Introduction

Conversational UIs (CUI) is the recent buzz; tech-wise, design-wise, and interaction-wise. We all are a part of a constellation of technological artifacts that can just be interacted with without physical touch, gestures, or movement, i.e. we can talk to them.

Problem

We were given an open prompt to address conversation user interfaces(CUI) from the perspective of JEDI. Prompt did not mention any specific goals or problems related to Conversation User Interfaces.
How can we make CUI by, for, and about JEDI (Diversity, Equity, Inclusion, Justice)

Insights

Most of our users stated that they were struggling to find the structure for their CUI design and they spent more weeks searching use cases and the user groups. Some users even said the online instructions for CUI are very generalized and they ended up confused with a lot of data.

Success Goals

Make research goals easy to locate and easy to read
Educate non-designer employees on the importance of JEDI
Foster the design for accessibility culture in the whole team and design process

Solution

The Outcome of this explorative project is a tool kit that will aid designers to consider and reflect on designing JEDI (Diversity, Equity, Inclusion, Justice) oriented conversation user interfaces.

INTRODUCTION CARDS
A set of cards to introduce designers with Conversation User interface and JEDI (Justice, Equity , Diversity, Inclisivity) principles .

VISUALISE WORKSHEET
A worksheet to establish and visualise the problem space and relation of JEDI with the CUI in order to know the problem and goals of the project .

ACTIVITY WORKSHEET
A worksheet to understand common interactions, determine a detailed map of user groups, and prepare designers for expert interviews with the excluded user groups.

CARD SORTING + EVALUATE
Set of cards that consists of various user groups , use cases, things to consider, evaluation points

As the first step of our investigation, we conducted secondary research through literature, journals, paper, videos, and articles.
The purpose of secondary research was to gather existing data and familiarize with current notions or predispositions around conversation user interfaces (CUI).
‍

CUIs are not accessible to extreme users such as people with limited abilities, young & old people.
CUIs are biased towards native English accents and lack datasets for other languages.
CUIs commonly confers to only two genders
CUIs fails to be adaptable according to users' need and contexts such as personal characteristic and situational state.

Accessibility

CUIs are not accessible to extreme users such as people with limited abilities, young & old people.

Existing Data Bias

The dataset used to train CUI algorithms consists of many biases such as race, languages, accents that are designed to best serve the majorities.

Context in Conversations

CUIs are not designed to consider various context in conversations such as age , emotion and utterances.

Adaptable and Personalized

CUIs fail to be versatile and adaptable as they do not account for variations in personal characteristics and emotional state

The prompt was too broad and we were asked to narrow it down to a specific use case and target audience given the project timeline and limited resources. We spent countless hours brainstorming just trying to find a small segment within JEDI for conversational UIs but found that there were too many moving parts to the problem and selecting one was turning out to be impossible.

We noticed that every team working on this design prompt was struggling with choosing a direction for the project. The lack of resources online the entire process much more difficult. Hence we decided to pivot from solving for a specific use case to helping designers behind these conversational UIs throughout their entire design process.

Through more research from journals and articles, we understood novice designers are trying to adjust and adapt to new tasks, methods, and tools with the introduction of a new design challenge. Designers are often searching for design methods to make their design process effective and managing their timelines a lot easier, so there’s the knowledge that designers are more likely to rely heavily upon while designing solutions.

We conducted semi-structured interviews with designers from various levels of expertise in this field. we chose new designers as our demographic because our goals during our interviews were to:
‍
1. Understand their current process and methods used to design conversation user interface.
2. Dive deeper into the problems designers encountered when trying to understand Conversation UI and JEDI Principles.
‍

“Till the time we understood CUI and JEDI it was already 4 weeks pass the timeline”
- Student 1

“We have so many methods online,but how do we apply that for JEDI and specifically for CUI” - Student 2

From our interviews and experience mapping, we observed a pattern where most users faced difficulties during the research and narrowing down stages of the design process and that there is a clear need for a method or a process that can guide designers in designing conversational interfaces in the aspects of JEDI. Some of the support that we observed the designers needed are:

Key Takeaways

1. Designers did not know the tangible forms of JEDI because the information available was vast and it is was challenging to narrow down to a particular user group or use case.
‍
2. Almost all the designers could not envision how CUI feels like in terms of JEDI
‍
3. Designers did not know how to approach the problem space.

I conducted a brainstorming session with my team and came up with 3 main opportunity spaces we identified based on inputs from stakeholders and our research including interview and experience map analysis.

Ambiguity

Designers don’t know where to start looking and are overwhelmed by the vague, unorganized information

Complexicity

Young designers need a lower barrier of entry for Designing CUI from the DEIJ perspective.

Inclusivity

Users don't know how to be inclusive of the DEJI principles while designing DEIJ.

