Feeling the Pull
Master's Thesis and Novel Co-creative Research Method
Research in design, emotion, and product experience has focused on establishing a connection between the aesthetic qualities of products and emotions. This research will ultimately assess the emotional values of 3D forms through the use of magnetic modeling; a methodological tool that allows participants to manipulate pieces of 3D forms into proposed emotional labels. The teapot was selected as the overall form to study based on its size, manual manipulation, user recognition, and potential usage as well as historical permanence. This study will evaluate how people interpret the emotionality of form in order to establish a veritable method for interpreting emotional variables throughout the design development process. The findings of this research establish prototypical associations in aesthetic traits and teapot piece combinations for each emotion category.
PROJECT CONTEXT
This project was completed as my Masters of Science in Design Research and contained both an applied project aspect and thesis literary review, as well as quantitative and qualitative research. My entire research process included an exploratory phase where the concept and secondary literature were evaluated for relevance, an image analysis of 500 teapot images, a card sorting task of 118 participants, and 13 videotaped interviews with participants using magnetic modeling to form emotions.
THE BIG IDEA
AESTHETIC PRODUCT EXPERIENCE
Everyday occurrences elicit emotions from the smell of a well-used coffee maker, to the smooth yet powerful touch of a rumbling orbital sander, or the view of the sculptural tail fins of a 1959 Cadillac slicing through the boulevard; people experience emotion as a result of the environment. Products in particular are soaked with experiential qualities as each feature imbues facets of influence throughout a lifetime of use. We have all had the experience where we see something and it sparks our interest, draws us in with its shapely qualities, makes us question what the object is, whether as a singular element or combined within the environment…the look of a product drives the initial step to building a lasting relationship.
"Aesthetic attributes are vastly regarded as the most important factor for attraction"(Adams & Van Gorp, 2012)
CONCEPTUAL FRAMEWORK
This model represents an exchanging relationship between the individual and the aesthetic attributes of the object or image. The resulting relationship is cyclical as the message is continually sent to the individual and aesthetic characteristics reinforce the continuous emotional perception of the object. This perception model was used to formulate the overall goal of the study.
RESEARCH QUESTIONS
As a majority of research into aesthetics and emotions was vastly limited to 2D representations of aesthetics qualities, I sought to break this barrier with magnetic modeling to make the results more verifiable to 3D forms. Additionally, I wanted to explore the perceptual biases one might have while creating 3D forms to expand the research in the field to understanding how people interpret emotional qualities in aesthetics on an intimate level. The following were my documented research questions.
1. Is there a difference between 2D and 3D emotional aesthetic perceptions?
2. What are the aesthetic elements that communicate particular emotions in form?
3. How do the perceptual biases of individuals alter the emotional perception of aesthetics?
EMOTIONS
CORE AFFECT
At the base of emotions, affect or an affective state, according to emotion theory, “is generally used to refer to all types of subjective experiences that are valenced, that is, experiences that involve a perceived goodness or badness, pleasantness or unpleasantness (Desmet, 2008b, pp 381).” Once it is determined whether the core affect is good or bad, it is then distributed across a level of arousal as organized by Yik and Russell (1980) below.
Core Affect (adapted from Yik & Russel, 1980)
THE ROLE OF EVOLUTION
Derived from evolutionary instincts, feelings, emotions, and other affective qualities guide us in our actions and choices. Through the evolutionary study of both humans and animals, there are prototypical patterns of behavior that are enacted during specified situations involving goal attainment. Below are the prototypical cognitive and affective processes of the emotions that are considered core to the evolutionary psychology.
Emotions Derived from Survival and Evolution (adapted from Drews, 2007)
PLUTCHIK'S EMOTION WHEEL
Plutchick (1982) proposed that there are “eight patterns of adaptive behavior (which) represent the basic dimensions of emotion applicable to all organismic levels.” These emotions result from one of an organism’s basic needs such as protection (fear) or reproduction (happiness) (Plutchik, 1965). Below is an adapted version of Plutchik's emotion wheel with the primary emotions represented in the most saturated colors.
Emotions Derived from Survival and Evolution (adapted from Drews, 2007)
THE PRODUCT RELATIONSHIP
SUBCONSCIOUS NOTIONS
Humans are driven by a layer of subconscious influences underlying every action. Physiological responses, genetics, cultural upbringing and sociological interaction combine to establish our personal cognitive constructs. As displayed in the following image, each product houses a series of influencing factors that affect the various realms of experience by the individual.
