environment

In this project, we are asking, how can a technology-based product help reduce individual contribution to climate change? We conducted interviews,observations, and an online survey to assess how participants search for information regarding climate change, the motivations behind reducing personal carbon footprints, and the challenges they face when reducing carbon footprints. We used these findings to inform design decisions for a persuasive technology that would help individuals limit their personal carbon footprints. 

All observation participants were male; observations lasted approximately 30 minutes. We used a combination of the AEIOU (Activities, Environment, Interactions, Objects and Users) and Contextual Inquiry frameworks to scaffold our data collection and analysis. 

We organized our observation findings into three themes: (1) all participants’ queries were phrased similarly; (2) similar query patterns all yielded first-page search results littered with sources that appeared to be less than credible; and (3) two participants opted to use online carbon calculators which require input of complex information that we felt users were unlikely to have on-hand; this along with poor usability caused both participants to abandon the calculators.

Following our observations, we conducted interviews with four participants that included two males and two females; interviews lasted approximately 45 minutes. We transcribed our interviews and coded them using inductive analysis, which included open and systematic coding methods. We then grouped these codes to categorize our findings into four common themes: (1) motivation for personal action; (2) importance of convenience and expedience; (3) similarity of perceived challenges and obstacles; and (4) common features for a to-be designed technology.

We then administered a survey using surveymonkey.com. The survey had a total of 52 respondents recruited from the CDM participant pool, friends, and colleagues. We analyzed the results through a combination of descriptive and inferential statistical analysis. We tested a total of four hypotheses investigating differences in carbon reduction activity levels and feature preferences, age and carbon reduction activity levels, age and feature preferences, and climate change news encounters and carbon reduction activity levels. We also created two personas and scenarios based off interview and survey findings to help inform how different user types will interact with our web based application. 

The findings from the observations, interviews, and online survey survey provided insight for to several design implications for a to-be designed product to help individuals limit their carbon footprints. The major design implications included: (1) creation of a responsive web application for both desktop and mobile interfaces, (2) optimizing our web application for search engine visibility, (3) creation of an eye catching website that is considered credible (.org/.edu), (4) inclusion of an accessible carbon calculator for casual users with options for advanced settings feature, (5) visualization of data to demonstrate how your carbon footprint affects the environment, (6) inclusion of information about saving money through lowering your carbon footprint, and (7) inclusion of gamification features such as competition, achievements, and daily/weekly/monthly tasks and goals for lowering carbon footprints.

Methods 

In this section, we discuss our observations, interview, and survey methods. 

Observation

The following section presents our recruitment methods, participant demographic data, data collection, as well as data analysis methods for the observations.

Participants

We recruited four male participants among our friends, colleagues and co-workers; all were in their twenties. Inclusion criteria were (a) availability, and (b) a belief in the existence of climate change and that human beings are contributing to that change.

Data Collection Methods 

We conducted one-on-one observations (researcher: participant) that lasted about 30 minutes. All four participants used their own laptop. After obtaining informed consent, we asked warm-up questions focused on what they were most curious about related to climate change. We then gave participants a scenario, asking them to describe what they saw and read (see Appendix E for the complete observation protocol).

Throughout the observations, each researcher took notes using the AEIOU framework, identifying actions, environments, interactions, objects and users. After the observations were complete, we asked wrap-up questions and offered time for participants to ask questions. 

Data Analysis Methods

We grouped the observation responses from the warm-up and wrap-up questions using Trello. Using the AEIOU framework, those responses were used in conjunction with the observation data to create an affinity diagram using the online tool ‘Stormboard’. We the compiled a list of activities, environments, interactions, objects and users. In Stormboard we consolidated duplicates and grouped items that we felt consolidated into patterns. (See Appendix I for Stormboard Affinity Diagram)

Interview

The following section presents our recruitment methods, participant demographic data, data collection, as well as data analysis methods for interviews.

Participants

We recruited two female and two male participants from friends, colleagues, and classmates; participants ranged from late 20s to mid 30s. The inclusion criterion was a belief in the science of climate change and that human beings are contributing to that change.

Data Collection Methods 

We conducted one-on-one interviews (researcher: participant) which, on average lasted about 45 minutes. After obtaining informed consent, we affirmed that our screener had been applied and began our audio-recording.

We started with warm up questions which inquired about perception, motivation, and basic understanding of climate change terminology and concepts. We asked participants to describe climate change, share their interests, and describe what has motivated them to take action. We also asked about methods they had used for addressing climate change.

We then explored tools and resources that participants used to learn about carbon footprints and climate change. We asked reasoning questions, e.g.,  why one source of information was trusted over another. We closed the deep focus with questions about potential design features and functionality that may be useful in an app or on a website that would help them reduce their carbon footprint. We then asked about participants knowledge of related topics; the questions included experiential questions having to do with their community, social pressure, and choices they have made about their own carbon footprint (see appendix D for the complete interview protocol). 

Data Analysis Methods

We used inductive methods to analyze interview data. Each audio recording was uploaded and auto-transcribed using Temi.com. After transcriptions were completed, we stepped through them carefully to ensure accuracy. Final transcripts and audio-files were then shared with teammates using Google Drive.

Each researcher imported the Word document versions of the transcripts into Atlas.ti and coded for structure, attributes, and attitudes. Our team chose to use open and systematic coding methods with a focus on language and content. Initially, we individually coded all four interviews. We then grouped our codes and made a list of the most common themes. The themes served as a starting point for persona spectrums; we developed a total of five spectrums which highlight issues of frequency of action, depth of knowledge, impact, and sense of urgency. We created two personas based on the patterns we discovered in the five spectrums.

Survey

The following section presents our recruitment methods, participant demographic data, data collection, as well as data analysis methods for surveys.

Participants

We recruited 52 participants, spanning 18 years to 59 years old, with incomes ranging from $5,000.00 to $250,000.00 per year. The participants were recruited from the CDM participant pool, friends and colleagues. The inclusion criteria was a belief in the climate change and a minimum age of 18 years old.

Data Collection Methods

We created an online survey comprised of 24 questions. On average, it took 11 minutes to complete. We monitored the total number of completed surveys, closing the survey before we received 60 participants.

We began the survey by asking about interest level and search habits related to climate change. We then asked participants about levels of urgency and advocacy regarding climate change. We proceeded with questions about motivation and level of carbon reducing activity. A majority of our survey questions were written as Likert scale and matrix questions. For a full survey protocol see Appendix F.

Data Analysis Methods

We administered the survey using SurveyMonkey.com. A number of data points that we collected required minor manipulation to make it usable in our statistical tests which we ran using SPSS; year born>age> age group, income>income group, activity level>average activity level> activity group. With the results from “what year were you born”, our team converted the answers to an age, and then defined two age groups; 18-30 years old, and 31 years old or older. We developed two income brackets based upon the self-reported income data. Lastly, we calculated an average of all responses to question number 18, thirteen questions total. Once an average of the thirteen questions was calculated, we placed participant activity levels in three activity groups; low (2.5 and lower), mid (2.6 to 3.4) and high (3.5 or higher).

Once our data was collated, we were able to create a series of tests including; a Kruskal Wallis comparing activity level and importance features, Mann Whitney U measuring age groups and importance to features and activity level, and a Mann Whitney U measuring climate change news encounters and activity level.