QuestionPunkQuestionPunk
Start free
All templates

AI Therapist Chatbot Experience and Trust Survey

Measures how people actually use an AI therapist chatbot, how helpful and trustworthy they find it, and whether it has ever felt unsafe or judgmental — for product and clinical safety teams. The AI follow-up interview reconstructs a specific real conversation instead of relying on vague satisfaction ratings.

Use this templateBrowse more templates

Sample questions

A preview of what’s in the template. Every question is editable before you launch.

13 questions · ~7 min
Q01
Message

Hi! We're trying to understand how people actually experience (Replace with your chatbot's name) — the good, the unhelpful, and anything that felt off. This takes about 5 minutes and your answers are confidential. There's no judgment here, honest answers help us the most.

Q02
Multiple ChoiceRequired

In the last 30 days, how often have you used this chatbot?

  • Daily
  • A few times a week
  • About once a week
  • A few times this past month
  • This is my first time using it
Q03
Multiple ChoiceRequired

What do you typically use the chatbot for? Select all that apply.

  • General stress relief
  • Managing anxiety
  • Coping with depression or low mood
  • Working through a specific problem or decision
  • Loneliness or wanting someone to talk to
  • Filling the gap between human therapy sessions
  • Crisis or urgent emotional support
  • Just curious or testing it out
Q04
MatrixRequired

How much do you agree with each statement about your experience with the chatbot?

5 rows × 5 columns
  • It understood what I was trying to say
  • I felt judged by its responses
  • I trust its suggestions almost as much as advice from a person
  • I feel comfortable sharing sensitive or embarrassing topics with it
  • I would recommend it to someone who is struggling emotionally
Columns: Strongly disagree · Disagree · Neutral · Agree · Strongly agree
Q05
Opinion ScaleRequired

Overall, how helpful has (Replace with your chatbot's name) been in helping you cope with what you're going through?

Scale: 1 – 7
Min:Not at all helpfulMax:Extremely helpful
Q06
Best–Worst Trade-off (MaxDiff)Required

Which of these matter most to you when you choose to talk to the chatbot instead of a person?

  • Available 24/7
  • No cost or low cost
  • Anonymity and privacy
  • Non-judgmental tone
  • Immediate response, no waiting
  • Personalized suggestions
  • No appointment or waiting list needed
  • Feels like talking to a real therapist
Pick best & worst per setBest:Matters most to meWorst:Matters least to me
Q07
Multiple ChoiceRequired

Has the chatbot ever given you a response that felt unsafe, inappropriate, or unhelpful for what you were going through?

  • Yes
  • No
  • Not sure
Q08
Opinion ScaleRequired

How likely are you to recommend this chatbot to someone else facing similar struggles?

Scale: 0 – 10
Min:Not at all likelyMax:Extremely likely
Q09
AI Interview

If the respondent flagged a response that felt unsafe, inappropriate, or unhelpful, reconstruct exactly what they typed, what the chatbot said back, and what they did next — ignored it, sought a human, kept using the app, or stopped. If they didn't flag one, probe the single most meaningful conversation they've had with the chatbot: what they were going through, what it said that landed or fell flat, and what a human therapist might have done differently in that same moment. Push past generic 'it helped' or 'it's fine' answers for concrete detail.

Q10
Multiple Choice

If a human therapist were just as affordable and available as this chatbot, what would you do?

  • Switch to a human therapist
  • Keep using the chatbot instead
  • Use both
  • Not sure
Q11
Multiple Choice

What's your relationship to human therapy right now?

  • Currently seeing a human therapist
  • Have seen one in the past, but not currently
  • Never seen a human therapist
  • Prefer not to say
Q12
Multiple Choice

Which age range do you fall into?

  • Under 18
  • 18-24
  • 25-34
  • 35-44
  • 45-54
  • 55-64
  • 65+
  • Prefer not to say
Q13
Message

Thank you for sharing this with us — especially anything that felt uncomfortable to admit. Your answers will be used to make the chatbot safer and more genuinely helpful, and any flagged safety concerns will be reviewed by our clinical team.

What’s included

  • AI follow-ups

    Adaptive probes on open-ended answers that pull out detail a static form would miss.

  • Attention checks

    Built-in safeguards against rushed answers and low-quality respondents.

  • AI-drafted copy

    Wording, ordering, and branching written by the AI — tuned to your research goal.

  • Auto report

    Themes, quotes, and a plain-English summary write themselves once responses come in.

How it compares

We reviewed the closest templates from other survey tools. Here’s what they do well — and where this template goes further.

Why this template

  • Reconstructs a specific real conversation the respondent had with the chatbot via an AI follow-up interview, rather than asking them to rate 'satisfaction' in the abstract
  • Directly asks whether the chatbot ever felt unsafe, inappropriate, or judgmental, and routes flagged respondents into a deeper AI-driven follow-up on that exact incident — built for clinical safety review, not just product feedback
  • Combines usage frequency, use-case selection, a matrix of trust/agreement statements, and a max-diff on why people choose the bot over a human, giving product and clinical teams multiple angles in one instrument
  • Closes with context on the respondent's relationship to human therapy, so 'would you recommend this over a therapist' answers can be interpreted correctly

SurveySparrow

Therapist Chatbot | Template for Counsellors

This is a static, fielding-ready survey template aimed at counsellors gathering feedback on a therapist chatbot, so it's a genuine topical match. It appears built around fixed question sets rather than reconstructing individual conversations, and is framed more generally for counsellor use than specifically for product/clinical safety teams flagging unsafe responses.

What it does well

  • Purpose-built specifically for therapist chatbot feedback, giving it direct topical relevance
  • Framed for counsellors, suggesting language and question framing suited to a therapeutic context
  • Likely quick to deploy given SurveySparrow's template-driven form builder

Where it falls short

  • No indication of adaptive AI follow-up questioning — respondents who flag an unsafe or unhelpful moment likely can't be probed further on that specific exchange
  • No visible mechanism for reconstructing what actually happened in a specific conversation, relying instead on standard rating-style questions
  • No published methodology or prompt transparency for how any AI-assisted elements (if present) generate or score responses

Ready to launch?

Open this template in the editor. Every part is yours to change before the first respondent sees it.

Use this templateStart free