A Framework for User-Product Experiences (adapted from DiSalvo, Hanington, & Forlizzi, 2004, p.253)
PROCESSING PRODUCTS EMOTIONALLY
In Donald Norman’s book Emotional Design: Why We Hate or Love Everyday Things, he describes three main levels of processing that are performed when an object is observed, used, and considered. He describes the visceral level as “fast: it makes rapid judgments of what is good or bad, safe or dangerous, and sends appropriate signals to the muscles (the motor system) and alerts the rest of the brain (Norman, 2004, p.22).” This results in the behavioral stage which frames the use or the function and performance of the product. The topmost reflective level focuses mainly on the highest levels of feeling, emotions, and cognition.
Norman’s Three Levels of Processing (adapted from Norman, 2004)
MEETING APEX NEEDS
In Maslow’s classic Pyramid of Needs (image 1), once the basic needs of food, shelter, protection, and stability are met, individuals will seek higher standards of needs such as friendship, beauty, and truth(Giard, 2005, p. 56). Modeling after Maslow, Jordan's Hierarchy of product needs argued that consumer’s had similar needs in relation to products. According to Jordan, pleasurable designs make a product more desirable either through aesthetics, functionality, physiological sensations, or an ideological emotional connection.
Maslow’s Hierarchy of Needs (adapted from Maslow, 1954) and Hierarchy of Consumer Needs (adapted from Jordon 2001)
BUILDING A RELATIONSHIP
Emotional relationships are cumulative and can generate a powerful impact on the users connection with the objects they own. Current design imperatives strive for long lasting products that can establish meaningful first impressions that continue to build into stronger bonds over time. As seen below, tiny interactions and emotional experiences we incur on a daily bases impacts our overall emotional relationship with our products.
Responses > Experiences > Relationships (adapted from Demir 2008)
MAKE, SAY, DO
Liz Sanders, the creator of Make Tools advocates that participants create their perceptions using tools, games, or activities to portray concepts with Make Kits as seen below. Sanders emphasizes “the need to be directly involved in the creation and production of goods and services is pointing toward a human-centered design revolution, with the act of co-creation between designers and everyday people being the end goal (2002, pg.2).”
Make Tools Velcro Modeling Kit (Sanders, 2000)
Through such techniques as observation, designers can view what people do, through interviews designers will hear what people say, and through co-creation designers can interpret what people make. In particular, "making" allows for latent reservations, like emotions and feelings, to be revealed.
Say, Do, Make Scaffolds (adapted from Sanders 2002, p.1-3)
IMAGE COLLECTION
In an attempt to confirm prototypical teapot shapes, roughly 1,000 images from the Metropolitan Museum of Art website, Ebay, and other shopping websites were scrutinized for both antique and currently manufactured teapot and tea kettle forms from the year 1700 to the current marketplace.
ARCHETYPE TAXONOMY
Archetypes were chosen based off of how often they appeared in the images viewed and how unique their aesthetic qualities were considered. The length the archetype had appeared throughout time and the prevalence of the forms found in geometry also aided in deciding on the familial structure of the teapot taxonomy as geometrically derived forms were considered unique.
Rules were written for each archetype and then the collection was reduce and evaluated for form accuracy. After each image was re-evaluated, 101 images remained that strongly represented their categories according to their rules. The following lists the number of images per base: 18 hemispheres, 17 cones, 16 cylinders, 17 cubes, 16 spheres, 17 pears, and 16 cylinders. The 101 images were then coded using the following code so that correlations could later be identified.
CARD-SORTING TASK
Optimal workshop’s online card sorting software Optimalsort was utilized to complete the digital closed card sorting task remotely. Participants were directed through a series of instructions and demographic questions and then were instructed to organize 20 randomly sorted images from the total of 101 teapot image collection. Plutchik's 8 primary emotional labels were chosen for their term recognition, and emotion-theory derivation. The image below is an example of the task.
During this first experimental phase, a digital card sorting task was initiated by 118 participants and completed by 86 participants. The experiment was made available for three months. The average time for completing the task was 5.6 minutes. Participants were gathered from online (Facebook and email) and word of mouth promotions of the study via the primary and secondary researcher. In addition, two classes of students attending ASU’s College of Innovation and Technology were allowed extra credit for completion of the task. Participant demographics are summarized below.
