Social Redundancy, Collective Learning, and The Influence of Architectural Design Features: a Literature Review

Introduction

I think it is important to note that my original research topic was aimed at the direct link between individual social learning, sustainability, and affordability within a co-housing community. Through my research, I discovered that social learning can be facilitated on a much larger scale than an individual, rather social collective and collaborative learning. Stevenson et al argue that if social and physical redundancy is created within a built environment than occupants will have the potential capacity to use said redundancy through co-producing and collective learning. Narozny et al believe that the framework for collective learning can’t be successful without understanding the barriers and opportunities of the learning process itself. Ergan et al, through a crowdsourcing experiment, finds that architectural features can have positive and negative influences on users, thus affecting the impact of a learning experience.

The purpose of this paper is to develop further research on the connection between human behavior, specifically social collective learning, and architectural design. The scope includes reviewing a quantitative study on the influence of architectural features, the need for physical and social redundancy, and the need for a social learning tool to be used in conjunction with present-day building evaluation surveys, such as Building Performance Evaluation (BPE) and Post Occupancy Evaluation (POE). Before we move on, I would like to define a few key terms regarding social learning that will aid in better understanding of this paper and help define the learning theories in the context of the reviewed literature.

• The Social Learning Theory (SLT), created by Albert Bandura (1977) assumes that people learn from each other through observation, imitation, and modeling. These three influencers must have the necessary conditions to be effective. These conditions are attention, retention, reproduction, and motivation.
• Collective learning involves the ability to share information between people efficiently and securely so that it can be passed down over time (Oxford Reference).
• Collaborative Learning involves two or more people who learn something as a group. Typically, those involved in collaborative learning tend to capitalize on each other’s skills and knowledge. What follows is a brief overview of the reviewed literature stating the premise, methodology, and findings. The first section that follows will be a combination and comparison of two articles, Stevenson and Narozny, as they share the same case study and both focus on the goal of collective learning. Proceeding that will be a review of the effects of architecture on human experience and the quantitative significance. Assumptions and limitations will be addressed sequentially.

Collective learning and a social learning tool

Stevenson et al (2016) used LILAC (Low Impact Low Living Affordable Community) as a primary example of how physical and social redundancy within a built environment can address the problem of perpetually shifting climate, an unstable economy, and reduced energy sources of the UK. The mostly qualitative research was conducted using questionnaires, participant observation, field study, house tours, and evaluations. The themes of physical and social redundancy were prevalent in the study. Regarding the physical aspect, LILAC provided emergency heating and water supply, independent emergency shelter, alternative non-electric ventilation systems. This physically built redundancy exists due to the collaboration of its occupant’s feedback.

Stevenson et al (2016) found that social redundancy refers to having “collected access to the accumulated experiences”. This means tenants were able to not only use their personal skills to help produce solutions but were also able to learn from each other resulting in a community melting pot of knowledge that will continue benefit them in the future. Narozny et al (2014) use LILAC as an example of the potential for collective learning opportunity through the initial use of a Social Learning Tool as an extension of BPE and POE. Using both quantitative and qualitative methods, such as Building User Satisfaction surveys and open-ended usability surveys, she explores the conceptual framework of collective learning in connection the need for a Social Learning Tool to better user experience and understanding of systems and their controls within a community. Themes include the role of motivation adjacent to the learning process and the decision-making process which is essential for collective home learning to exist. Narozny (2014) conducted a partial application experiment, focusing on the MVHR system and the usability of its controls based on user’s experience, to see if the application of a SL Tool as an extension of BPE would be beneficial to its users. Using a case study methodology, Narozny (2014) found that the occupants had a consistent lack of understanding how to use the system due to the ineffectiveness of LILAC’s hand over procedures, thus resulting in lower quality environments. The SL Tool, which specifically focuses on user skills and understanding, monitoring system meter readings specific to the MVHR system, decision making procedures and mapping home environment, pointed out barriers in LILAC’s system, but also allowed opportunities for improvement such as making users aware of the different uses of each system in order to prevent long-term bad habits based on incorrect assumption on how a home system is used.

LILAC vs LILAC, MVHR, and user experience

Both authors discussed the initial Home User Guidance (HUG) and the collective learning application in relation to the Mechanical Ventilation and Heat Recovery Unit (MVHR). LILAC’s hand over procedure involves a verbal how-to guide on systems throughout the community, including the MVHR, and is conducted by the subcontractor on a selected group of occupants. This information is then is passed on to community members; ideally on move-in-day, or at least within the first 8 months of occupancy.

Stevenson (2016) applauded this method of collective learning. Learning through observation and the being passed down over generations. This is the ideal definition of collective learning. He (2016) also stated that due to “comfort issues”, the community rejected the subcontractor’s advice that the MVHR should be running all the time to prevent mold due to the timber and straw construction of the dwellings.

Narozny (2014), argues that the reason for turning off the system was not because of “comfort issues”, but rather a lack of understanding of how to use the system. It was just easier to shut it off. According to her study, she discovered through her usability survey that those who did not feel competent using the system found their open ways to solve their issue, e. g. shutting off the system. This runs parallel with her research that the current hand over method being used is not working correctly but has the potential to be successful through adaptation of a social learning tool. Over the course of 15 months, a research team did a field study which included shadowing these HUG demonstrations. Through this, main issues were identified: lack of hands-on experience, trouble focusing due to environment or character of the person giving the demonstration, and lack of experience by the guide which could lead to misconceptions.

