No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Towards a More Semantically Transparent i* Visual Syntax
Abstract
[Context and motivation]
i* is one of the most popular modelling languages in Requirements Engineering. i* models are meant to support communication between technical and non-technical stakeholders about the goals of the future system. Recent research has established that the effectiveness of model-mediated communication heavily depends on the visual syntax of the modelling language. A number of flaws in the visual syntax of i* have been uncovered and possible improvements have been suggested. [Question/problem] Producing effective visual notations is a complex task that requires taking into account various interacting quality criteria. In this paper, we focus on one of those criteria: Semantic Transparency, that is, the ability of notation symbols to suggest their meaning. [Principal ideas/results] Complementarily to previous research, we take an empirical approach. We give a preview of a series of experiments designed to identify a new symbol set for i* and to evaluate its semantic transparency. [Contribution] The reported work is an important milestone on the path towards cognitively effective requirements modelling notations. Although it does not solve all the problems in the i* notation, it illustrates the usefulness of an empirical approach to visual syntax definition. This approach can later be transposed to other quality criteria and other notations.
... A conceptual model should be a conscious construction to achieve a goal that goes beyond the making of the model itself, ranging from intermediate artifacts in systems analysis and design to sense-making and communication, model simulation, quality assurance, and requirements specification [34, p. 186]. An intuitive modeling language notation is important to achieve the goal of efficient communication [8,18,47] as it establishes the "first contact of the users with the modeling language" [13, p. 123]. A well-designed graphical notation is the first precondition for adoption and correct use of a modeling method [5]. ...
... Already in 1968, Howell and Fuchs were concerned with the challenge of developing "efficient signs or symbols for use in visual communication" [26, p. 310]. Their sign production method is based on a series of experiments aiming to identify population stereotypes, a concept first described [65] Existing notation This approach uses a simple matching technique where participants need to match notation with a term it represents [8,18] -Each participant draws one sample notation (I), which were then classified and statistically analyzed to identify Stereotypes. Participants then chose the best stereotype, thereby defining a new Prototype notation (III). ...
... [13] -Similar to [8,18], participants first draw samples which are then analyzed for identifying Stereotypes and Prototypes. Other participants then interpreted the prototypes [29] Existing notation Participants are shown sample models and asked to respond to comprehension questions [43] Existing notation Participants are shown sample models and asked to respond to comprehension questions [7] Existing notation The approach shows sample models with three or more alternative notations the participants need to choose [38] Existing models Participants are shown small sample models and need to respond to comprehension questions [61] Existing notation Participants needed to define the meaning they derive from a presented notation. ...
The notation of a modeling language is of paramount importance for its efficient use and the correct comprehension of created models. A graphical notation, especially for domain-specific modeling languages, should therefore be aligned to the knowledge, beliefs, and expectations of the targeted model users. One quality attributed to notations is their semantic transparency, indicating the extent to which a notation intuitively suggests its meaning to untrained users. Method engineers should thus aim at semantic transparency for realizing intuitively understandable notations. However, notation design is often treated poorly—if at all—in method engineering methodologies. This paper proposes a technique that, based on iterative evaluation and improvement tasks, steers the notation toward semantic transparency. The approach can be efficiently applied to arbitrary modeling languages and allows easy integration into existing modeling language engineering methodologies. We show the feasibility of the technique by reporting on two cycles of Action Design Research including the evaluation and improvement of the semantic transparency of the Process-Goal Alignment modeling language notation. An empirical evaluation comparing the new notation against the initial one shows the effectiveness of the technique.
... The principle is defined as: "…the extent to which the meaning of a symbol can be inferred from its appearance" [113]. When signs indicate low ST, not only does the cognitive load increase, but users also have difficulty to differentiate between concepts and need more time to learn a new language [56,57,114,116]. The principle has been recognized as essential to ensure that non-experts can better understand visual notation [96,117]. ...
... 2.2), we included all those synonyms in our search string. ST is supposed to increase cognitive effectiveness [56,57,114,116] and decrease cognitive load [24,116]. Thus, these are the outcome criteria we are investigating while focusing on results in published evaluations and applications of ST (which are the context criteria). ...
... 5.3.1. [99,137] Naturalness [64,117] Perceptual immediacy [12,114] Semantic distance [34,50] Visual/perceptual directness [55,116] Visual onomatopoeia [24,56] Closeness of mapping [111] Intuitiveness [17] Pragmatic quality [17] Readability [17] Number of studies listing synonyms 12 ...
