Introducing our latest innovations
At EyeQuestion, we are committed to continuously improving our platform to make sensory and consumer research more efficient, insightful, and user-friendly. This year, we are excited to introduce a range of powerful new features designed to streamline panel management, enhance data quality, and optimize research workflows.
From advanced quota management and multi-appointment scheduling to AI-powered analysis and dynamic data cleaning, these enhancements are built to provide greater flexibility, efficiency, and accuracy. We’ve also expanded our reward tracking, panel filtering, and project design capabilities, ensuring that researchers have the tools they need to conduct studies seamlessly.
In this article, we’ll explore these key updates in detail and how they can elevate your research experience. Let’s dive in!
Quota
Quota is commonly used in market research and consumer research to recruit specific demographic groups for sensory and consumer studies. With this feature, users can set up filtering criteria within a survey or product-related questionnaire. These criteria may be based on personal attributes (such as age and gender), category-based questions, or general questionnaire responses.
Users can define the screened-in actions for panellists. For example, panellists who meet the criteria can:
- Continue answering the questionnaire
- Be redirected to another questionnaire
- Schedule an appointment
Monitoring quota progress and panellist status is simple through the Panel tab.
Once panellists have completed the questionnaire, users can analyze results in EyeOpenR (EOR). Some EOR analyses offer the option to split results based on quota, generating separate tables for screened-in and screened-out panellists.
Multiple appointments
Some research projects require panellists to evaluate products over multiple days, meaning they need to visit the facility several times. To streamline this process, we have introduced the multi-appointment functionality.
With this feature, users can enable panellists to book multiple appointments—either within the same day or across different days. Additionally, appointment scheduling can be restricted based on specific criteria:
- Same timeslot: Panellists can select a specific timeslot that applies to all their appointments.
- Same half of the day: If a panellist selects morning hours for one appointment, they will only be able to choose morning hours for the other days as well.
This functionality simplifies scheduling, ensuring a more structured and efficient booking process for both panellists and researchers.
Rewards export
When a panellist participates in a sensory or consumer study, they may be compensated for their participation. Our existing rewards functionality allows users to grant and pay rewards to panellists. To enhance reward tracking, we have introduced a new export type called Rewards Export that records all rewards granted and paid to panellists, providing a clearer overview of compensation history.
In Home Use Tests (HUTs), panellists may sometimes respond too quickly, indicating a lack of attention, or too slowly due to distractions or complex questions. In such cases, the collected data may be unreliable. To address this issue, we have introduced the Data Cleaning feature. Once data collection is complete, users can set minimum and maximum response times or use EyeQuestion’s recommended values to automatically flag inconsistent sessions.
Additionally, sessions can be flagged based on specific category data. For example, in a questionnaire about the liking of plant-based meats, a question may ask how often a panellist consumes them. If a panellist reports consuming plant-based meats only once a month, their data may be considered unreliable. These flagged sessions can then be disabled or deleted, ensuring they are not included in the final analysis.
AI Coder
To enhance the analysis of open-ended questions, we have introduced an AI-powered labeling improvement using a Large Language Model (LLM) called the AI Coder. This feature allows users to efficiently categorise panellist remarks by applying relevant labels. Users can either provide their own labels or select from a default list. The AI model then analyses the panellists’ remarks and the provided labels, automatically linking them to the corresponding responses. Additionally, the AI performs sentiment analysis, helping users determine whether a label is associated with a positive, negative, or neutral sentiment.
Finally, users can run a CATA (Check-All-That-Apply) analysis to calculate label frequency and assess statistical significance.
Panel filtering
EyeQuestion has a robust panel management tool, we have upgraded it by allowing creating static and dynamic panel based on restrictions on past and future appointments. This will ensure diverse group of panellists participate in a study. It is also possible to exclude panellists based on their participation in a specific project with panel filtering.
TCATA Dynamic Feedback and TCATA analysis
The Temporal Check-All-That-Apply (TCATA) method expands on the traditional CATA (Check-All-That-Apply) approach by introducing a dynamic feedback mechanism. With this enhancement, if an attribute is selected for longer than a predefined duration, panellists receive a visual reminder to deselect it if the sensory perception is no longer present. In the advanced settings of a question, users can define the maximum feedback time. Once this limit is reached, the selected attribute will glow at maximum intensity, prompting panellists to reassess and adjust their selection if needed. This feature helps improve data accuracy by encouraging panellists to provide real-time, precise feedback throughout the evaluation process. Read more about TCATA Dynamic Feedback and TCATA analysis.
New Designs
To effectively group similar samples together in project design, we offer a design type called “Balanced Blocks.” We have expanded this functionality by introducing three new Balanced Blocks designs:
1. Balanced Blocks (BB) Random Blocks: Similar products are grouped together, and similar samples are randomized within each group. Additionally, the blocks themselves are randomized.
2. BB Random Products: Each block contains random samples, and these samples are randomized within the blocks.
3. BB Random Products and Blocks: In this design type, both the products and the blocks are randomized.
