CaMeLOT: An educational framework for conceptual data modelling

https://doi.org/10.1016/j.infsof.2019.02.006Get rights and content

Abstract

Context

Teaching conceptual data modelling (CDM) remains a challenging task for educators. Despite the fact that CDM is an integral part of software engineering curricula, there is no generally accepted educational framework for the subject. Moreover, the existing educational literature shows significant gaps when it comes to pursued learning outcomes and their assessment.

Objective

In this paper, we propose an educational framework for conceptual data modelling, based on the revised Bloom's taxonomy of educational objectives, and provide necessary examples of systemized learning outcomes.

Method

We utilized the revised Bloom's taxonomy to develop an adapted framework specifically for learning outcomes related to CDM. We validated the framework by mapping learning outcomes distilled from the existing course material to the framework, by presenting the framework for feedback to the experts in the field and further elaborating and refining it based on the feedback and experiences from these validation activities.

Results

CaMeLOT is an adaptation of the Bloom's taxonomy specifically for learning outcomes related to CDM. We identified different content areas and indicated the necessary scaffolding. Based on the framework, we worked out 17 example tables of learning outcomes related to content areas at different levels of scaffolding, exemplifying the different knowledge and cognitive levels. We clarify the differences in learning outcomes related to different knowledge and cognitive levels and thereby provide a domain specific clarification of the classification guidelines.

Conclusion

CaMeLOT gives educators an opportunity to enhance the CDM part of software engineering curricula with a systemized set of learning outcomes to be pursued, and open the path for creating more complete, useful and effective assessment packages. The adoption of our educational framework may reduce the time spent on designing educational material and, at the same time, improve its quality.

Introduction

Conceptual data modelling can be considered a crucial part of software engineering curricula, being “the phase of the information systems development process that involves the abstraction and representation of the real world data pertinent to an organization” [1]. Such development process involves solving ill-structured problems and implies activities at high levels of abstraction, which poses a substantial challenge for educators. Moreover, due to the ill-structured nature of modelling problems, which are in most cases context-dependent, software engineering students have to grasp not only the modelling techniques themselves, but also certain specifics of the domain and context in which the task is situated [2].

Though data modelling is part of the majority of software engineering curricula, no generally accepted educational framework for data modelling exists to this moment. Subsequently, educators become the ones responsible for setting the learning outcomes to be pursued in the data-modelling course and for designing the entire course, based mostly on their own experience and professional judgement. However, not every educator has necessary experience and time resources to come up with a full set of learning outcomes, even though clearly set learning outcomes are essential for the learning process [3]. The path between the two states of “novice modeler” and “expert modeler” remains relatively unclear despite the fact that the differences between them have been already explored in a large number of scientific studies (e.g. [4], [5], [6]). Such differences include the stages of modelling implemented by the experts, the differences in modelling patterns, and strategies.

While many educators rely on existing educational literature on data modelling as a basis for their course material and design, the initial study of current educational resources on conceptual data modelling has shown a considerable number of gaps in the learning outcomes presented in those sources [7]. As the research showed, one of the most important types of tasks – evaluation of models, – was significantly underrepresented in the practical educational materials. Moreover, while the tasks related to the ultimate goal – creation of the model, were present in all types of educational resources, the scaffolding steps and the lower levels leading to the mastery of creative skills in modelling remain unequal and lack consistency.

Student experiences from conceptual modelling courses have also not been flawless. Despite the fact that the history of conceptual modelling education spans more than several decades, the most common difficulties and mistakes in novice modellers’ solutions persist: the difficulties they faced more than 20 years ago, such as understanding the requirements, or choosing the right multiplicity [8], are very similar to those they faced 10 and 20 years later [9], [10]. These persistent difficulties and errors correspond to the initial skills that the students have to acquire before being able to create fully functional models.

These concerns suggest that the scaffolding process in conceptual modelling education requires a thorough review and improvement. This study aims at filling the identified gaps by presenting a systematic educational framework for data modelling education, based on the revised Bloom's taxonomy of educational objectives [11]. The work includes the identification of scaffolded content areas, a domain-specific definition of the Bloom's taxonomy's levels, the development of recommendations for learning outcomes classification as well as examples of learning outcomes to be pursued in the identified particular content areas of the course.