Numerous visual notations are present in technical and business domains. Notations have to be cognitively effective to ease the planning, documentation, and communication of the domains’ concepts. Semantic transparency (ST) is one of the elementary principles that influence notations’ cognitive effectiveness. However, the principle is criticized for not being well defined and challenges arise in the evaluations and applications of ST. Accordingly, this research’s objectives were to answer how the ST principle is defined, operationalized, and evaluated in present notations as well as applied in the design of new notations in ICT and related areas. To meet these objectives, a systematic literature review was conducted with 94 studies passing the selection process criteria. The results reject one of the three aspects, which define semantic transparency, namely “ST is achieved with the use of icons.” Besides, taxonomies of related concepts and research methods, evaluation metrics, and other findings from this study can help to conduct verifiable ST-related experiments and applications, consequently improving the visual vocabularies of notations and effectiveness of the resulting diagrams.
... This study aims at designing symbols that are semantics-oriented and user-oriented (see Section 2) as outlined in [15]. Based on the suggestions and ndings presented in [13, 16, 17], we designed two studies to obtain graphical symbols for 11 security concepts. Subsequently, we evaluated the symbol set via expert interviews. ...
... The rst experiment addresses research question RQ1 to identify which symbols can be used to represent security concepts (see Section 3). For this purpose, we adapted the experiment design of the rst experiment presented in [16]. ...
... To answer research question RQ2, Experiment 2 is concerned with producing stereotypical symbols out of the sketches of Experiment 1 (adapted from [16]). ...
In recent years, business process models are used to define security properties for the corresponding business information systems. In this context, a number of approaches emerged that integrate security properties into standard process modeling languages. Often, these security properties are depicted as text annotations or graphical extensions. However, because the symbols of process-related security properties are not standardized, different issues concerning the comprehensibility and maintenance of the respective models arise. In this paper, we present the initial results of an experimental study on the design and modeling of 11 security concepts in a business process context. In particular, we center on the semantic transparency of the visual symbols that are intended to represent the different concepts (i.e. the one-to-one correspondence between the symbol and its meaning). Our evaluation showed that various symbols exist which are well-perceived. However, further studies are necessary to dissolve a number of remaining issues. © IFIP International Federation for Information Processing 2013.
... In recent years, we have seen several studies to identify language improvement opportunities, identifying problems with their concrete syntax and how they impact developer experience. These studies have covered a diversity of languages, including UML [16], [21], BPMN [22], [23], KAOS [24], [25], [26], i* [17], [27], [28], [29], OutSystems BPT [30], and SEA_ML++ [5]. Some of these languages were also analysed from the perspective of the impact of diagram layout in the understandability of models, namely UML [31], [32], [33] and i* [34]. ...
Context: The "Physics of Notations" (PoN) supports a systematic improvement of the cognitive effectiveness of visual modelling languages. Problem: PoN focuses on the concrete syntax of a language, building on a predefined abstract syntax. We should also consider the abstract syntax of a language when developing efforts to improve it by choosing the most adequate language constructs (concepts and their relationships). We instantiate this challenge by comparing two Multi-Agent Systems Domain Specific Languages: SEA_ML++ and DSML4MAS, and assessing the extent to which their respective constructs affect the developer experience. Method: We will perform a quasi-experiment for comparing how practitioners use both languages to solve similar modelling challenges. The experiment will have a cross-over within-subjects design and will focus on the extent to which the different language constructs impact on developer experience. These tasks will be monitored, so that we can assess their success and effort involved, including eye-tracking information. Results: This paper reports on the planned study design for this empirical comparison of two DSLs for MAS.
... Objective: Conduct a comparative experiment in order to evaluate the impact of using piStar in terms of performance and perceptions from a practitioner point of view. A4: iStar 2.0 notation and Moody's visual notation [15] [16] Description: The standard iStar 2.0 notation uses standard shapes (from the original version of iStar) to represent concepts. For example, circles represent actors, stadium shaped nodes represent goals, and hexagons represent tasks. ...
