosophy becomes more obvious: the focus will be on eliminating waste between and in predefined learning activities and the aspect of incorporating the customer in the process often is suitable.
6 clustering of the needs of learners for personalized learning environments. Another contribution regarding the theoretical gap is the development of traditional operation management models towards models that can be used in a service context with human interactions.
1.1. Problem Definition
This research intends to design leagile based methods for Kunskapsskolan secondary education schools which can support them in (1) identification of the individual learning demand and (2) clustering the aggregated learning demand of the learners by performing a design method study.
For this research, it is important to create a distinction in parts of implementing personalized learning: the educational aspects and the organizational aspects. The educational aspects can be related to the didactics and content of learning. The organizational aspect can be considered as the practical organization of all learning activities. For this research, the focus will be on the organizational aspect of personalized learning within the boundaries of the operations management discipline.
1.2. Research Questions
The main research question that will be central in this research is: “How can a school identify and cluster the learning demand of individual learners by using methods based on leagile philosophy?”. This question can be divided in several sub questions:
1. Which factors do influence the learning demand and the supply of learning activities?
2. Which design constraints are necessary for constructing a method/design artifact that can be used in the educational context?
3. Which tools, principles and theory do exist about identification and handling of the product demand in a make-to-order environment? The focus will be placed on a lean, agile or leagile philosophy and on tools, principles and theory that seems to be useable in the educational context.
4. How does a Make-to-order production environment use lean, agile or leagile philosophy for the identification and handling of the product demand? The focus will be placed on tools, principles and theory that seems to be useable in the educational context.
5. How can the demand of learners and the supply of learning activities be captured in a design artifact with the input of RQ 1-4 and the use of leagile philosophy?
7
1.3. Methodology
The research questions do all need a well-fitted research method to answer the question. Diverse types of research questions do need other methods and just one method cannot answer all types of research questions (Karlsson, 2016). For this research, a multi-method research is the outcome.
General description
This multi-method research can be regarded as an explanatory study, where the intention is to derive a model based on empirical findings and current literature. Furthermore, it can be considered as theory building research, where inductive reasoning will be applied (starting from observations, via abstractions to a theory).
The study can be regarded as a design method study, because methods will be designed for the identification and clustering of the demand of learners in personalized learning environments. Therefore, this research can be classified as qualitative research. The choice for a design study is legitimated in this case, because the maturity of existing knowledge in this research field can be regarded as nascent theory: Karlsson (2016) proposes longitudinal field studies, design research and clinical management research as the most appropriate research approach for the existing base of knowledge. A longitudinal field study approach is not the best solution, because the problem definition does not aim to study a set of variables over a longer period. The choice between design research and clinical management research can be made with determining the position and role of the researcher in both approaches as defined by Schein (1987): process consultation aims on helping the client solve own problems (can be compared to design research), clinical management research puts the researchers into a doctor or expert role, the expert consultation. The aim of this research is supporting schools with the created methods and is therefore more comparable with the process consultation: the choice for design research is hereby legitimated.
Selection and description unit of analysis
For conducting the design method research, one school was selected. There are a few requirements for the selection of the school:
• The personalized learning environment/activities should be planned during the complete day, in other words: personalized learning should be structurally implemented.
• Previous research (Vermeulen, 2016) and experts in personalized learning emphasizes the role of coaching. The selected school should apply coaching in an active form, because coaching could also be one of the input variables for the methods that will be created.
8 After six school visits (see appendix A), it is determined that the Picasso Lyceum school in Zoetermeer is the best option for conducting the research, because the school applies to all above mentioned requirements.
The Picasso Lyceum offers TOPmavo (vmbo-tl), havo and vwo education to 1089 learners in the school year 2016-2017 (the school year where this research will be performed). A number of 220 learners do follow their education in the first and second year (brugklas). At this moment, 55 of those 220 learners are working with the Kunskapsskolan system in the first year. This group of 55 learners starts every day with a scheduling session where they plan learning activities together with a teacher. After this session, every learner will follow own routes in lectures, tutorials, group sessions and self-study activities. All learning activities are predefined in the learning portal and their progress is registered in the STROOM portal. Every learner will have a weekly coaching session. Subjects are divided in theme subjects (for instance biology): the subject consists of different modules, all modules discussing other themes. Another possibility can be that subject is a step subject (for instance a language related subject), where all steps build up in levels of difficulty. Learners can follow the step subjects in their own pace. The school is well-known for their extensive experience with customized learning paths to facilitate learners with top sport ambitions, learning disabilities and talented learners before they started with implementing the Kunskapsskolan system. In the upcoming school year (2017-2018), the number of learners involved in the Kunskapsskolan system will be upscaled to approximately 250 learners. The Dr. Nassaucollege school will be consulted for the validation of some parts of this research. This school has 228 learners, 40 learners are involved in the pilot with the Kunskapsskolan system. In the upcoming school year (2017-2018), the number of learners in the Kunskapsskolan system will be upscaled to approximately 70 learners.
Phasing of the research
Karlsson (2016) proposes four steps in design research: (1) constructing the action, (2) planning the action, (3) taking the action and (4) evaluating the action.
First the constructing and planning steps: for finding the criteria/requirements for the design artifact (research question five), the set of the first four research questions should be discussed first. The fifth research question will be based on creating the method (taking action step) and the last sixth question consists of the evaluation, validation and generalization step.
The research will be conducted in three phases:
9 3) Testing, validation and generalization of the action: Answering RQ 6
For every phase, the method, activities and the intermediate validation will be discussed. A literature review will accompany almost every method/phase in this research. This will be done because it can be stated that literature reviews aim to construct legitimacy and authority of the research (Karlsson, 2016).