Section snippets

Background and related work

In the last 2 decades, numerous studies have been made on software engineering pedagogy, exploring various approaches to curriculum development and enhancement (e.g. [12], [13], [14]). However, very few of them explore the pedagogy related to the data-modelling phase of software engineering in depth. In 2005, Cowling [15] argued that modelling in general had not been given sufficient attention in the software engineering curriculum, while the role of modelling is crucial in the software

Design methodology

CaMeLOT's design was performed using the design science approach adapted from the Information Systems Research Framework developed by Hevner et al. [27]. The choice of the approach was determined by the nature of the aim of this study: development of a taxonomy of learning outcomes for future practical use by the conceptual modelling educators, taking into account the current state of the art and needs of the modelling community.

As described by Hevner et al., design science research (DSR) is

CaMeLOT: a revised Bloom's taxonomy for conceptual data modelling

Next to an adapted definition of the knowledge and cognitive levels, CaMeLOT also proposes the identification of a set of distinct content areas and their scaffolding.

Example sets of learning outcomes

To validate the classification framework as a practical tool to be used by educators, we have developed example sets of learning outcomes for six content areas covering two subdomains of conceptual data modelling. For each content area, one or two tables with a line of example learning outcomes per knowledge level is given. Not every content area would have all the knowledge levels presented and certain intersections between content areas are expected in many cases. The exception is

Evaluation

To perform the preliminary evaluation of the current version of the framework, seven experts in the field of conceptual modelling education were contacted and asked to participate in the framework evaluation in a form of an interview that followed the presentation of the framework and the experts’ familiarization with the guidelines. Three experts were found willing to participate in an in-depth interview. The summary of the interview gathering process is presented in Table 18.

As a result of

Discussion

The set of tables presented in Section 5 provides examples and guidelines to conceptual data modelling educators wishing to enrich their educational practice with new types of exercises and assessment items or to identify the gaps in the existing curriculum.

Conclusion

In this study, we have developed CaMeLOT as a revised Bloom's taxonomy adapted for conceptual data modelling. Besides a reinterpretation of the knowledge and cognitive levels, the framework also identifies and scaffolds content areas in the domain of conceptual modelling. While the content area of Model Creation was elaborated specifically for data modelling, it can easily be extended for other types of modelling (e.g. business process modelling). The content area of General Modelling, as its

Acknowledgment

This research is funded by grant C24/16/002 of the KU Leuven Research Council.

References (49)

  • F. Leung, N. Bolloju, Analyzing the quality of domain models developed by novice systems analysts, in: Proc....
  • M. Kayama et al.

    A practical conceptual modeling teaching method based on quantitative error analyses for novices learning to create error-free simple class diagrams

  • D.R. Krathwohl

    A Revision of Bloom's Taxonomy, Theory Into Practice

    (2002)
  • A.J. Cowling, A multi-dimensional model of the software engineering curriculum, in: Proc. Eleventh Conf. Softw. Eng....
  • R. Adcock et al.

    Curriculum guidelines for graduate degree programs in software engineering

  • J.T. Gorgone et al.

    MSIS 2006

    ACM SIGCSE Bull

    (2006)
  • M. Ardis et al.

    Curriculum guidelines for undergraduate degree programs in software engineering

    Comput. Curric. Ser.

    (2014)
  • Data modelling and design DTAN — SFIA, (n.d.)....
  • B.S. |An et al.

    Handbook on Formative and Summative Evaluation of Student Learning

    (1971)
  • A. Crowe et al.

    Biology in bloom: implementing Bloom's taxonomy to enhance student learning in biology.

    CBE Life Sci. Educ.

    (2008)
  • A.Y. Zheng et al.

    Application of Bloom's taxonomy debunks the "MCAT

    Science

    (2008)
  • A.R. Thompson et al.

    The Blooming anatomy tool (BAT): a discipline-specific rubric for utilizing Bloom's taxonomy in the design and evaluation of assessments in the anatomical sciences

    Anat. Sci. Educ.

    (2015)
  • N.B. Zaidi et al.

    Climbing Bloom's taxonomy pyramid: lessons from a graduate histology course

    Anat. Sci. Educ.

    (2017)
  • C.W. Starr et al.

    Bloom's taxonomy revisited

  • Cited by (22)

    View all citing articles on Scopus
    View full text