Conceptual modelling languages are typically taught in graduate and
postgraduate information systems programs. In the specific case of postgraduate
information systems students, the lecturer has the opportunity to exploit conceptual
modelling for educational purposes that go beyond learning how to apply the
modelling paradigm or specific languages. In this paper, we report on our employment
of the recently proposed iStar 2.0 in the master-level course Advanced
Research Methods at Utrecht University in the Netherlands. During this course,
the students conducted a design science project where iStar 2.0 artefacts are evaluated
in an experimental setting. We present the course (intended learning objectives,
learning materials, etc.), we explain how we employed iStar 2.0 therein,
and we discuss our teaching experience including lessons learnt.
... comparison; (ii) Accessibility to participants-finding the appropriate participants to the profile required by the study is difficult. Besides that, we have to consider the sample size; (iii) Researcher tendency-if the researcher is testing something that s/he has developed, some care must be taken not to be biased in planning and even analyzing the study results; Complementing these properties, we claim that in researches of evaluation and design of modeling languages, focusing on the visual aspect, the importance of the stakeholder involvement and application of empirical studies are significant, as in ( GENON et al, 2012) and ( CHO et al., 2012). So, they should be considered when evaluating the resulting modeling language, besides assessing some quality properties. ...
Em Modelagem Conceitual, tem ocorrido um interesse crescente em Linguagens de Modelagem Visuais Específicas de Domínio (Domain-Specific Visual Modeling Languages (DSVMLs)) e no suporte que elas provêem para compreensão do domínio de um problema e comunicação entre modeladores e interessados. Assim, é importante providenciar diretrizes para o design de DSVMLs. Por muitos anos, o foco de pesquisa tem sido na sintaxe abstrata, enquanto a sintaxe concreta tem recebido menor atenção. Isso é um infortúnio, pois a sintaxe visual impacta significativamente a capacidade de comunicação e de resolução de problemas de modelos conceituais Um dos trabalhos mais disseminados para análise e design de aspectos visuais de linguagens de modelagem é a Física das Notações (PoN). PoN define um conjunto de princípios usado para analisar e projetar notações visuais cognitivamente eficientes. Contudo, PoN tem lacunas, tais como: (i) Falta um método para aplicar seus princípios; (ii) O design de símbolos não cria sistematicamente símbolos que refletem entidades do mundo real. Nesta pesquisa, nós apresentamos a Física das Notações Sistematizada (PoN-S) para resolver a lacuna (i). PoN-S estabelece um conjunto ordenado de atividades de design e sugere quando aplicar os princípios de PoN. Ela também propõe grupos de princípios de PoN. Outra maneira de melhorar a qualidade de DSVMLs é a aplicação de teorias ontológicas, mas ontologias podem ser aplicadas com sucesso no design de sintaxes concretas como tem sido aplicadas para a sintaxe abstrata? Guizzardi (2013) propoem diretrizes ontológicas baseadas na Ontologia de Fundamentação Unificada (UFO) para auxiliar no design de notações visuais. Contudo, tais diretrizes também tem lacunas: (iii) São diretrizes isoladas e não parte de um processo de design; (iv) O conjunto de distinções ontológicas, é restrito; (v) As diretrizes ontológicas são restritas a serem aplicadas no estabelecimento dos símbolos de uma DSVMLs. Para resolver as lacunas (ii) até (v) nós combinamos as diretrizes ontológicas baseadas em UFO com PoN-S, originando a Física das Notações Ontologizada e Sistematizada (PoNTO-S). PoNTO-S é um processo de design sistematizado para sintaxes concretas de DSVMLs que conecta a sintaxe concreta com o significado do mundo-real (isto é, o significado ontológico). Este projeto é um processo de Design Science com diferentes iterações, cada uma produzindo artefatos próprios. O problema de design é o design de sintaxes concretas de DSVMLs. Os artefatos são melhorias de duas soluções existentes: PoN e diretrizes ontológicas baseadas em UFO. PoN-S e PoNTO-S são classificados como teorias de design, visto serem processo de design. Nós também investimos em estudos empíricos. Foram executados estudos exploratórios para dar suporte as indicações coletadas durante a revisão de literatura e guiar algumas decisões. Após desenvolver versões de PoN-S e PoNTO-S nós aplicamos novos estudos empíricos que geraram evidências para concluirmos que PoN-S e PoNTO-S são utéis, e que tais abordagens podem evoluir, dando origem a abordagens ainda mais úteis.
