Glossary
Survey glossary
The terms you'll meet when designing and analyzing surveys, in plain language.
- Survey
- A structured set of questions used to gather information from a group of people about their opinions, behaviors, or characteristics.
- Sample
- A subset of people selected from a larger population to take part in a survey, chosen to represent the whole group.
- Sample size
- The number of people who actually complete a survey. A larger sample size generally makes results more reliable, though with diminishing returns. Calculate the sample size
- Margin of error
- A range, usually written as plus or minus a percentage (for example, ±5%), that shows how much the sample results may differ from what you would get by asking the entire population. Calculate the margin of error
- Confidence level
- The probability, commonly set at 95%, that the true population value falls within the survey's margin of error if the survey were repeated many times.
- Response rate
- The percentage of people invited to a survey who actually complete it. A low response rate can reduce how reliable the results are.
- Net Promoter Score (NPS)
- A loyalty metric based on a single question: how likely someone is to recommend on a scale of 0 to 10. It is the percentage of promoters (9 and 10) minus the percentage of detractors (0 to 6), and ranges from -100 to 100. Calculate your NPS
- CSAT (customer satisfaction)
- A measure of how satisfied customers are with a product, service, or interaction. It is usually asked with a rating scale and reported as the percentage of satisfied responses.
- CES (customer effort score)
- A measure of how much effort a customer had to make to complete a task, such as resolving an issue or making a purchase. It is usually asked with a single question on an easy to difficult scale.
- Likert scale
- A rating scale, typically with five or seven points ranging from "strongly disagree" to "strongly agree", used to measure attitudes or opinions.
- Open-ended question
- A question that lets people answer in their own words instead of choosing from preset options. It provides rich detail but is harder to analyze.
- Non-response bias
- A distortion that occurs when the people who do not respond differ systematically from those who do, so the results no longer represent the whole group.