To design a meaningful IoT-based assistant for public speaking, our team paired surveys with a Cultural Probe / Diary Study to capture both scale and depth. The survey offers structured data on challenges, habits, and preferences, while the diary study provides contextual, real-time insights. Together, they help validate findings, reveal behavior patterns, and inform human-centered design.

• What are the biggest challenges professionals face during public speaking?

• How do professionals prepare, and what tools do they use?

• What real-time feedback would support skill improvement?

• What are users’ preferences for an IoT-based assistant (wearable vs. non-wearable, feedback style)?

• What adoption barriers exist?

This research enables a deep understanding of user needs, ensuring that our IoT device concept addresses real-world challenges with features that matter most—like filler word reduction, real-time pacing feedback, and posture tracking. By grounding design in both broad trends and lived experiences, we aim to create a product that is intuitive, inclusive, and impactful.

We selected a diary study because it allows us to collect real-time, longitudinal insights into user behavior in a natural context. This method helps uncover patterns in how individuals prepare for and reflect on public speaking events. Since our project aims to help users improve their public speaking skills through personalized feedback, the depth and structure of diary entries align well with our goals.

In addition, our team is interested in understanding the motivations and emotions users experience during their public speaking journey. A diary study gives us insight into how users' preparation methods evolve as their speaking events approach, offering a holistic view of their process.

Relevant public speaking scenarios include:

• Presentations

• Team meetings

• Client pitches

• Conferences or seminars

• How do participants currently prepare for public speaking tasks?

• What are the most common challenges and pain points users face when speaking publicly?

• What do users need from a tool designed to improve their public speaking skills?

This is a diary study focused on capturing self-reported reflections on public speaking preparation, performance, and mindset. By examining these experiences over several days in real-life settings, we can identify patterns in habits, challenges, and emotional responses.

Type of Research

Primarily qualitative, with some quantitative components (e.g., rating scales).

Steps in the Process

1. Recruit 6 working professionals who frequently engage in public speaking tasks (e.g., meetings, presentations).

2. Conduct a brief onboarding session (virtual or in-person) to explain the study and expectations.

3. Distribute the diary (via Google Form or printed copy) for participants to complete daily over 5 days.

4. Collect and analyze all diary entries using thematic analysis.
   
   

Duration per session

Each daily entry should take approximately 5–10 minutes to complete, with 7 questions, 3 of which are open-ended.

Team Roles

• Moderators (Melp and Jay): Send reminders and guide participants throughout the process.

• Data Analysts (All team members): Review responses, identify themes, and synthesize insights.
  
  

Participant Interaction

Moderators will interact with participants before, during, and after the study via email, sending daily links and reminders.

Bias Prevention Measures

• Entries will be anonymized to encourage honest responses.

• Questions are neutrally worded to minimize leading participants.

• Researchers will perform independent (silent) analysis before group discussions to avoid bias in interpretation.

Recruitment Channels

We’ll use:

• Personal networks (friends, family, colleagues)

• Online professional groups (MSI Slack channel, Discord servers with working professionals)

Ideal Number of Participants

6 participants (with a minimum of 5 in case of dropout)

Eligibility Criteria

• Working professionals who engage in public speaking at least once a week, such as:

  • Presentations

  • Talks

  • Client meetings

  • Team discussions

• Did you spend time today preparing for a public speaking event? [Yes / No]

• Describe the public speaking event you were preparing for (include date, time, length). [Paragraph]

• How did you feel during your preparation today? (e.g., confident, anxious, overwhelmed) [Short Answer]

• How successful did you feel your preparation was? (1–5 scale)

• Describe how you prepared. What tools/methods did you use? [Paragraph]

• What challenges did you face during your preparation? [Paragraph]

• Did you have a public speaking event today? [Yes / No]

• How successful were you in the event? (1–5 scale)

• Explain why you rated your performance that way. [Paragraph]

• Did you notice yourself using filler words during your speaking? [Yes / No]

We’ll use thematic analysis to explore and categorize responses, looking for:

• Patterns in preparation methods

• Emotional states and confidence

• Shifts over time as the speaking date approaches

• Differences across types of events

Process

1. Break down entries into individual data points

2. Silent Grouping: Team members independently group data points into themes

3. Collaborative Discussion: Group reconvenes to refine categories and adjust clusters

4. Labeling Clusters: Assign descriptive names that capture the core insights within each group

Design Dimensions Matrix
Identified 3 core dimensions to guide scenario design:

• Automation Level (High to Low assistance)

• Public Speaking Scenario Type (Casual to High-Stakes events)

• Handling Unexpected Changes (Adaptability under pressure)

Speed Dating Matrix
Mapped user preferences and expectations across combinations of the three dimensions to guide scenario development.

User Enactment (UE) Scenarios
Designed 8 user enactment scenarios, each testing a specific interaction between user needs and PitchPerfect’s support system.

Props and Prototypes
Listed general and specific props to simulate realistic use of PitchPerfect in various contexts (real-time feedback, haptics, AR, etc.)

1. Adaptive Feedback

• Users may benefit from real-time spoken feedback in some cases (e.g., interviews or big speeches), but may find it distracting in others (e.g., networking).

• Flexibility in switching between modes (spoken, haptic, visual) is key.

2. Context-Specific Coaching

• Different situations demand different types of help:

  • High-stakes events → Structured + Real-time coaching

  • Casual settings → Discreet, minimal prompts

  • Unpredictable situations → On-the-fly adaptive tools

3. Personalization & Control

• Some users want control over their pace and style; PitchPerfect should support autonomy.

• Feedback should adjust to experience level and communication goal.

4. Managing Uncertainty

• Users benefit from PitchPerfect’s help during last-minute changes, noisy environments, and unexpected challenges.

• Scenarios test how AI tools can calmly assist users when stress levels rise.

5. Multimodal Interaction

• Incorporating different modes of feedback (voice, haptics, text, AR visuals) to match user environments and preferences.

6. Collaboration & Shared Experience

• PitchPerfect can support not just individuals but teams (e.g., group presentations) by offering personalized and collective feedback.

7. Emotional Intelligence in Coaching

• Emotion and tone awareness (e.g., practicing enthusiasm vs. professionalism) increases speaker impact and engagement.

1. Live Feedback Felt Disruptive
   Users found real-time verbal feedback during practice to be intrusive and distracting, often breaking their speaking flow rather than helping it.

2. Perception of Judgment
   Some users reported feeling judged or overly corrected when receiving live feedback, which negatively impacted their confidence and comfort.

3. Verbal Interruptions Break Focus
   Interruptions—especially spoken ones—were seen as breaking the user's concentration, making it harder to stay in the rhythm of their speech.

4. Feedback Lacked Usefulness
   Feedback was often not perceived as actionable or timely. Users wanted clearer, more specific suggestions that could be implemented immediately or reflected upon afterward.