... Other studies focused on how to improve the notation of languages: works on I * [Genon et al. 2012], UML [Khendek SLE'17, October 23-24, 2017, Vancouver, Canada Marco Brambilla, Jordi Cabot, Javier L. Cánovas Izquierdo, Andrea Mauri 2015, WebML [Granada et al. 2017], and BPMN [Genon et al. 2011] examined the visual notations using the "Physics of Notations" theory [Moody 2009], highlighting problems related to their complexity. All of them conclude suggesting that the involvement of non-expert people could increase the cognitive effectiveness of the language. ...
Crowdsourcing has emerged as a novel paradigm where humans are employed to perform computational tasks. In the context of Domain-Specific Modeling Language (DSML) development, where the involvement of end-users is crucial to assure that the resulting language satisfies their needs, crowdsourcing tasks could be defined to assist in the language definition process. By relying on the crowd, it is possible to show an early version of the language to a wider spectrum of users, thus increasing the validation scope and eventually promoting its acceptance and adoption. We propose a systematic method for creating crowdsourcing campaigns aimed at refining the graphical notation of DSMLs. The method defines a set of steps to identify, create and order the questions for the crowd. As a result, developers are provided with a set of notation choices that best fit end-users' needs. We also report on an experiment validating the approach.
... For example, one could test how these BPMN security extensions are handled and utilized during the act of process modeling (e.g., on paper or on computer). Another experimental setup (e.g., [23]) could be the evaluation of hand-sketched drawings of people. ...
Enhancing existing business process modeling languages with security concepts has attracted increased attention in research and several graphical notations and symbols have been proposed. How these extensions can be comprehended by users has not been evaluated yet. However, the comprehensibility of security concepts integrated within business process models is of utmost importance for many purposes such as communication, training, and later automation within a process-aware information system. If users do not understand the security concepts, this might lead to restricted acceptance or even misinterpretation and possible security problems in the sequel. In this paper, we evaluate existing security extensions of Business Process Model and Notation (BPMN) as BPMN constitutes the de facto standard in business modeling languages nowadays. The evaluation is conducted along two lines, i.e., a literature study and a survey. The findings of both evaluations identify shortcomings and open questions of existing approaches. This will yield the basis to convey security-related information within business process models in a comprehensible way and consequently, unleash the full effects of security modeling in business processes.
... Unfortunately, the proposal is not adapted for an integration into large scale projects as the ones targeted by RUP/UML. Finally, the question of an alternative representation of i* elements has been envisaged in [7]. Their proposal is based on an empirical study of preferred icons; we point to the use of the ones that are already adopted in the industry by RUP/UML practitioners. ...
[Context and Motivation] Business modeling is nowadays a common approach in huge enterprise software developments. It notably allows to align business processes and supporting IT solutions at best, to produce a documentation of the company’s “savoir-faire” and to look for possible optimizations. The business modeling discipline of the Rational Unified Process (RUP) has enriched the semantic of the Unified Modeling Language’s (UML) use case diagrams for the special purpose of representing the organization’s processes with accurate elements. [Question/Problem] RUP/UML business use case semantics are nevertheless only intended to further stereotype use case models and not to be used for reasoning. In parallel and in line with artificial intelligence concepts, researchers have developed the i* framework enabling the evaluation and decomposition of multiple design opportunities. RUP/UML business use case semantics could be used more efficiently to integrate the latter benefits. [Principal ideas/results] Through a systematic mapping of elements from i* on the one side and of the RUP/UML business use case model on the other, we have set up a RUP/UML graphical notation for i* elements. Applicability has been shown on an illustrative example. [Contribution] The main contribution of the framework is allowing to model in an i* fashion using CASE-tools meant for RUP/UML and proposing an interface for forward engineering the produced model in a classical UML requirements model. Future work is required to fully validate the proposal, notably to measure the method’s efficacy.
Guaraná domain-specific language is aimed to design enterprise application integration models. Although the language has been broadly used in the industry and academy, the design of its graphical notation was based in the common sense and social opinion of its developers rather than theoretical principles and empirical evidence. Thus, Guaraná becomes as an option for studying, modifying, and improving its concrete syntax. In this paper we conducted an experiment with 85 participants in order to evaluate the semantic transparency of the concrete syntax of Guaraná. The study concludes that the language is not found to be semantically immediate because most of the language constructs used in the graphical notation of Guaraná are not semantically immediate, that is, a novice reader is not able to infer their meaning from their appearance alone.
