Creating professional questionnaires takes time, and choosing the right response scale can be a challenge even for experienced researchers. We want to make this process as easy and fast as possible, which is why we are introducing a new artificial intelligence-based feature to the StartQuestion platform: the "Generate responses" feature.
From now on, our AI assistant can step into the role of your personal quantitative research methodology expert and automatically suggest optimal answer options for your questions.
How does it work?
The new feature is already available in the questionnaire editor for the two most popular question types:
Single-choice questions
Multiple-choice questions
To use it, simply enter your question in the appropriate field and then click the new purple "✨ Generate responses" button located just below the answer fields.
The artificial intelligence will analyze the context of your question and propose a balanced set of answers in a matter of seconds. Our model has been instructed to provide options that are:
Professional and neutral – we avoid colloquialisms and bias.
Mutually exclusive – so the respondent has no doubts about what to select.
Balanced – the system is great at creating classic research scales (e.g., the Likert scale), tailoring them to the specific question.
By default, the system generates 3 to 5 short answers that best fit the given problem, which helps maintain high quality and readability of the survey.
What do you need to keep in mind? Limitations and security
We care about transparency, which is why, when implementing AI-based solutions, we have introduced specific rules and limitations that you should know about before using the new feature.
1. Replacing current answers
Using the "Generate answers" button will replace (overwrite) all the answers you have manually added to a given question so far. Use this option at the beginning of creating a question or when you want to completely change the existing scale.
2. Protection of personal and sensitive data
Answers are generated using a secure connection with the GPT model from OpenAI. However, remember privacy rules: do not enter personal data (e.g., names, surnames) or sensitive data into the question text unless you have an appropriate, documented legal basis to do so.
3. Quality and security filters
Our AI assistant is an expert focused exclusively on research. We have built rigorous filters into it:
If the entered text is not a question or is incomprehensible, the assistant will not generate answers.
The system completely ignores and blocks attempts to introduce profanity, offensive content, or programming code.
The artificial intelligence will not exceed the limit of 20 answer options per question (although in 99% of cases, it will suggest a few – the most optimal ones for the respondent).
Test it in practice!
The new feature is an ideal solution when you have a question in mind but lack ideas for comprehensive answer options. Log in to your account on Startquestion.com, create a choice question, and let the AI prepare the response options for you.