Phase 1 – Educational requirements and practices from MtO environments (RQ 1-4)
In this phase, the goal is to compose the design constraints for the methods that will be created and factors for the learning demand and supply of learning activities from the educational. It is important to get those requirements and factors in a short time period and based on the collective opinion of teachers. The requirements are difficult to observe, so together with the fact that it should be performed in a short time period and should be based on a collective opinion, a focus group method will be the best solution (Chatrakul Na Ayudhya, Smithson and Lewis, 2014). The first focus group session was organized at the Picasso Lyceum. A literature review is conducted to find existing knowledge for this topic. The testing of those findings are guaranteed by distributing a digital interview in the Zo.Leer.Ik! Network and with organizing a focus group at the Dr. Nassaucollege in Norg with three teachers.
The research questions three and four are focused on the practices in industry regarding make-to-order environments when it comes to deriving of the demand and clustering the aggregated demand. The core of design research is applying an observational approach; it serves as theory building and as validation of experimental work (Cash, Hicks, Culley and Adlam, 2015). Therefore, company visits to three make-to-order production companies were conducted: Neopost in Drachten, Lely Maassluis and Norma in Drachten. Those companies are selected, because Neopost and Norma are part of the Innovationcluster Drachten. Lely Maassluis can be regarded as one of the most innovative production facilities that uses methods such as lean (Petovic, 2017; Dotman, 2016). Therefore, it can be concluded that those three companies can provide up-to-date and therefore relevant insights for this research. In addition to those visits, a literature study is performed.
Phase 2 – Designing and applying the methods (RQ 5)
10
Phase 3 – Validation (RQ 6)
The end of this research is concluded with two validation sessions at other schools than Picasso Lyceum or the Dr. Nassaucollege. In design research, it is important to test after the implementation if requirements are satisfied (Shabi, Reich and Diamant, 2017). In addition to that, Gibson (2017) mentions that face-to-face meetings can help to extent findings of the research to other settings. Therefore, the methods were presented to teachers of other schools and they had the opportunity to test the methods. This step has also the characteristics of a generalization step, because both schools do apply personalized learning in other forms compared to Kunskapsskolan system.
Visualization of methodology
In figure 1.1., the methodology of this research will be summarized and linked to all concepts.
1.4. Outline of this thesis
By now, the introduction, problem definition, research questions and methodology is discussed in chapter one. The second chapter of this thesis will discuss the current base of literature, the two focus groups and the digital interviews in the educational context. It will answer the first and second research question regarding the demand and supply of learning activities.
Chapter three will use a literature study and observations during location visits at three make-to-order companies to answer the third and fourth research question regarding lean, agile or leagile tools in make-to-order companies that can be suitable for the demand and supply of learning activities in the educational context.
Chapter four will use all insights from chapter two and three in order to create a design artifact. The implementation of this design artifact will be discussed in chapter five. In chapter six, the implemented design artifact will be validated and chapter seven will end this thesis with the conclusion.
Creating and implementing
the methods Zo.Leer.Ik! network
Presentation, testing and discussion with other schools RQ 5 Design of the methods RQ 6 Generalization & Validation Dr. Nassaucollege Focus group RQ 1 + 2
Design constraints and factors personalized learning Picasso Lyceum Focus group Zo.Leer.Ik! network Digital interviews PHASE 3 RQ 3 + 4 MtO production environment visits Literature study Neopost/Norma/Lely Location visits PHASE 2
Figure 1.1. Model of this research
Picasso Lyceum
Validation of the method Literature study
11
2. Educational context
This chapter will answer the first and second research question: it will provide the factors which can be used for the identification of the learning demand and it will discuss several design constraints that will function as the borders for the design artifact (chapter four) in this thesis.
Paragraph 2.1. will consist of the theoretical background about learners’ demand in the educational context. However, the current base of literature is not completely answering the research questions of this research as will be explained below. Therefore, paragraph 2.2. will explain the factors for the learning demand in terms of learning activities, based on focus groups and digital interviews. This chapter will end with the design constraints for the design artifact.
2.1. Theoretical background about Learners’ demand and supply of activities
The scientific literature does contain some literature about the needs of learners. A significant amount of this literature has a focus on electronic/virtual learning environments, for instance the work of Xu and van der Schaar (2016) that is performed in higher education institutions. This research aims to find a method for determining the best course sequence and combines information about course availability, prerequisite requirements and performance of a learner. The objective is to increase the overall result and to minimize the time to graduation. Insights of this research can be valuable for curriculum planners and planners who allocate teaching resources (Xu and van der Schaar, 2016). The work of Corbalan, Kester and van Merriënboer (2006) is focused on the role of a teacher that makes sequencing decisions of learning activities based on the learner’s expertise. Therefore, this research can be closely related to the system of the Kunskapsskolan where coaching is essential. Corbalan et al. (2006) created a model where the teacher selects a subset of learning activities based on performance scores and mental effort. With this model, the efficiency and effectiveness of instruction can be improved and the level of control in the system (which is important for personalized learning) can be guaranteed. The task selection is based on previous task performance, invested mental effort and level of expertise. The insights of Corbalan et al. (2006) are incorporated in an electronic learning environment.
12 which is adaptable and therefore suitable for personalized learning. Combining the research from Xu and van der Schaar (2016), Corbalan et al. (2006) and from Xu, Wang and Wang (2005) can provide insights for the requirements of a method which incorporates the coach of a learner, the possibilities of the curricula and can therefore provide real-time learning plans defined for the needs of learners. Xu, Wang and Wang (2005) do identify some aspects regarding the pedagogical model in selection of learning activities, however, they assign those pedagogical characteristics to complete learning plans: not to specific learning activities. Xu, Wang and Wang (2005) mentions those characteristics as the instructional strategy of the learning plan. Spronk et al. (2016) confirms this aspect with mentioning the preferences of learners for different pedagogical learning styles.
