Browse Content by Topic

Usability Testing (19)

Sample Size (18)

UX (12)

Task Time (11)

Survey (11)

Usability (11)

SUS (11)

Questionnaires (9)

Usability Problems (9)

Rating Scale (8)

Net Promoter Score (7)

Statistics (7)

Remote Usability Testing (7)

Usability Metrics (6)

Problem Discovery (6)

Satisfaction (6)

Confidence Intervals (5)

Task Completion (5)

Formative Testing (3)

Regression Analysis (3)

User Research (3)

Summative (3)

Loyalty (3)

Heuristic Evaluation (3)

SUPR-Q (3)

Margin of Error (2)

Credibility (2)

Card Sorting (2)

Qualitative (2)

Information Architecture (2)

Sampling (2)

Crowdsourcing (2)

Clicks (2)

PhD (2)

Benchmarking (2)

Mobile Usability Testing (2)

SEQ (2)

Keystroke Level Modeling (2)

Problem Severity (2)

Cognitive Walkthrough (2)

Task Times (1)

Task Failure (1)

Salaries (1)

Ordinal (1)

Completion Rate (1)

Z-score (1)

UI Disasters (1)

Interval (1)

Marketing (1)

Top Box Scoring (1)

Monte Carlo (1)

Return on Investment (1)

Ethonographic Research (1)

Geometric Mean (1)

Key Drivers (1)

Median (1)

Methods (1)

Evidence Based Design (1)

A/B Testing (1)

iPhone (1)

Analytics (1)

Lean UX (1)

Correlation (1)

Errors (1)

Reliability (1)

Think Aloud (1)

Post-Task Ratings (1)

Quality Assurance (1)

Survey Topics

How to estimate a survey response rate

Jeff Sauro • June 15, 2011

Using the lower-boundary of a confidence interval on a pre-test response rate will provide an accurate estimate of your full-survey's response rate. You can use this to estimate how many total invites you need to maintain a margin of error in your responses.[Read More]

Should you care if your rating scale data is interval or ordinal?

Jeff Sauro • June 1, 2011

It's fine to compute means and statistically analyze ordinal data from rating scales--the numbers don't know where they came from. But just because one rating is twice as high as another does not mean users are really twice as satisfied.[Read More]

How to interpret survey responses: 5 techniques

Jeff Sauro • May 10, 2011

Closed ended rating scale data is easy to summarize and hard to interpret. These five approaches provide meaning to raw responses and often generate similar results.[Read More]

8 Research Based Insights for User Experience Surveys

Jeff Sauro • May 4, 2011

It's easy to get derailed when writing a survey or questionnaire when measuring the user experience. You need to worry about what to ask, who to ask and how what you're asking affects the responses. These eight insights will help make things a bit smoother.[Read More]

Survey items should include a neutral response: Agree, Disagree, Undecided?

Jeff Sauro • April 19, 2011

Research generally shows that including a neutral response will affect the distribution of responses and sometimes lead to different conclusions. However, this is less important when assessing usability as you're usually more concerned about comparisons over time or against a benchmark than the percent of users who agree to statements[Read More]

How to quantify comments

Jeff Sauro • February 21, 2011

Quantifying the frequency of comments with a binomial confidence interval helps you estimate a sentiment in the total user population and prioritize findings.[Read More]

Top-Box Scoring of Rating Scale Data

Jeff Sauro • December 14, 2010

Top-box and top-two-box scoring is appealing for summarizing responses in the absence of benchmarks or comparisons. Top-box scoring has the disadvantage of losing information about precision and variability. Reducing 5, 7 or 11 response options to two or three options can mask real changes in attitudes.[Read More]

How many people cheat in online surveys?

Jeff Sauro • December 7, 2010

Around 10% of paid respondents to online surveys are rushing through surveys to get the honorarium. With longer surveys and more complicated questions, closer to 20% of responses will be unusable.[Read More]

That's the worst website ever!: Effects of extreme survey items

Jeff Sauro • September 21, 2010

Items in the System Usability Scale (SUS) were rephrased to either all extreme positive or all extreme negative wording and average scores were compared. Users disagree more with items that are worded in the extreme and resulted in significantly different SUS Scores.[Read More]

Survey Respondents Prefer the Left Side of a Rating Scale

Jeff Sauro • September 14, 2010

There is a modest bias towards the left side of a rating scale. This matters most for stand-alone surveys or questionnaires when no comparisons are being made. In such cases, the number of "agree" or top-box statements will be higher if placed on the left-side of a scale.[Read More]

Should you use 5 or 7 point scales?

Jeff Sauro • August 25, 2010

7 point scales tend to perform slightly better than 5 point scales. The benefit is too small to change your existing questionnaires if you have historical data. Having more points will provide the biggest benefit when you have only a few or one question in your questionnaire. Focus more on what you'll do with the results than whether 5 or 7 points is better.[Read More]