# Survey Question Bias Detector MCP for AI Agents MCP > Survey Question Bias Detector helps you clean up your research. It uses NLP to find leading language, double-barreled questions, and unbalanced response scales in surveys and NPS questions. Stop getting skewed data by catching methodological errors before you launch your study. This MCP ensures your questions are clear and your results are actually reliable, saving you hours of manual auditing. ## Overview - **Category:** nlp - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_v0qgJaxUfO7Htt4jKvIXE7AO1ifqZf1XEP8pymDX/mcp - **Tags:** surveys, nps, bias-detection, linguistics, data-quality ## Description When you're designing a survey, it's easy to accidentally nudge people toward an answer you want to hear. You might use a word that feels slightly biased or combine two different questions into one, which makes the data messy and hard to interpret. This MCP fixes that by acting as a rigorous editor for your research questions. Instead of manually scrubbing every line of text, you can have your agent run a check on your draft to find hidden issues. It looks for things like double negatives that confuse respondents or scales that don't offer a neutral middle ground. By using this tool through the Vinkius catalog, you ensure your data is actually clean and your results are reliable. It takes the guesswork out of survey design, making sure your questions are clear, balanced, and scientifically sound. You'll spend less time worrying about skewed results and more time acting on high-quality insights. It handles the heavy lifting of linguistic auditing so you can focus on the strategy behind your research. It provides a consistent second set of eyes to keep your methodology sharp and your findings trustworthy. ## Tools ### identify_negation_confusion Finds double negatives that might confuse your respondents. This helps ensure your data remains accurate and easy to interpret. ### scan_leading_language Detects biased or loaded words that nudge users toward a specific answer. It helps you maintain a neutral and objective research stance. ### verify_scale_symmetry Checks if your response scales are balanced and offer a fair range of options. This prevents you from accidentally skewing the results. ### check_conjunction_complexity Identifies double-barreled questions that ask two things at once. It helps you split complex prompts into clear, single-topic questions. ### evaluate_cognitive_load Measures the complexity and length of your questions to prevent respondent fatigue. It helps you keep your survey engaging and easy to finish. ## Prompt Examples **Prompt:** ``` Is 'Don't you agree that our service is excellent?' biased? ``` **Response:** ``` Yes, this question is biased. It uses loaded language ('excellent') and a leading 'Don't you agree' structure. This creates a high risk of skewed results. ``` **Prompt:** ``` Check if 'How satisfied are you with the food and service?' is double-barreled. ``` **Response:** ``` Yes, this is a double-barreled question. It asks about two distinct topics (food and service) in one prompt, which will make it impossible to tell which one the user is rating. ``` **Prompt:** ``` Evaluate the cognitive load of: 'In the event that you find yourself in a situation where the product fails to meet your expectations, how would you rate us?' ``` **Response:** ``` The cognitive load is high. The sentence structure is overly complex and the word count is high for a simple rating question. It's likely to frustrate respondents. ``` ## Capabilities ### Spot leading language Identifies biased or loaded words that nudge respondents toward a specific answer. ### Identify double-barreled questions Detects questions that combine two different topics into one prompt. ### Check response scale balance Ensures your scales offer a fair and symmetrical range of options. ### Assess cognitive load Measures the complexity and length of questions to prevent respondent fatigue. ### Detect double negatives Finds confusing phrasing that might lead to incorrect data. ## Use Cases ### Auditing large-scale usability surveys A UX researcher has a 50-question survey and wants to make sure none of them are double-barreled. They ask the agent to run `check_conjunction_complexity` on the whole list to find questions asking about two different features at once. ### Validating NPS questions for bias A product manager is worried their NPS questions are too pushy. They use `scan_leading_language` to see if words like 'amazing' or 'must' are skewing the results. ### Reducing survey fatigue in long forms A market researcher needs to know if a survey is too hard to read. They use `evaluate_cognitive_load` to identify which questions have too much text or complex sentence structures. ### Verifying response scale integrity A data scientist finds that a survey's results are leaning heavily one way. They use `verify_scale_symmetry` to see if the response options are actually balanced. ## Benefits - Get cleaner data by catching double-barreled questions with `check_conjunction_complexity` before you send your survey out. - Improve respondent completion rates by using `evaluate_cognitive_load` to simplify overly complex phrasing. - Eliminate researcher bias by using `scan_leading_language` to find and remove loaded words from your drafts. - Ensure balanced results by using `verify_scale_symmetry` to check that your Likert scales are fair. - Reduce respondent confusion by using `identify_negation_confusion` to catch tricky double negatives that skew results. ## How It Works The bottom line is you get high-quality, unbiased research data without the manual headache of linguistic auditing. 1. Paste your draft survey questions into your AI client. 2. Ask the agent to run a bias and complexity check on the text. 3. Get a detailed report on linguistic flaws and suggested rewrites. ## Frequently Asked Questions **What can the Survey Question Bias Detector do for my market research?** It finds linguistic flaws like leading language and double-barreled questions. It helps you ensure your survey data is clean and unbiased before you launch it. **How does the Survey Question Bias Detector help with NPS scores?** It scans your NPS questions for leading words that might nudge customers toward a positive rating. This ensures your feedback is honest and actionable. **Can the Survey Question Bias Detector find double negatives?** Yes, it specifically identifies phrases like 'not unhappy' that might confuse people. This makes your survey easier to understand and improves completion rates. **Will the Survey Question Bias Detector help reduce respondent fatigue?** Yes, it assesses the cognitive load of your questions. It points out long or complex sentences so you can simplify them and keep people engaged. **How do I use the Survey Question Bias Detector to check my survey scales?** It checks if your response options, like Likert scales, are balanced. This prevents you from accidentally leading people toward one end of the scale. **Is the Survey Question Bias Detector good for academic research?** It's great for ensuring methodological rigor. It helps you catch structural flaws and linguistic biases that could invalidate your research findings. **How does it detect leading language?** The `scan_leading_language` tool compares your text against a lexicon of loaded words to identify potential bias. **Can I check for double-barreled questions?** Yes, use the `check_conjunction_complexity` tool to find conjunctions like 'and' or 'or' that merge multiple concepts. **Does it support NPS scales?** Yes, use `verify_scale_symmetry` to check if your response labels are balanced and include a neutral midpoint.