Previous research regarding the choice preferences of learners is performed by Spronk et al. (2016). First of all, this research defines the learner as the customer and not as the product. The focus in this research was from the point of view of teachers and a bit more on the level of choices between subjects. An important insight of this study is that the motivation for a certain subject or learning activity depends also on the preferences and choices of peer learners. Other influencing factors that are named by teachers were: speed of learning (that is also influenced by a lot of other factors), the planning for assignments and exams, advices of the coach and the level of difficulty for a subject. An interesting insight is also provided by the work of Hahn (2016) regarding the batching of learners for learning activities. An important criterion in this research in determining the form of learning activities is the number of learners.
13 the learner can choose from. Xu, Wang and Wang (2005) reveals this aspect of teachers’ expertise in the fact that there is an instructional strategy behind every learning plan.
The three factors for identifying the learning demand derived from current scientific literature are by now: cognitive performance of a learner, dependency and planning in learning activities and the advice of teachers and coaches.
The previous discussed literature is sometimes focused on higher education or other forms of education, also in other countries than the Netherlands with other educational systems. The literature is sometimes also focused on electronic learning environments only. The only literature that fits perfectly for this thesis is the research of Spronk et al. (2016) and Hahn (2016). Furthermore, regarding the small base of the literature and given the fact that the research question is not completely answered by using literature that is focused on other forms of education than personalized learning, further research is conducted in the form of an empirical study organized via focus groups and digital interviews.
2.2. Empirical findings
regarding Learners’ demand and supply of activities
This paragraph will discuss the empirical findings regarding the learners’ demand in the educational context and will confirm and/or elaborate more on the findings of the previous paragraph regarding the theoretical background.
During the first focus group at the Picasso Lyceum, it became clear that there should be a division between learner and teacher driven factors in determining what is needed for a learner. The following factors can be identified, based on focus groups at different schools (see appendix D and E for transcriptions) and were tested with digital based interviews (see appendix F and G) at other schools in the Zo.Leer.Ik! network.
Demand: Learner driven factors for the learning demand of a learner
Learners’ motivation
Interests in a subject
Insights in the learning goals Lack of knowledge
Planning of other activities (named during focus groups, but less in digital interviews)
14 motivation. It became clear during a focus group with a school with a lot of top sports players, that their level of motivation was significantly higher: there seems to be a relation with the discipline of a learner. A more or the less comparable factor is the interests in a subject.
The insights in the learning goals of a learner is an important aspect in general in personalized learning environments. It helps learners to get insights in what is needed for passing assessments and reveals the relevance of why a learner needs to complete tasks/steps in their learning process.
Another factor that drives the learning demand of a learner is awareness of the lack of knowledge, for instance if a learner is not able to process the learning materials.
The planning of other learning activities is named during the two focus groups and was also observed by the researcher of this thesis. A lot of learners named the upcoming presentation for instance as a reason to choose a certain learning activity. This is confirmed by the teachers during the focus groups. This factor is not confirmed in the digital interviews.
Supply: Which teacher driven factors do influence the supply of learning activities for a learner?
Effectivity of learning activities based on (historical) experiences of teacher
Foreknowledge of a group
Number of learners
Age of learners
Exit tickets
The effectivity of educational methods based on the experience of a teacher is a crucial factor, it entails the fact that a learning activity should be shaped around the content of the learning materials. To illustrate this with an example: a biology teacher knows that an instruction of a microscope can be better done via a practicum or tutorial session instead of a lecture.
The foreknowledge of a group tells a teacher what learners already know and what they need (in terms of content/learning material) during a learning activity. It determines the content of a learning activity and therefore it has influence on choosing the right activity, because of the link with the previous factor: the effectivity of learning activities based on the experience of a teacher.
The number of learners is a factor for supplying learning activities because of the organizability. An activity for hundred learners is more likely to be a lecture, an activity for a few (five learners for instance) will be more likely a tutorial session or a moment at self-study hours.
15 organize more tutorial sessions. Older learners at the end of their career at secondary education do have more lectures. It can also be linked to the notion of the experience in taking own responsibilities and level of self-steering as mentioned in the focus group at the school in Norg.
The tool ‘Exit tickets’ consists of a last question for learners at the end of a learning activity, for instance solving or answering a question that the teachers poses. The answer can be input for a next learning activity, based on how successful the learners were in answering the question. It provides detailed information for the teacher where the lack of knowledge is at the side of learners.
Considering the before listed factors, it can be concluded (and it is concluded during one of the focus groups) that there are a lot of factors influencing the demand of a learner and the supply of learning activities. Furthermore, it is difficult to measure those factors in a coherent way and in a way that it would be measured more frequently.
2.3. Merging empirical findings & theory: Factors for demand and supply of activities
In this paragraph, the theoretical background will be compared with the empirical findings. The result will be a list of factors that do drive the learners’ demand and the supply of learning activities. This paragraph will answer directly the research question: Which factors do influence the learning demand and the supply of learning activities?
The learners’ motivation is named in the focus groups and by the study of Spronk et al. (2016). The study of Spronk et al. (2016) also confirms the interests in a subject of a learner as one of the factors that drive a learner. The insights in the learning goals is mentioned in the focus groups, but not mentioned in the current base of literature: this could be because it is a factor that is prevalent in the personalized learning environment where the focus is far more on learning goals, compared to the more traditional forms of secondary education. The lack of knowledge as is mentioned in the focus groups can be directly linked to the cognitive performance factor that is found in the literature. The learner driven factor planning of other learning activities is not confirmed by the digital interviews, but it is named during focus groups and observations in schools. It is also confirmed by the literature in the theoretical background by Spronk et al. (2016). Therefore, there are sufficient reasons to assume that it is a relevant factor.