We plotted a user journey map to identify different entry points where designers felt either overwhelmed or confused about further steps while designing the conversation user interface. We made use of the Miro board as it is accessible and more collaborative within the team. The black sticky note below shows the major output in the process.

User Journey

Concepts

I decided to conduct conceptual relays to further brainstorm concepts. Basically, one member of the team presents a quick concept based on his understanding of the problem. The second team member then adds to that concept and then the other teammate does the same. This allows every team member's ideas to be heard and discussed. After completing a round of the conceptual relay we came up with 3 interconnected concepts.
‍
Though all concepts were a possible way to help designers to reflect and encourage accessible and universal design, we decided to work on Concept 3 as one of the important pain-point was that the designer was clueless about how to start the CUI Design from a JEDI perspective.
‍

Concept 1

Create a video tutorial series that contains some common information about CUI and DEIJ. Video can be used to inspire young designers to be more considerate of JEDI.

Concept 2

As part of this activity, we help the designer arrive at principles for his user group in the aspects of DEIJ in designing a CUI by answering a set of predefined questions via cue cards

Concept 3

An activity-based of worksheets and a set of reflective card decks to arrive at the intended solution. Create a design process that can guide and make designers reflect and think from JEDI's perspective.

From our user journey map, we found 3 main intervention points. So we started drawing initial drawings of how the worksheets will look. We performed another round of secondary research to understand UX method kits available online. The main reason was to get an idea about the visual aspects of how to worksheets look and common patterns that follow to using them.

Whiteboarding on miro

We started laying out structures for the worksheet and also discuss n what kind of content should go on these worksheets. Once we had our initial drafts of the worksheets and cards, we performed user testing to understand the interactions and user experience of the worksheets. We asked users to use the worksheets and we observed where the users were facing difficulties and getting stuck. The major issue we found was that the instructions were not clear and the visual flow of the worksheet was hard to follow.
We decided to use different combinations of F - pattern & Z - Pattern of visual hierarchy to iterate our worksheets.

1. The initial worksheet & process was too long and confusing. It was explaining unnecessary information.
2. It wasn't an option for users to skip from onboarding, so I included the skip instructions.
3. On the worksheet, users prefer to have information either up/down or left/right on the worksheets.

The toolkit consists of 4 phases containing worksheets and cards to help the team to establish the problem space in the scope, user group and relate JEDI with the CUI in order to know the problem and goals of the project effectively.
Prerequisite: Designers go through the Orient Card that understands the basics of CUI and JEDI.

W1: Establish the Design Space (supported by User cards, Use case cards, and considerations cards)
W2: Understanding Interactions
W3: Determining Persona Spectrum and Persona Network
W4: Experts Interview

Worksheet 1

Establish the Design Space

Why do you use this activity ?

REFLECT & VISUALISE as a team to establish the problem space in the scope, user group and relation of JEDI with the CUI in order to know the problem and goals of the project effectively.

When to use this activity ?

We recommend using this activity as early as possible. This will allow team to have a clear idea about the problem and the include JEDI thinking from the beggining of the project.

Worksheet 2

Understanding Interactions

Why do you use this activity ?

UNDERSTAND as a team some common interactions that is intended with the bot and relate it with the common human interaction.

OBSERVE & DISCUSS challenges & Friction between the two roleplays to identify opportunities and improvements.

When to use this activity ?

We recommend this activity after you have completed the Establish phase and have thought of some common interactions for the CUI. If you haven’t, think of a few common interactions for 5 minutes. before this activity

Worksheet 3

Building the Persona Spectrum

Why do you use this activity ?

DETERMINE as a team, a user group and their connected Persona Spectrum & Persona Network.

Teams can use the User Card Decks for brainstorming the excluded User group

When to use this activity ?

We recommend this activity to DETERMINE the the excluded user group and their connected persona network that will be the intended user of CUI.
This Activity will help the team during the Expert Interview phase when you empathize with these excluded user groups

Worksheet 4

Expert Interviews

Why do you use this activity ?

Now that you have your User Persona & User Spectrum, Perform EXPERT Interviews to get a better understanding of their needs and pain points. The instruction in the worksheet will help your team to initiate the complete process .

When to use this activity ?

Getting insights from the Excluded user in the beginning phase of the CUI design can help designers to think from the JEDI Perspective.
As Microsoft mentions Design for Exclusions and in turn, you design for all.

Support Card Deck

Why do you use this activity ?

When to use this activity ?

1. I gained an in-depth practice of all processes for UX Design, except for formative usability testing because the format of our solution is hard to simulate in a test set.
2. For future steps, I would like to conduct in-depth usability testing together with more designers to iterate and refine our assets and also work on visual design.
3. A ToolKit of this scale requires a lot of time & research to make it more extensive and universal.
4. This project has had a special place in my heart since I got to work and understand JEDI(Justice, Equity, Diversity, Inclusion) Principles that will help me to design more accessible for my future projects.
5. Explore options to make it more general and not limited to the conversation user interface