Visual Instruments for Personal Finances
Creating immersive experiences through discovery, verification and understanding while handling personal financial information.
This thesis investigates optimized models for reasoning on complex information through the use of visual representation, exploratory interaction and collaborative evaluation. The work is grounded on existing research on visual analytics for learning and education, the role of interaction in information visualization and the insights gained from a user centered design approach.
Exploratory Tools are a series of connected interactive visualizations for portable, multi-touch capable devices. The aim of these visualizations is to facilitate accessible, memorable and actionable understanding and to transform this into applied knowledge. The present application is focused exemplary on the handling of financial information for laymen and encompasses theoretical fundamentals, personal data and the integration of a social layer for validation and comparison.
The influence of visualization and interaction attributes on the perceived understandability of information systems has been evaluated in a series of experiments. Subsequently, practical design guidelines were derived from the validated best practices. The aim of these principles is to guide future designers of such systems towards leveling the user’s capabilities with the complexity of the information domain.
We experience the multifaceted impact of money on our society and economy on a daily basis in our role as consumer, investor, professional or citizen. Whether it's controlling the household budget, planning the financing of education and provision or selecting assets for investment. In recent years numerous initiatives and programs have been launched around the world to promote financial education. This trend was reinforced by the most recent th financial crisis. The aim of these efforts is to increase financial literacy of broad sections of the population. Meanwhile countless governmental and private institutions for financial education, including international organizations, central banks, finance and education ministries, financial institutions and non-profit foundations are involved. Financial literacy allows us to better understand options and risks and therefore to make more informed and expedient decisions.
This is of increasing importance due to the continuously increasing diversity and complexity of available financial products. Moreover, an international trend regarding professional pension exists to transfer the liberty to decide and thus the risk to the individuals. A more careful and more conscious financial behavior of many individuals may also positively impact the national economy. The expected consequences include not only a reduction of personal debt and financial fraud, but furthermore an improved efficiency and stability of financial markets. In addition, a higher savings rate may lead to increased investment and economic growth. (Manz, 2011)1
The abstract nature of finance in conjunction with the myriad of options for earning, spending and saving and the unforeseeable impact of choices makes it hard for inexperienced people to make informed decisions. That's why, acting based on financial information requires for most people to collaborate with an intermediary such as a financial advisor. In this relationship, the laymen trade a share of their independency to gain security. The consequent dependency between them and the expert potentially deprives them from self-determination and exposes them to harmful overreaching. On the other hand, the unassisted access to raw information is, due to its' complexity, not sufficient to be understood, discussed and verified and thereby to be transformed into applied knowledge (Judelman, 2004).2 This kind of knowledge is necessary to foster a more critical handling of finances and consequentially enable an independent capacity to act for untrained people.
We are faced with two different challenges. The first is to adapt abstract information to fit the user's context and capabilities. Ideally, model abstractions should correspond to the viewer's mental models of a domain to aid reasoning. Reliable discoveries arise from people's ability to scrutinize both data and model, and to verify that a visualization shows real phenomena rooted in appropriate model assumptions (Chuang, Ramage, Manning, Heer, 2012).3 The second is to strike the balance between the user's personal knowledge about his financial situation and options and the user's trust in external assistance to guide him. To justify the trust in this kind of counsel, the counsel itself needs to be verifiable. Trust building verification of preprocessed information constitutes of the indication of the author, citation of the source of underlying data and through the assessment of the information with alternative parameters.
The essential question of this thesis is, how the content and representation model of personal financial information influences understanding for inexperienced people. Central to this is the question, what specific configuration of information is relevant, useful, and desirable to the user in the different contexts of handling finances. Concomitant, the question will be answered, how the visual and functional design of an instrument for the exploration, evaluation and communication of personal financial information needs to be, to facilitate an intuitively accessible, memorable, and actionable comprehension.
State of the Art
In-person counsel of an advisor, representing a financial institution like a bank, an insurance company or a private consulting firm, can be seen as the most wide spread way for people to get support in financial matters. Placing reliance on the expertise of a third party makes sense considering the complexity of dependency between decisions and their long-term impact. Furthermore, an intimate dialog between two humans enables honest discussions of sensitive topics like desires and fears. Psychological features like this need to be taken into account by the advisor to fully understand the person's behavior, goals and risk awareness.
Digital tools that support the registration, calculation and review of financial flow have evolved significantly over the past two decades. Harnessing the power of computing, connectivity and automation, they provide the user with more effortless means of tracking their financials. Solutions are provided by a range of different entities, ranging from e-finance services of banks to personal mobile, web- and desktop applications of independent software companies. If we look at the objectives of these applications, we are not only able to distinguish them by functionality and design, but also by their value proposition. Some tools are designed as utilities in the purest sense with affordances optimized for the user and mechanisms optimized for the subject matter. Others emphasize on achieving defined personal goals or changing spending behavior. They have systems in place to evaluate the user’s behavior by contextualizing and grading the collected data and putting the findings into perspective to the goals. Applications like this are built using persuasive design principles to positively influence the user’s decisions. It’s important to note that there is no strict separation between the two groups. We’re looking rather at a linear transition from one to the other. The position on this scale can be defined on a macro level, looking at the overall objective of the service, or on a micro level, looking at the individual components of the application (Risdon, 2012).4
Personal consulting has an unparalleled advantage in the qualitative assessment of a persons characteristics and the resulting financial situation and outlook. The remuneration model of a significant part of financial advisory services holds the potential for recommendations biased toward options that are lucrative to the financial institution. Tapping into the power of the collective intelligence by connecting the people seeking for advise could minimize this potential danger. Finding ways to share anonymized data over a neutral, trustworthy and community supported network would be beneficial to the individual and simultaneously build up a collective consciousness.
