MeasuringU

We’ve developed methods for the comparison of Net Promoter Scores with benchmarks. measuringu.com/nps-benchmark…

We ran two studies (with 1749 participants across 18 products) to determine the best method for constructing NPS confidence intervals. We recommend the adjusted-Wald method. measuringu.com/nps-three-con…

We often think of Think Aloud as a single method, but there are many ways to think aloud. measuringu.com/many-ways-of-…

Despite its known potential for overlaps and gaps in the classification of UI problems, interrater agreement (reliability) for heuristic evaluation was comparable with those reported for the CUP and OSUDC methods. measuringu.com/frameworks-fo…

Mobile apps are big business, so it’s important to have a method to measure the #UX of mobile apps that is quick to administer and easy to interpret. measuringu.com/how-to-score-…

User Experience Salaries & Calculator, 2026 edition. measuringu.com/salary-survey…

One prominent scholar has said the NPS is “widely discredited in the academic literature.” We reviewed more than 20 academic papers and found qualification, not discrediting, of the NPS. measuringu.com/nps-discredit…

NPS designations of detractor and promoter are based on referral and repurchase behaviors and the sharing of negative comments, but some critics are not convinced. measuringu.com/detractors/

Randomizing tasks minimizes unwanted effects that can distort your data, like participants learning or simply being fatigued. measuringu.com/scientific-th…

Small changes to survey items can have big impacts on responses. Does changing the wording from “will recommend” to “would recommend” have an impact? measuringu.com/will-would

Responses to the NSAT are based on a four-point scale, from “very satisfied” to “very dissatisfied.” Four points is not a lot for a single item measure, but it at least captures intensity, unlike a three-point scale. measuringu.com/microsoft-nsa…

Our investigation of how well digital twins predict human SUS scores for business software found average AI errors of 10 points with measurable differences in AI and human distributions. measuringu.com/digital-twins…

We analyzed data from seven large-sample #usability studies and found that the first five users uncovered 91% of the most common problems but only 22% of the least common ones. measuringu.com/sample-size-r…

Sometimes #UX researchers use five-point agreement scales that range from −2 to +2 with 0 in the middle, but are they any better than standard five-point numeric scales that range from 1 to 5? measuringu.com/standard-vers…

Independent review of 16 unmoderated #usability test videos by four evaluators uncovered 38 unique usability problems with 41% any-2 agreement among the evaluators. measuringu.com/examining-the…

Here’s how to use banner tables to present #survey results. measuringu.com/how-to-use-ba…

Abelson’s eight laws are not only witty, they’re also useful constructs for researchers who use statistical analysis to guide their research narratives. measuringu.com/abelsons-laws…

There is a lot of speculation (and only a little research) on what effect neutral point labels have on Net Promoter Scores. So of course we ran our own studies to find out. measuringu.com/labeling-nps-…

Not everyone likes the NPS, but many organizations rely on it. The SUPR-Q contains the 11-point likelihood to recommend item as part of its score. measuringu.com/10-things-sup…

Do eleven-point scales take longer to respond to than three-point scales? In our analysis we found they take 100 to 300 milliseconds longer to respond to. measuringu.com/3-point-nps/