MAGNETIC MODELING DEVELOPMENT
Magnetic modeling uses 3D printed pieces of an existing product shape that may be manipulated to represent specific concepts proposed by the researcher. Fifteen interchangeable teapot pieces were 3D printed in plastic and lined with neodymium magnets and steel bolts. Four bases, three spouts, four handles, and four lids were identified in the card-sorting task as strongly attributed to specific emotions. To account for the variations in attachment, 9 interconnection pieces were also developed to go between the spout/handle and the base of the teapot. All pieces were printed in plastic, sanded smooth, primered white, and then sealed. Below shows the final pieces and their functional capabilities.
MAGNETIC MODELING INTERVIEWS
Over the course of two weeks, thirteen video recorded interactions/interviews took place at the participants’ homes to ensure they each felt comfortable in their surroundings. The length of the interviews ranged from the shortest lasting 42 minutes and the longest lasting one hour and 48 minutes. Demographics from a preliminary survey are provided below.
During the experiment, participants were told to randomly select four cards from a deck of eight again containing the same emotions from Plutchik. The participants were instructed to model one teapot at a time with each model reflecting one emotion and they were encouraged to think allowed during the task. Once each teapot was constructed, a semi-structured interview took place and then the pieces were reset to allow for reuse. Below is a series of frames taken from one interview.
RESULTS
CARD-SORT PIECE FREQUENCIES
As frequency was recognized as a strong theme within the popular placement groups, the rate of occurrence for each piece was put into dimensional bar charts. The numbers represent the amount of images that were categorized most often in the corresponding emotion and are indicative of images with high selection rates by the participants. For example, there were seven images with a spherical base that had a strong relationship with joy.
CARD-SORT PATHFINDER NETWORK
The popular placements data from Optimal Sort was processed through the networking software Pathfinder which enabled a visual of piece relationships to their corresponding emotion placement. The links that are depicted in the following network show thematic distances between the images and emotions. Thus, the shorter the distance of the image to the emotion, the stronger the association between the two objects. The image was then colorized to visually depict the relationships more clearly.
MAGNETIC MODELING PERSONAS
While transcribing and coding interviews, personal cognitive biases were discovered in how the participants framed their construction of the emotions. Therefore, a sliding scale was constructed as a result of this phenomenon in order to accurately portray each participant’s range of perception.The participants generally had one strong perceptual bias paired with degrees of the other personas. Below is the scale, as well as an example.
Each persona has a core thematic trait that the participants exhibited during the interview through statements, justifications, and form combination results. Below is a generalized description of each persona followed by a series of video clips of each persona.
MAGNETIC MODELING IMAGE CLUSTER
The pictures taken from the magnetic modeling task were collected, sorted into their emotions, and then outlined with the color that matches the persona of the participant who modeled the teapot. The green represents the functionalists, the purple indicates the conceptualists, the pink shows the emotionalists, and the orange photos were modeled by the aestheticists.
2D VS 3D COMPARISON CHARTS
In order to provide a structured comparison of the card-sort and magnetic modeling tasks, I created a radial graphic in order to rate the dominance and stimulus characteristics of each emotion. The Pathfinder web and the modeling map emotion sections were rated separately and then the results were overlapped. In the images below, the lighter shaded image is the card-sorting result and the darker shaded image is the magnetic modeling result.
CONCLUSIONS
The results of the Card-sorting network and the Magnetic modeling cluster revealed which aesthetic attributes of the teapots communicated basic emotions. The content of these aesthetic traits were organized thematically as to coordinate which ones were deemed as positive, negative, dominant, or submissive. Below is a graphic depicting the thematic conclusions of the aesthetic content.
As shown in the following figure, all of cognitive perceptions layer upon each other to develop the emotional aesthetic. Piece selections occur within the individual’s perceptual bias which infuses the idealistic quality of the piece. Each layer of perception can add or subtract from the one before it and the larger layers project through all of the layers placed on top of them. Therefore, while applying emotional aesthetics, dominant features are considered first, followed by functional elements, and then topped with less significant details. The overall process is additive and collectively communicates the emotional qualities of the form.
PLACING THE PIECES
