If you're new to Research Flow, this glossary explains the main terms you'll see when creating a study, recruiting respondents, running sessions, and reviewing results.
Tip! To find a specific term quickly, use Command + F on Mac or Ctrl + F on Windows.
Core terms
| Term | Definition |
| Study | The complete set of activities involved in conducting and analyzing research in Typeform, led by an AI-moderator. This includes tasks, questions, insights reports, and study settings |
| Researcher | The user of Typeform/Research Flow who creates and manages the study. |
| AI moderator | The AI that interacts with respondents during a session. |
| Train/Context | The background and instructions you give the AI moderator so that it knows how to interact and behave during the session. |
| Research goal | The main question or topic that the researcher wants to answer by conducting the study. |
Recruitment terms
| Term | Definition |
| Panel recruitment | The service you can use to recruit respondents through Typeform’s built-in recruitment partners. |
| Recruitment group | A set of respondents selected using specific audience filters. |
| Target audience | The group of people you want to learn from, such as current customers or prospects. |
| Respondents | The people who participate in the study. |
| Panel | A pre-recruited group of people who have agreed to take surveys or research studies. Researchers use panels to find qualified respondents faster. |
| Sample | The qualified group of respondents who take part in your study. It represents the larger population that you want to learn about. |
| Sample size | The number of people included in your study. In general, a larger sample size can give you more confidence in the results. |
| Gen Pop | Short for general population. It means a broad audience of adults, rather than a narrow or very specific group. |
| Quota | The target number of responses from a specific subgroup. (For example, 100 responses from customers and 100 from prospects). |
| Subgroup | Part of your target audience that shares something in common, such as being customers, prospects, or people in a specific role. |
| Incidence rate | The percentage of people who qualify for your study after answering the screener questions. |
Session setup terms
| Term | Definition |
| Session | A session is one end-to-end interaction between the AI moderator and a respondent. A session can be Active, Partial, or Complete. Active: The session is currently in progress. Partial: The respondent started the session but did not finish it. Complete: The respondent finished the session. |
| Sessions | Sessions are the full list of sessions (interactions between the AI moderator respondents) collected in a study. |
| Session format | A session format is how a respondent can answer questions: through text, voice, or voice and video. |
| Session guide | A session guide is the ordered list of steps and instructions the AI moderator follows during a session. |
| Session tasks | A session task is a single part of a session, made up of instructions the AI moderator follows. For example the Welcome, Screener, Greeting, Questions, and Ending. |
| Personality | Personality is the set of instructions that shapes how the AI moderator interacts with the respondent. |
| Moderator voice | Moderator voice is the feature that lets the AI moderator speak its questions out loud instead of showing text only. |
| Research question | A research question is a specific question the AI moderator asks during a session. You can choose the answer format and add follow-up questions. The AI moderator keeps the main structure of the question, while allowing for some flexibility in tone and context. |
| Follow-up questions | Follow-up questions are extra exploratory or probing questions based on the main research question and the respondent’s response. You can control how in-depth follow-up questions should be. |
| Stimuli | Stimuli are visuals shown to respondents so they can share feedback on things like packaging, branding, or other concepts. |
| Screener questions | Screener questions are questions asked before the session to decide whether a respondent qualifies to take part. This helps ensure your insights are based on responses from people who fit your study criteria. |
Response and session terms
| Term | Definition |
| Respondent experience | The session experience from the participant's point of view |
| Recruitment in progress | This means the study is live and actively collecting responses. |
| Completes | The number of fully completed and usable responses in a study. |
| Completion rate | The percentage of people who start a study and finish it. A low completion rate can signal that the study is too long, confusing, or poorly designed. |
| Soft launch | A small test run of a study before fully launching it. It can catch issues like confusing questions, broken logic, or poor-quality responses. |
Insights terms
| Term | Definition |
| Insights report | The report created from your study data to help you review the findings. |
| Executive summary | The short recap of the full insights report. |
| Highlight reels | Curated, shareable collections of key moments from one or more sessions. They’re made from selected video and audio highlight clips to tell a clear story or highlight a theme, and they connect back to the executive summary in the insights report. |
| Highlight clips | Short excerpts from a session—video, audio, or transcript snippets—that capture a meaningful quote, reaction, or moment related to the research goal. |
| Highlight mentions | Automatically detected moments where a response strongly relates to a topic or insight. They’re surfaced as suggested highlights, often as a quote or clip reference, for review and inclusion. |
| Topics | AI-generated themes that group related responses across sessions, making it easier to analyze patterns and find relevant responses. |
| Sentiment | The AI’s interpretation of the emotional tone of a participant's response, such as positive, neutral, or negative, often with a confidence score. |