16 performance of a learner that is mentioned in the literature and are therefore also legitimated factors. The number of learners is found as a factor during the focus groups and is confirmed by the work of Hahn (2016). The age of learners as a factor relies on the span of attention that a learner can have during a given learning activity but is also linked to the experiences in taking responsibility and level of self-steering. Therefore, this factor can be linked to the findings about the cognitive performance in the theoretical background. The exit tickets do identify where a learner and teacher should pay more attention to (related to the content of a subject) and is therefore closely linked to the cognitive performance of a learner.
Some factors of the empirical findings are only confirmed by a single research of the theoretical background. However, the concerning factors do relate to the research of Spronk et al. (2016) and Hahn (2016). Those studies were performed in the same context and environment of this research, which makes it valid to incorporate those factors in the comparison of the main findings of this research.
The comparison between empirical results and theoretical background as discussed before is summarized in table 2.1. For this research, the naming of those factors will be based on the empirical findings, because those names relates closely with the personalized learning environment.
Learner driven factors - demand of learning activities Teacher driven factors - supply of learning activities
Empirical findings Theoretical background Empirical findings Theoretical background
Learners’ motivation
Motivation
Only mentioned by Spronk et al (2016)
Effectivity of learning activities based on experiences of teacher
Advice of teachers and coaches
Interests in a subject
Motivation
Only mentioned by Spronk et al (2016) Foreknowledge of a group Cognitive performance of a learner Insights in the learning goals
Not mentioned in current literature, because this factor is inherent to personalized education in the context of this research: it fulfills a theoretical gap
Number of learners Number of learners
Only mentioned by Hahn (2016)
Lack of knowledge Cognitive performance of a learner
Age of learners Cognitive performance of a learner
Planning of other learning activities
Planning of other learning activities
Exit tickets Cognitive performance of a learner
17
Intermediate conclusion
The answer on the research question: ‘Which factors do influence the learning demand and the supply of learning activities?’ is therefore: The learner driven factors are the insights in the learning goals, the lack of knowledge, learners’ motivation, interests in a subject and planning of other activities. The teacher driven factors for the supply of a learning activity are the effectivity of learning activities based on (historical) experiences of a teacher, the foreknowledge of a group, the number of learners, the age of learners and the exit tickets. There are more factors related, this conclusion consists only of factors where previous literature has found consensus (over more studies) and which were also identified in the empirical study of this research. A lot of factors do influence the demand of a learner and the supply of learning activities. It is difficult to measure those factors in a coherent way and in a way that it would be measured more frequently.
The frequently measurement of those factors should be ensured in a design artifact. Therefore, the measurement of those factors will be included as a design constraint, which will be discussed in the next paragraph.
2.4. Design constraints for a design artifact
For the construction of a design artifact, different design constraints are necessary which will define the boundaries and specifications of the design artifact (Karlson, 2016). This paragraph will provide the design constraints based on observations in schools and focus group sessions. It will answer directly the research question: ‘Which design constraints are necessary for constructing a method/design artifact that will be used in the educational context?’
General principles of the personalized learning context
The design artifact should support the personalized learning environment. As mentioned in the introduction in the description regarding personalized learning: Miliband (2004) emphasizes that personalized learning should be regarded as the practice of shaping the teaching activities according to the way individual learners learn. A learner will compose a planning every day and combines the planning with the pre-planned activities. The process of supplying learning activities and composing the planning every day for every learner can be regarded as the process of matching supply and demand in personalized education. Furthermore, Prain et al. (2013) mentioned that problems arise regarding the flexibility of teachers (adaptability), student grouping and how the needs of different learners can be incorporated in the organization of personalized learning. The design artifact should contribute to the solving of those problems.
18 organization of the school: the supply of learning activities should be shaped towards the needs of learners in terms of what, when and how (pull) and should not be based on the teachers’ plans only (push). Furthermore, Dam (2016) advices to create a tool that aggregates the individual learning demand towards a cumulative clustered demand of a set of learners. Finally, Dam (2016) shows the distinction between standard learning activities and ‘specials’: special learning activities. Some standard learning activities can be handled by every teacher and the special learning activities can be handled by teachers with the best characteristics or knowledge in order to handle those special learning activities (like a multi-mixed line in operations management terms, with standard operating lines and branches for specials). During the observations in schools, it became clear that those standard activities were pre-planned and the special learning activities were planned on a later moment and shaped to the new demand of learners. Therefore, this research will use the term ‘pre-planned activities’ for the standard activities and the term ‘post-planned activities’ for the specials.
Learners’ ownership and learning goals
The personalized learning concept of Kunskapsskolan is created with the ownership of a learner in mind (Schouwenburg, 2017). Every learner will get the trust that the learner will develop discipline, self-knowledge and a feeling of responsibility in the educational route, own possibilities and development of talents. Schouwenburg and Keyser (2017) mentions that the support of a coach and preformulated learning goals will help a learner in learning effectively. This results in a plan-do-check-act cycle (Deming cycle) repeating during the whole time of a learner present at a school.
The design constraint for the design artifact that can be derived from this is that it should support (and not prohibit) the ownership of a learner.
Teachers’ autonomy
During several focus groups, it became clear that the autonomy of a teacher is important: a teacher is willing to make decisions based on own experience and the teacher should always be able to influence possible outcomes. This is also supported by a literature review conducted by Parker (2015), where teacher’s autonomy is defined as the teacher’s ability to balance between experience, training, personality and the requirements of the educational context regarding the control in their work environment. Professional autonomy of a teacher is defined as the ability of believing the authority and influence on individuals, ideas and ideals. A strong defense for establishing and maintaining professional autonomy is important for the future of the education system (Parker, 2015).
