Skip to content

Joni Salminen Posts

Factors that affect evaluation complexity

evaluation complexity increases when: *task is more subjective*you have more coders*you have more categories*you have more data to code chance-adjusted agreement (as opposed to percentage agreement) metrics typically aim to consider these factors in their calculation (though there’s no easy way to quantify “subjectivity” except after the calculation; a low score can be of indicative of the subjective nature of…

The Dance of Dynamics

Research involves some interesting “back and forth” dynamics. Two of them are: 1. “Zoom in, zoom out” dance 2. Expand, simplify The first one deals with how to write paragraphs. For example: “There is little research on the effects of persona modality on users’ attention” (<= zoom out sentence = talks about the big picture). “However, of those studies that…

“AI didn’t replace thinking—it demanded more of it”

A quote from a panel discussion I participated recently:“AI didn’t replace thinking—it demanded more of it.” It is indeed our solemn duty, now more than ever, to ensure that the future generations get educated in thinking, whether logical, critical, or hypothetical. I don’t believe in the “calculator hypothesis” (=you don’t need to count because we have calculators; same way you…

How to cope with failure?

Failing is painful in academia. Rejection hurts, mentally and almost physically. And you get plenty of it. You get it MUCH MORE than success. So, it’s very tough. We need to learn to cope with failing to survive. Here’s a simple formula: Recharge … regroup … retry 1. Recharge = rest enough, take some time off from the screen, eat,…

Personas and startups

Presentation slides of “personas and startups: some ideas” talk at ERLab Days 2025 in Heilbronn, Germany. Personas and StartupsDownload

On the place called extreme (or “let’s not break the head”)

This is a snippet of an email I periodically send to my PhD students. “Let’s not break the head” is a good related quote, from my PhD student, Ilkka Kaate. Dear team, (cc other supervisors) you’ve all seen this place. You know you’re there when weird things start to happen: the research comes into your sleep. You can’t focus on…

When to Defend Your Ideas (As a PhD Student)

When to defend. Now, there’s a thing with being a PhD student and a supervisor. There’s some information asymmetry: the supervisor has more experience and knowledge. So, should the student do everything the supervisor asks? No. Definitely not, because the supervisor is still a fallible human being. The goal is to somehow surpass the supervisor! (it has to be 🙂…

Concerns of Complexity as a Competitive Factor in CHI Conference

CHI research has become more complex (but not more insightful). The reason? Complexity is a competitive factor: it helps separate your paper from others. Lengthy papers with more substudies and details are more likely to do better than short but insightful work that might have more relevance for HCI theory and/or practice. Reviewers are instructed to pay attention to HCI…