Disponível em: http://repositorio.utfpr.edu.br/jspui/handle/1/4281
Esta dissertação propõe um método e uma linguagem de modelagem gráfica de requisitos de software e sistemas sob o nome de RIMON (Requirements and Interdependencies MOdeling Notation). Esta linguagem permite representar requisitos e suas interdependências de forma sistemática, precisa e expressiva, visando contribuir para a melhoria da qualidade da especificação de requisitos de softwares e sistemas. Originada a partir de conceitos da abordagem RON (Requisitos Orientados a Notificações), a RIMON foi criada para ser visualmente atrativa em possíveis usos comerciais. Essa atratividade decorre de um desenvolvimento fundamentado na teoria de física das notações (Physics of Notations – PoN), que fornece princípios destinados a produzir notações visuais eficientes em termos de comunicação. Adicionalmente, a linguagem é definida por meio de uma sintaxe abstrata (metamodelo) e de uma sintaxe concreta (sintaxe visual) complementada por um mapeamento semântico preciso. Em relação ao método proposto, este consiste em um ciclo iterativo de atividades para identificação e análise dos dados de entrada e modelagem gráfica dos requisitos e suas interdependências. A RIMON oferece características tais como suporte a modelagem de requisitos funcionais e não-funcionais, entidades, atributos, pré-condições, pós-condições e identificação de conflitos. Como diferencial, a modelagem de pré-condições possibilita representar restrições relativas a requisitos não-funcionais por meio de atributos contendo condições lógico-matemáticas. Por fim, este trabalho apresenta três experimentos de modelagem usando RIMON, tanto de requisitos de sistemas (extraídos da literatura) quanto de requisitos de software (projeto real), destinados a demonstrar e verificar as capacidades da linguagem.
Available in: http://repositorio.utfpr.edu.br/jspui/handle/1/4281
This dissertation proposes a method and a language for graphic modeling of software and systems requirements, under the name of RIMON (Requirements and Interdependencies MOdeling Notation). This language allows to represent requirements and their interdependencies in a systematic, precise and expressive way, aiming to contribute to the quality improvement of software and systems requirements specifications. Sourced in concepts of RON (Requirements Oriented to Notifications) approach, RIMON was designed to be visually attractive in its possible commercial usage. This attractiveness results from a development grounded in the Physics of Notations (PoN), which provides principles for creating efficient visual notations in terms of communication. In addition, the language is defined by an abstract syntax (metamodel) and a concrete syntax (visual syntax), complemented by a precise semantic mapping. Regarding the proposed method, it includes a set of steps for identification and analysis of input data, graphical requirements and interdependencies modeling. RIMON offers features such as support for modeling functional and nonfunctional requirements, entities, attributes, preconditions, postconditions and conflict identification. As a notable characteristic, the modeling of preconditions allows the representation of constraints related to non-functional requirements by means of attributes containing logical-mathematical conditions. At last, this work presents three modeling experiments using RIMON, including system requirements (extracted from the literature) and software requirements (real project) environments, designed to demonstrate and verify the capabilities of the language.
This paper presents a critical reflection upon the concept of 'physics of notations' proposed by Moody. This is based upon the post hoc application of the concept in the analysis of a visualisation tool developed for a common place mathematics tool. Although this is not the intended design and development approach presumed or preferred by the physics of notations, there are benefits to analysing an extant visualisation. In particular, our analysis benefits from the visualisation having been developed and refined employing graphic design professionals and extensive formative user feedback. Hence the rationale for specific visualisation features is to some extent traceable. This reflective analysis shines a light on features of both the visualisation and domain visualised, illustrating that it could have been analysed more thoroughly at design time. However the same analysis raises a variety of interesting questions about the viability of scoping practical visualisation design in the framework proposed by the physics of notations.
Goal-oriented modelling is one of the most important research developments in the requirements engineering (RE) field. This
paper conducts a systematic analysis of the visual syntax of i*, one of the leading goal-oriented languages. Like most RE notations, i* is highly visual. Yet surprisingly, there has been little debate about or modification to its graphical conventions since
it was proposed more than a decade ago. We evaluate the i* visual notation using a set of principles for designing cognitively effective visual notations (the Physics of Notations).
The analysis reveals some serious flaws in the notation together with some practical recommendations for improvement. The
results can be used to improve its effectiveness in practice, particularly for communicating with end users. A broader goal
of the paper is to raise awareness about the importance of visual representation in RE research, which has historically received
little attention.