Technology supported data collection, processing and representation already empowers user to precisely track their financials. Few of the existing tools go further by providing actionable means of analysis, though. As user's context plays a vital role in how information is being perceived, finding the appropriate configuration of content and presentation is key for the application to be usable and useful (Lockton, Harrison, Stanton, 2010).5 Using information visualization that leverages the principles of visual perception has been proven to be indispensable in reading large and complex data sets. It is planed to investigate how the direct and directed interaction between the human and the visualization aids understanding of actions and consequences. Such actions should be testable in a lightweight manner to reduce the barrier of entry and to increase engagement. A playful, in the sense of engaging and motivating, experience could shift the user's state of mind while interacting with financial applications from 'Work' to 'Exploration', 'Discovery' or 'Learning'. The hypothesis of this paper is that, equipped with a coherent instrument that utilizes the above mentioned concepts, people with limited financial literacy will be supported to conduct a more informed, considered, critical and trained interaction with money.
Analysis of target audience
The sound understanding of the target audience forms the groundwork of all consecutive activities and will be achieved with the analysis of the human factors and the user scenarios. An in-depth investigation using contextual inquiry and cultural probes with a user group of five people will be used to articulate the user archetypes as well as their scenarios. The contextual inquiry happens in the real environment of user and will be documented using video, photography and transcriptions of the dialog. Cultural probes are self-conducted observations by the user of his or her behavior and will be documented using photography and a textual journal. This reveals why, how, when and where people are working with their finances and what information is most important for the tasks at hand.
Analysis of existing instruments
A collection, categorization and expert review of the current state of the art in research and industry, displays the discrepancy between demand and supply and by that, defines the room for innovation for the following process of conceptualization. As the existing solutions are diverse and access often is limited, expert reviews of the most successful options will give a concise overview.
Conception and development
With a user centered instead of a technology driven approach, a set of prototypes on multiple levels of fidelity will be tested in a series of experiments with the user group. These prototypes are built to satisfy different user stories and thereby analyze subsections of a holistic system for the access and handling of personal financial information. Each iteration will be subject to honest evaluation and therefore not only stand as an exemplary application, but also produce data about its' potential impact on the users' behavior and influence the consecutive development. Measurable metrics like performance and memorability as well as soft metrics like engagement, favorableness, sense of trust and security will be collected in parallel. The applied testing methods vary depending on the state of development and can include Competitive Testing, Walkthroughs or A/B Testing.
The evaluation of the prototypes will be conducted using a nested model for visualization design and validation with four layers: "The top level is to characterize the problems and data of a particular domain, the next level is to map those into abstract operations and data types, the third level is to design the visual encoding and interaction to support those operations, and the innermost fourth level is to create an algorithm to carry out that design automatically and efficiently" (Munzner, 2009).6 This nested model ensures that the content as well as the representation are error free and satisfy the identified user needs. The proven characterization of the task and data, abstraction of operations and data types, visual encoding and interaction techniques, and algorithms will be consolidated into a coherent application in the final phase of the development process. In a second round of user tests, the specific qualities of this application as identified in the prototypes and as proposed in the research question will be validated. The results from this user test will provide qualitative and quantitative data that forms the foundation for the consecutive discussion.
Verification with real users will prove appropriateness and usefulness of content and representation for inexperienced people, while expert reviews will ensure its' accuracy and utility. Based on the findings from this dyadic evaluation, a set of design recommendations for interactive visual analysis tools that support an intuitively accessible, memorable and actionable comprehension will be distilled. Although, the resulting instrument will be optimized for the exploration, evaluation and communication of personal financial information, it is planned to draw conclusions for a potential adaption and application in other domains. Personal financial information thereby serves as an instance of a topic area with an extensive taxonomy, multifaceted user groups that require specific means of access, and a polyvalent relevance in our society. Comparable areas can be found in healthcare, justice, or politics.
Manz, M. (2011). Financial Education – Rolle und internationale Entwicklung. Die Volkswirtschaft, 2011/6, 57-60. ↩
Judelman, G.B. (2004). Knowledge Visualization: Problems and Principles for Mapping the Knowledge Space. MSc Thesis Dissertation, International School of New Media, University of Lübeck, Germany. ↩
Chuang, J., Ramage, D., Manning, C.D., Heer, J. (2012). Interpretation and Trust: Designing Model-Driven Visualizations for Text Analysis. ACM Human Factors in Computing Systems (CHI). ↩
Risdon, C. (2012). Behavior Change as Value Proposition. Retrieved April 25, 2012, from http://adaptivepath.com/ideas/behavior-change-as-value-proposition. ↩
Lockton, D., Harrison, D.J., Stanton, N.A. (2010). The Design with Intent Method: a design tool for influencing user behaviour. Applied Ergonomics, Vol.41 No.3, 382-392. ↩
Munzner, T. (2009). A Nested Model for Visualization Design and Validation. IEEE Transactions on Visualization and Computer Graphics, Vol.15 No.6, 921-928. ↩