19
Frequently measurement of learner and teacher driven demand factors
The demand for learning activities should be measured frequently, because the demand of learners can change over the time. The factors that do influence this demand are listed in paragraph 2.3 of this research.
Intermediate conclusion
In this paragraph, the research question is answered which relates to the design constraints of the design artifact in this research. There are some typical characteristics of personalized learning where the design artifact should apply to: creating pull, difference between standard and special learning activities and it should be a design artifact that is able to aggregate the individual learning demand towards a cumulated learning demand. Furthermore, the design artifact should incorporate the ownership of learners and should have a link with the learning goals that are used in the educational context (supported by focus groups, observations ant literature). The autonomy of teachers is an important aspect when the design artifact will be designed and constructed (based on focus groups and literature). The last design constraint is the frequently measurement of the learning factors that were derived in paragraph 2.3. of this research.
2.5. Main conclusion about the educational context
The design constraints for a design artifact are that it should confirm to the typical characteristics of personalized learning: a focus on the learning goals is important, the design artifact should shape learning activities based on the learners, a pull flow should be the result, the design artifact should aggregate the individual learning demand towards a cumulative clustered demand of a set of learners and the distinction between standard learning activities and ‘specials’ should be made. The second design constraint is that it should confirm to the teachers’ autonomy and learners’ ownership. The third and last design constraint is that the design artifact should be able to measure the previous learner and teacher driven factors frequently. The learner driven factors for the learners’ demand are the insights in the learning goals, the lack of knowledge, learners’ motivation, interests in a subject and planning of other activities. The teacher driven factors for the supply of a learning activity are the effectivity of learning activities based on (historical) experiences of a teacher, the foreknowledge of a group, the number of learners, the age of learners and the exit tickets.
20
3. Make-to-order production context and leagile philosophy
This chapter will discuss the literature and observations in make-to-order companies as is performed, to answer the third and fourth research questions, about the literature and practices in lean, agile or leagile philosophy in make-to-order production environments. The previous chapter about the educational context states the design constraints for the design artifact. This chapter will explore solutions that potentially can be used as starting point for the design artifact for the educational context and therefore, the design constraints will be already deployed as boundaries for this chapter. The first design constraint is that the artifact should fit in the personalized context and should therefore ensure a pull flow and it should cluster the aggregated learning demand. The second design constraint is that that the teachers’ autonomy and learners’ ownership should be ensured and the last design constraint is that the learning demand should be measured frequently.
The first paragraph will discuss the current base of literature regarding the research questions of this research and limited by the boundaries of the design constraints. The second paragraph will discuss the observations in three make-to-order environments and will show some best practices regarding the problem definition in the educational context (again, the design constraints will be used as a boundary for discussing the observations). This chapter will end with a conclusion about the insights from literature and practice regarding lean, agile or lean philosophy in the identification and handling of the demand.
3.1. Theoretical background
21 choice for agile tools combined with lean factors more legitimated: the focus should be placed on an agility tool.
Agility tools
The educational context of this research can be characterized as a service-delivering industry. Koch (2005) mentions different agility tools that are used in the software development industry, an industry that is also focused on delivering a service instead of a physical product. This industry can be regarded as engineer-to-order (EtO) companies, a decoupling point placed before the make-to-order customer decoupling point. For this research, it is also suitable to include literature of environments where the decoupling point can be placed before the make-to-order (MtO) point such as the engineer-to-order point, because the focus is on the transition of a make-to-stock decoupling point towards decoupling points like EtO or MtO where the customer (the learner in this case) has more influence on the outcome in an early phase of this process.
The software development industry does share characteristics with the context of personalized learning: in both, the demand can change over time and in both contexts, a service will be provided (and not a physical product). Agility methods as mentioned by Koch (2005) are: adaptive software development, dynamic system development method, extreme programming, feature-driven development and Scrum.
Adaptive software development is focused on the learning experience in the development of software and, therefore, not relevant for the teachers in the educational context: it doesn’t solve the problem as mentioned in the problem definition of this research about the identification and clustering of the demand. The same is applicable for the dynamic system development method and in addition to this, this method is also focused on delivering a minimal viable product in a brief time period which can be tested at the end user. That is also not relevant for the educational context as mentioned in the problem definition or design constraints, because it is not a design constraint that a minimal viable product should be delivered. Extreme programing is a collection of best practices in programming and, therefore, also not applicable because the focus differs too much from the educational context. The feature-driven development method has the restriction that it places focus on implementing one feature at a time and furthermore, the first activity in this method is creating the architecture for a software system: both aspects are not applicable for matching the supply and demand of learning activities.
22 that a teacher has always a maximum number of learners, but in practice, a teacher is also able to change the forms of learning activities in order to serve more learners at once (switching from tutorials towards lectures for instance). Koch (2005) mentions also that this method does not address software development specifically. Therefore, the Scrum method is interesting to explore as a base for the design artifact in relation to the problem definition of this research: it is focused on the planning and can be used in other contexts than software development.
Scrum principles
The principle of Scrum is explained by Sutherland (2015), the principle starts with a product owner. This person is the stakeholder of the result and determines what is needed for the product. What is needed for the product is determined in user stories. Those are short descriptions of product features or wishes formulated in a format, it should always be formulated as a need for a feature to fulfill a goal. All those user stories will be placed in the product backlog. Some of those user stories can be grouped together in themes, when they have the same purpose or characteristics. Next to themes, those themes can be grouped in epics. An epic is a bundled set of themes that consist of a complete workflow for a product owner or end user. All user stories, themes and epics are placed on the product backlog: the to do list for the developer team.