Stevenson stated that occupants went a step further with collaborative learning when they each shared with each other information on how they used their windows as ventilation systems in lieu of the MVHR. With this information, 17 configurations were shared and learned. He also stated that “comfort issues” caused by the MVHR encouraged the community to take action. This resulted in recommissioning the MVHR system, removing suspended ceilings, opening dampers in fan units, and instruction all occupants to use the fan above their stove when the cook. Narozny (2014) believed that if occupants knew how to use the system in the first place via the SL Tool then they wouldn’t have had to make any of these changes.

Stevenson et al’s (2016) study included data collected from the overall community while Narozny (2014) only focused on six dwellings (out of 20). This limits her ability to understand the rest of the occupant’s skills in relation to the user experience. However, assumptions by Stevenson included that all users had the same or higher-level skill sets and were able to understand and reproduce home demonstrations with ease. Both authors discovered similar a limitation: lack of time. LILAC occupants must have a collective desire to learn and to communicate to others their needs. LILAC provides a task management team to solve the issue of time. The benefit of this is that an issue is quickly resolved, and the resolution is documented, thus providing the next-generation with the same solution to the same problem. This is a good example of collective learning. However, the task management team focus on present individual cases instead of forward planning by studying how occupants consistently use their home, which could lead to future improvements both on the built environment and social collective learning. Through the SL Tool, we enable users with a better understanding of home use, thus creating a user with higher quality skills. Once the quality of a group is controlled, then we can begin to achieve the level of collective and collaborative social learning that both Stevenson and Narozny want to create.

Another aspect that can be controlled is the human experience within the built environment. Stevenson (2016) briefly discusses several of LILAC’s architectural features, such as an open kitchen and easily accessible central pond and garden, that supported social and physical redundancy. Semiha Ergan et al (2018) discovers the correlation between human experience and the influence of architectural design.

Evidence of architectural influence

Ergan (2018) examines the gap that exists between architectural design features and impact on the human experience. He believes, simply put, that buildings impact people. While current building standard measures, such as WELL buildings, exist to attempt to improve the future design, these do not address how architectural design features can enhance or diminish human experiences. Through a quantitative study, Ergan sets out to define the architectural design features that users are most responsive to in conjunction with a substudy of bipolar scales of each feature in relation to identifying preferred configurations.

The first step was to do define types of influence and architecture features to be used. Through extensive background research, vigorous literature reviews, and elicitation from 18 experienced architects, Ergan et al (2018) came up with four types of influences of architectural design features on user experience: Restorative, Stress/Anxiety, Aesthetics/Pleasure, and Motivation. While Ergan (2104) goes into detail about each user experience, motivation is the most critical in relation to this paper as it is a key concept linking each author. Motivation is directly associated with decision making and indirectly linked with social pressure and savings potential, specifically within a cohousing community. As stated earlier, motivation is necessary for collective learning. If future built environments can provide design features that enhance motivation, then collective learning can ensue, thus creating a social redundancy and physical resilience. Ergan et al (2018) define the architectural design features of motivation as a) color, texture, and materials of surfaces b) ease of access to space c) movement patterns and spatial connectivity. The positive end of each bipolar scale for each feature is thought to promote productivity.

After experiences were defined, architectural features were then created. Each architectural feature (i. e. presence/absence of light, the color of surfaces, etc. ) was decided on based on the features most utilized by the interviewed architects throughout their professional lives. Each feature has a bipolar scale that is rated based on experience. For example, the presence/absence of windows could be rated on a scale of “pleasant – unpleasant”. Participants were given dual images for each design feature set (e. g. symmetry/a-symmetry) and asked to define and rate their feelings in a blind survey. The feature being evaluated was not given to the participant. This information was then used as a measure of effect against a particular design feature. Responses were analyzed statistically to quantify human experience. Analytical computer software was used to evaluate the significance of results. Each architectural design feature and the significance between its bipolar configurations was as also analyzed through computer applications. Responses were formed both descriptively and analytically. Participant written responses revealed that the preferred configuration of the architectural features were the ones that related to the more positive end of the bipolar scales. For example, people liked the presence of natural light and large open windows as opposed to a windowless room with artificial lighting. Now that we know there is preferred configuration, analytical data quantifies the difference between preferences. Ergan et al (2018) found that the results of the analytical analyses showed that the ratio between the selection of the two configurations is statistically significantly different. Using partial eta squared measurements, each design feature was given a value and ordered based on influence and noticeably levels. Nature imagery and exposure to nature were listed as the top influencer on human experience. These findings serve as valuable information to practitioners that should be used, or at least considered, on future projects. This would ideally create the physical redundancy needed to affect housing resilience.

Key findings and notable commonalties

Stevenson et al (2016) and Ergan et al (2018) share one similar limitation: the belief that users/occupants/participants innately have or have acquired sets of skills and knowledge that give them the ability to perform or correctly communicate information. They assume the user experience is already understood. For example, a participant looking at an image of a crowded football stadium may have never been to a football game, thus their response to the architectural feature is null. When a user doesn’t know the answer because they don’t understand or have no experience regarding the question they don’t look for a solution. A Social Learning Tool as an extension of practicing building standards can facilitate in solving this drawback.

Key findings include:

• There is a direct correlation between architectural design features and user experience
• User experience regarding skill sets and knowledge varies. Using a SL Tool to measure these gaps to facilitate a capacity for collective learning is necessary.
• Collective and collaborative learning is effective in terms of social redundancy and can aid in sustaining a built environment.
• There is potential for a better system regarding understanding the user experience.