KeywordsGoal modelling-
i*-Visualisation-Visual syntax-Evaluation-Visual notation-Visual language
Goal modeling languages are now an integral part of requirements engineering. They allow for the systematic capture of rationales for stakeholder needs and enable the reasoning about potential system solutions. The goal- oriented Requirement Language (GRL) is both a typical goal modeling lan-guage and an international standard, yet it suffers from many shortcomings, akin to deadly sins, that relate to its concrete syntax, semantics, approach to modularity, analysis, and extensibility. Based 011 10 years of experience using GRL, we discuss several shortcomings and point to relevant future work areas.
Requirements are usually understood as stating what a system is
supposed to do, as apposed to how it should do it. However,
understanding the organizational context and rationales (the
“Whys”) that lead up to systems requirements can be just as
important for the ongoing success of the system. Requirements modelling
techniques can be used to help deal with the knowledge and reasoning
needed in this earlier phase of requirements engineering. However most
existing requirements techniques are intended more for the later phase
of requirements engineering, which focuses on completeness, consistency,
and automated verification of requirements. In contrast, the early phase
aims to model and analyze stakeholder interests and how they might be
addressed, or compromised, by various system-and-environment
alternatives. This paper argues, therefore, that a different kind of
modelling and reasoning support is needed for the early phase. An
outline of the i* framework is given as an example of a step in this
direction. Meeting scheduling is used as a domain example
The present paper is concerned with the development of efficient signs or symbols for use in visual communication. In particular, it describes and illustrates a technique for generating such signs, and presents a series of experiments undertaken to evaluate this technique. Although the application of these methods was limited to a specialized set of concepts (military intelligence terms), and the test experiments for signs developed were highly selective, it was believed that the principles involved have implications for several rather divergent areas of psychology which share an interest in language and communication.
Subjects generating pictograms of abstract verbal concepts showed a preference for using concrete rather than abstract representations, even for very low concreteness values. As level of concreteness decreased, a greater variety of ideas was produced in the pictograms, but this was accompanied by a decrease in stereotype strength. There was a significant correlation between the judges' rankings of the best representation of the meaning of a concept and its stereotype strength. However, in about one fifth of these judgements, a unique idea (i.e. produced by only one subject) received the highest ranking, suggesting that concrete representation of an abstract concept may in some cases be related to creativity.
Visual notations form an integral part of the language of software engineering (SE). Yet historically, SE researchers and notation designers have ignored or undervalued issues of visual representation. In evaluating and comparing notations, details of visual syntax are rarely discussed. In designing notations, the majority of effort is spent on semantics, with graphical conventions largely an afterthought. Typically, no design rationale, scientific or otherwise, is provided for visual representation choices. While SE has developed mature methods for evaluating and designing semantics, it lacks equivalent methods for visual syntax. This paper defines a set of principles for designing cognitively effective visual notations: ones that are optimized for human communication and problem solving. Together these form a design theory, called the Physics of Notations as it focuses on the physical (perceptual) properties of notations rather than their logical (semantic) properties. The principles were synthesized from theory and empirical evidence from a wide range of fields and rest on an explicit theory of how visual notations communicate. They can be used to evaluate, compare, and improve existing visual notations as well as to construct new ones. The paper identifies serious design flaws in some of the leading SE notations, together with practical suggestions for improving them. It also showcases some examples of visual notation design excellence from SE and other fields.
Requirements are usually understood as stating whata system is supposed to do, as opposed to how it shoulddo it. However, understanding the organizational contextand rationales (the "Whys") that lead up to systemsrequirements can be just as important for the ongoing successof the system. Requirements modelling techniquescan be used to help deal with the knowledge and reasoningneeded in this earlier phase of requirements engineering.However, most existing requirements techniques are...
Goal-oriented modelling is one of the most important research developments in the RE field. This paper conducts a systematic analysis of the visual syntax of i*, one of the leading goal-oriented languages. Like most RE notations, i* is highly visual. Yet surprisingly, there has been little debate about or modification to its graphical conventions since it was proposed more than a decade ago. We evaluate the notation using a set of evidence-based principles for visual notation design. The paper identifies some serious flaws in the i* visual notation together with some recommendations for improvement. A broader goal of the paper is to raise the level of debate and stimulate discussion about visual representation in RE research.