The developer team is the team that will do the work for the product owner. This team will discuss and reformulate the user stories together with the product owner. Furthermore, one of the roles in Scrum is the role of the Scrum master: this person will coach the rest of the team in using Scrum. The focus of this person is on eliminate anything that is slowing the team down.
The backlog will be divided in different sprints, a sprint is a period where a defined set of user stories/themes or epics will be completed. The sprint will always have the same length and will always be less than a month. Most of the teams do run an one-week or two-weeks sprint. During meetings, the next sprint will be planned and the team will review their work. Reviewing the work can be done in the daily stand up where every member of the developer team explains what is completed, what should be done next by this person and what blocks him or her from doing tasks.
23
Scrum in the educational context
Applying Scrum methods is not new in the educational context. Delhij, van Solingen and Wijnands (2015) explained the concept of eduScrum: a so called Scrumbut (an altered form of the original Scrum method) for education. The tree pillars of the eduScrum method are transparency, inspection and adaption. Transparency means that a defined language should be used between all participants in the learning process and all participants should use the same definitions of stages, processes and statuses. Inspection refers to the fact that a frequently inspection towards learning goals is necessary in order to detect deviations. Those frequently inspections should be done by teachers and students. The third pillar is adaption: if a student or teacher identifies that the learning process does not fit the situation of the learner, adjustments should be made in order to minimize further deviation. It is remarkable how this approach of eduScrum does apply to the design constraints of this research: the pillar ‘adaption’ and the pillar ‘inspection’ complies with the design constraint that the learning demand should be identified frequently. The pillar of ‘adaptability’ can also be linked with the design constraints about the teachers’ autonomy and learners’ ownership.
However, eduScrum is focused on the planning of all activities inside one subject. That does not comply with the problem definition of this research where the design artifact should solve the problem for the planning of all subjects. The eduScrum method is also more focused on the content of a subject and the self-organization of project teams in the subject for learning. Therefore, implementing eduScrum does not answer the research questions and problem definition of this research and is more focused on the pedagogical aspects of learning. However, the example of eduScrum and the close link with the design constraints of this research brings us to the conclusion that the Scrum method can be successful applied to the educational context. Therefore, it is interesting to explore if this eduScrum approach can be extended to the educational context of this research.
Based on the before mentioned arguments about agility tools and the successful example of eduScrum, this research will use the Scrum method as a base for the design artifact. In other words, the design artifact will be a Scrumbut: an altered form of the original Scrum method).
Lean principles as addition to agility tools such as Scrum
Parnell-Klabo (2006) states that metrics of the lean philosophy about throughput times, waiting times and cycle times can be helpful in eliminating waste when applying agility tools such as Scrum.
24 For a make-to-order company, Bateman, Philp and Warrender (2016) developed a communication board for the different cells of the production environment. The board consists of three parts: cell overview (cell members, products, process flow, attendance planner), the cell performance (weekly efficiency, weekly down time and quality concerns) and the cell improvements, history and plans. More insights in addition to this board are provided by Galsworth (2005): the distinction between cell metrics, machine metrics and management metrics only is created. Remarkable is that Galsworth (2005) shows only examples of those three metrics where the goal or plan is compared with the actual output. The explanation is that this information does improve productivity. Next to this real-time information, there is also a category of metric displays with stationary information: metrics that monitor only. For most of the time, this is paper based and shows the performance of the previous week for instance. Galsworth (2015) stresses the need for both forms of visualization: it is the baseline for problem solving in lean production environments.
Visualization combined with workload control
Another visualization technique that can be discovered in lean and make-to-order environments is a production system with the use of cards, like the Kanban system (Hopp and Spearman, 2008). A station, machine or cell will only produce parts if a card from a downstream station is received, therefore it is also a method of workload control and avoidance of the waste of inventory. The card can also contain information about the number of parts that should be produced. With this system, the demand is visualized with cards at every station. The Kanban system reduces the amount of inventory between stations (Hopp and Spearman, 2008). Some implementations of Kanban places a maximum on the work in progress, as mentioned before: such kind of implementations are not useful for the educational context, because a teacher can change the forms of learning activities in order to serve more learners at once.
25
Intermediate conclusion
It can be concluded by now that the Scrum approach will be the best agility tool as underlying principle for the design artifact of this research. Different options of visualization and the visualization of metrics are discussed that can make the agile tool Scrum a more leagile approach, as discussed by Parnell-Klabo (2006).
For the three company visits, it is interesting to find out how they visualize metrics on the work floor and how they organize their process of demand identification. Those observations can give more direction for the to be developed metrics in the design artifact of this research.
3.2. Observations in make-to-order production environments
In this paragraph, the observations in three make-to-order production environments will be described. The earlier mentioned design constraints for a design artifact will be used as a boundary for what is and what is not discussed in this paragraph. During the company visits, attention is payed to the way the production environment uses metrics and the visualization of those metrics (as also discussed in the previous paragraph). Furthermore, it is observed how those companies do visualize the demand and specifications on the work floor.
In all three companies, an order specification form (see appendix C for an example) is used for the identification of the demand in the production process. Those forms contain the information about the product family or name, customizations needed, the status of the product during the production process and does function as a checklist for checking if a product has completed all different production steps. One of the companies has created a digital task list where all steps of a product are automatically visualized for every department. It is also good to mention that this company has installed a very detailed ERP system which is able to compose the planning for a given product and is able to readapt the planning if a given production step faced delays or could handle the step earlier than planned: this seems very relevant and useful for the educational context. A learning factor derived from the input of this company is that an information system should provide only relevant information for an employee in the shortest way, to prevent an overflood of information.
Another interesting observation is that two companies used the concept of Kanban actively as a workload control system, for instance between the pre-assembly department and the final assembly department of a process line. Kanban is also used in the control of inventory of raw material in the processes.
