Measuring Usability With Only 7 Questions

Dear readers & friends,

Smashing Magazine has published a new article of mine, which I co-authored with the fantastic Johanna Jagow.

The article introduces Iɴᴜɪᴛ, a questionnaire (or instrument) for measuring usability. I developed Iɴᴜɪᴛ already during my Ph.D. studies and presented it at various academic conferences, but only recently did Johanna and I have the chance to test it in a more longitudinal, actual industry setting. With this article, we want to make the questionnaire more available and accessible to practitioners in the field.

However, this is not just about advertising a new tool. The article also explains in detail what needs to be done so that one can be sure a questionnaire actually measures the thing it’s supposed to measure, and reliably so. If you’ve never heard of “confirmatory factor analysis” or “Cronbach’s α,” we provide you with a (hopefully) understandable introduction to those as well.

The bottom line is, you can’t just come up with a set of questions and start asking those. It’s important to rely on scientifically founded questionnaires if you wanna be sure to measure the right thing in the right way. We hope our article also does a good job explaining this.

And, as usual, if my own article is not enough reading material for you, here are my reading and tool recommendations. Today’s topics are ChatGPT (can’t get rid of it) and accessibility.

Reading Recommendations

📖 What Is ChatGPT Doing … and Why Does It Work? by Stephen Wolfram provides a detailed explanation of how a Generative Pre-trained Transformer (GPT) works internally. The article explains that ChatGPT, like other large language models (LLMs), generates text by producing a “reasonable continuation” of the given input, where “reasonable” is defined as what one might expect someone to write after seeing what people have written on billions of webpages.

📖 AI Is First New UI Paradigm in 60 Years by Jakob Nielsen discusses how AI is introducing the third user-interface paradigm in computing history, shifting from command-based interaction to intent-based outcome specification. This means users tell the computer what they want, not how to do it, reversing the locus of control.

Tool Recommendations

🛠 Randoma11y is a tool that lets you generate accessible color combinations for web design. It aims for a 4.5:1 ratio between the foreground color (e.g., text, links, etc.) and the background color to ensure people with moderately low vision can differentiate colors and see your content.

Ah, and one more thing: Markus Pirker interviewed me for his UX Heroes podcast, where I talk about the balance between data-drivenness and innovation, and how “scientific” UX research needs to be. However, the episode is in 🇩🇪 German only.

Enjoy reading/listening and until next time. As always, I welcome your feedback and comments on my articles and this newsletter. Simply reply to this email to share your thoughts.

Cheers,
Max