26 the state of the orders where the orders were divided over three columns: Upcoming, processing or finished.
One of the companies incorporated Andon in almost every production station. This company used it as a tool to call a supervisor for further assistance or the ‘waterspider’: a colleague responsible for the distribution of raw material, products and tools in the plant.
3.3. Main conclusion about the make-to-order production context
The main conclusion answers the following research questions: Question three, ‘Which tools, principles and theory do exist about identification and handling of the demand in a make-to-order environment that uses a lean, agile or leagile philosophy?’ and question four, ‘How does a Make-to-order production environment use lean, agile or leagile philosophy for the identification and handling of the demand?’.
The use of (insights from) Scrum, Andon and Kanban is discussed in literature and is observed in three different make-to-order companies. Therefore, the use of those three tools do answer both research questions three and four.
27
4. Model
This chapter will continue with the construction of the design artifact in this research and therefore will answer the fifth research question: ‘How can the demand of learners and the supply of learning activities be captured in a design artifact with the input of the first four research questions and the use of leagile philosophy?’.
From chapter two, the factors for learning demand and supply are known and the design constraints can be applied to the design artifact. Chapter three provides the insight that the Scrum method will be adapted towards the context of personalized education and the design artifact will have additional lean principles and tools.
The first paragraph will briefly summarize the input from chapter two. The next paragraph will start with the conceptualization of Scrum elements towards the design artifact in the educational context. The third paragraph will construct the design artifact with adding the additional lean tools in order to create a leagile based design artifact. This chapter will end with a schematic overview of the building blocks of the design artifact, a reflection on how the design constraints are met and a conclusion.
4.1. Overview of design constraints
The design artifact should confirm to the context of personalized learning: a focus on the learning goals is necessary and a pull flow should be implemented. The design artifact should aggregate the individual learning demand towards a cumulative clustered demand and the distinction between pre-planned learning activities and post-planned learning activities should be made. The second design constraint is that it should confirm to the teachers’ autonomy and learners’ ownership. The third design constraint is that the design artifact should frequently measure the learners demand. The learner driven factors for the learners’ demand are the insights in the learning goals, the lack of knowledge, learners’ motivation, interests in a subject and planning of other activities. The teacher driven factors for the supply of a learning activity are the effectivity of learning activities based on (historical) experiences of a teacher, the foreknowledge of a group, the number of learners, the age of learners and the exit tickets.
4.2. Conceptualization Scrum towards the design artifact
The elements of Scrum as mentioned before can be conceptualized towards the environment of personalized learning. In this paragraph, all earlier discussed Scrum elements will be coupled to an element of the personalized learning environment.
28 The product owners in a Scrum method is the person who determines together with the client what is needed for the product (software, or in educational context: the learning path). In personalized learning, this will be the learners together with the coach. A learner will compose the wishes/needs and acts based on those needs by discussing this with teachers and by composing the own planning. Together with a coach, those wishes, needs and actions will be discussed on a weekly basis. The product owners will formulate their user stories.
User stories are short formulated texts that determines what should be delivered by the developer team. Those user stories are the wishes or needs of a learner as discussed by the product owner element of the Scrum method. Therefore, those user stories are in fact the demand of a learner. The product backlog consists of all user stories that should be handled by the developer team. Therefore, this list of user stories is the overview of all individual learners’ demand in the educational context.
The Scrum method uses themes and epics as a clustering method for user stories. User stories with the same characteristics can be grouped together in themes. Furthermore, themes with the same characteristics (on a higher level) can be grouped again in epics. This concept can be applied in the educational context by stating that themes should be all user stories that can be handled with the same learning activities. All themes related to a subject in the educational context can be grouped into the epics, so every epic is a subject. A given theme will be handled with a to be determined learning activity. The question regarding the educational context now is where a developer team should look for, in clustering the user stories into themes. One of the design constraints is that the design artifact should support the use of learning goals. Furthermore, it can be concluded that the learning goals can be compared with product specifications as we would consider it in a make-to-order environment. Clustering in make-to-order environments can be done by clustering the products with the same specifications, which needs the same handling during the process (Cachon and Terwiesch, 2009). Sprints are blocks of a given time period that will content some user stories, themes or epics that will be handled during that time period. Those sprints are planned in a sprint planning. During the daily Scrum stand-up, every participant of the developer team will reflect on what is done, what will be done next and what’s blocking the developer from doing what should be done. In the personalized learning context, those sprints should be the planning of themes for the upcoming weeks: themes are converted to learning activities, those learning activities will be planned. During the weekly team meetings (that are already organized in the schools), the sprint planning can be done and the daily Scrum stand-up can be performed.
29
Scrum element Explanation Personalized learning element
Developer team Who constructs the product or
service? Teachers
Product owners Who determines what is
needed for the product? Learners with coaches. User stories Wat should be delivered by the
developers? Demand of a learner.
Product backlog All user stories to be
implemented. Overview of all individual learners’ demands. Themes and epics User stories with the same
characteristics can be grouped together in themes. Themes can be grouped into epics.
Themes: group of individual demands which can be handled with same learning activity.
Epics: a subject.
Sprints User stories, themes and epics
will be planned/distributed in different sprints.
Planning of themes for upcoming weeks. Sprint planning & daily Scrum
stand-up The planned user stories, themes and epics in the sprint planning will be discussed every day.
Weekly team meetings
Table 4.2. Conceptualization of the SCRUM approach in personalized learning
4.3. Constructing a leagile based design artifact
After the conceptualization, it is important to shape the design artifact in such a way that a workable solution will be constructed that complies to the design constraints. Furthermore, lean principles and tools (already mentioned in the theoretical background of this research) will be introduced to control the workload for teachers. The Scrum elements will be renamed to more explanatory terms in order to make the working of the design artifact more clear for learners and teachers.
First step: identification of the learners’ demand
30 activities will be the learning activities that should be shaped towards different or deviating needs of learners.
The next question is: how can we make the distinction between pre-planned learning activities and post-planned learning activities for learners? For this distinction, the lean tool Andon will be useable. A learner will follow the pre-planned learning activities and will trigger the Andon if the demand of the learner requires a post-planned learning activity. The concept of Andon is also wishful based on the design constraint about the ownership of a learner. The learner triggers the Andon by itself. A user story for a post-planned activity will be formulated after a learner has triggered the Andon. An important design constraint in the general characteristics of personalized learning is the use of learning goals. Therefore, a user story should be coupled to a learning goal.
Workload control is now necessary in order to prevent too much work-in-progress for teachers. Therefore, a one-bin Kanban system is applied for the user stories of a single learner. If a set of user stories is submitted to a teacher or subject, the learner must wait with sending new user stories for the same teacher or subject. New user stories can be send when the previous user stories are planned to finish. There is also a very practical reason from the educational context for this one-bin Kanban system: a teacher will determine a learning activity based on the received user stories from a learner. There is the possibility that a teacher will receive conflicting user stories if it is possible to send more user stories after each other for a subject. It is also not efficient for a learner to send new user stories requesting a post-planned learning activity before previous user stories are handled: a learner should formulate and send the user stories, requesting for a special activity and should then continue with other tasks for that subject or with other subjects. It is not efficient for the learner to continue and struggle on their own with the specific task if user stories requesting a post-planned learning activity were sent to the teacher. This form the one-bin Kanban of workload control will be tested in the validation phase of this research.
The whole concept of user stories, Kanban and Andon will be renamed towards the term ‘Signals’. Those signals are the so-called emergency brakes for a learner when the learner needs a post-planned learning activity. A signal will be formulated by the learner in relation to a learner goal and will be send to the teacher.
Second step: clustering the signals (user stories) into themes
31 possible for the teachers to create those themes automatically based on the learning goals of the signals in the inbox. All signals will be grouped together in themes based on their learning goals. A teacher will always have the autonomy to bring signals to other themes afterwards. It is also possible to redirect signals to the coach if the signal is not useable or relevant for the teacher. Another possibility is to redirect a signal to the self-study hour, for instance if the number of learners for a learning activity is too low (one of the factors for the supply of learning activities found in this research).
Another factor for the supply of learning activities found in this research, is the known effectivity of a learning activity for a given content of a lesson based on the experience of a teacher. Therefore, it is completely up to the teacher to give a theme the right form of a learning activity and a time duration. The Scrum terms like themes and epics are also unknown for the teachers. Therefore, the term ‘themes’ will be replaced with the word clusters: a cluster of signals that can be handled by a learning activity. The ‘epics’ in Scrum will be the same as subjects in the educational context.
Third step: planning
The last step in the design artifact and in this Scrum approach will be the planning. All clusters of signals are linked to learning activities with a form and a time duration. Those learning activities will be planned. In the Scrum method, a Scrum master will collaborate with the developer team to plan their sprints. In the context of personalized learning and this design artifact, the role of the Scrum master will be taken by the project manager of personalized learning or the department leader. It could also be one of the teachers who take the role of organizing the Scrum meeting and the daily Scrum stand-up. During those team meetings, the sprints will be planned. In terms of personalized learning: the planning of learning activities. All the to be planned learning activities will be listed in the backlog, this backlog will be renamed to the Planning inbox. By focusing on the tempo in the Scrum method, the lean principle about eliminating the waste of waiting will be applied. The throughput time as a lean metric will function as an informational metric for the teachers and project manager/department leader about how long it takes before a signal of a learner is planned.
Building blocks of the design artifact
32
SCRUM elements/agility Lean principles/tools Design artifact elements
User stories Andon
Kanban Signal from learner to teacher
Themes Clusters: grouped user stories
with same learning activity
Epics Subjects
Backlog Planning inbox
Scrum master Project manager / dept. leader
Sprints Planning of learning activities
Tempo Waste of waiting, throughput
time Amount of planned learning activities
Burndown Cumulative amount of planned
learning activities
Developer team Teachers
Table 4.3. Elements of the Design artifact
Figure 4.2. Building blocks and workflow of design artifact
Plan learning activities during team
meetings First step: identification Second step: clustering Third step: planning
Learning goals Construction of signals Andon Yes No Kanban
Signal for subject already sent and not
completed?
Process terminated: Learner will continue with other tasks or subjects
Signal for self-study hour?
Cluster signals to learning activities, determine form
and time duration
33
Reflection regarding the design constraints
The last question in this paragraph is now: does the design artifact meet the design constraints of this research?
The design artifact places the focus on learning goals, by formulating signal in relation to the learning goals. If a teacher assigns the correct learning activity to a set of signals, the design artifact shapes indeed the learning activities based on the learners. The teacher still has all autonomy to decide for the best suitable learning activities. The pull demand is ensured because the learner formulates the signal (which triggers the process) by themselves, the ownership is still at the learner. The design artifact can cluster the individual learning demand in the second step of the method and by incorporating Andon, the distinction between pre-planned and post-planned learning activities is made. The method is able to measure the learning demand frequently, because a learner can place a signal when it is needed. The factors for the demand and supply, such as the effectivity of a learning activity, the numbers of learners, the insights in learning goals and the lack of knowledge is all incorporated: partly because the learner should relate the signal to learning goals and the other factors can be measured because the learner is able to state their signal in a free text field and can therefore formulate the learning demand freely, the teacher can decide autonomously based on those text messages.
