I started this blog in August 2023. After 2 years I’d like to highlight a few posts. I hope these will have continuing interest for anyone who missed them; and also may be good introductions for new subscribers (“subscribe” is always free here; click the envelope ✉️+ icon in the upper right!)
I sort these articles into categories of:
I’ve added a brief note about each topic as a preview or comment. I hope you find something new and provocative in these articles!
The End of Tech as a Big Family: https://quantuxblog.com/the-end-of-tech-as-a-big-family. This post reflects on the personal anxieties and changes in Big Tech that were exposed during recent layoffs.
We’re Far from AGI: https://quantuxblog.com/were-far-from-agi. As a psychologist and sometimes philosopher, I explain why proponents of AGI and “superintelligence” don’t understand what they’re claiming (or more precisely, they don’t understand what they are not discussing).
Quant Methods
Multidimensional Sentiment Analysis: Part 1, Part 2. If I were to sort methods into quadrants on a 2×2 grid, arranging them by powerfulness vs. frequency of usage, then multidimensional sentiment analysis would score as a “high opportunity” — very powerful but surprisingly rarely used. I’d like to change that!
Easy MaxDiff in R: https://quantuxblog.com/easy-maxdiff-in-r . If you’d like to understand MaxDiff from the ground up, this short code explainer may help (note: the code is for educational purposes, not production). It looks at the observations that MaxDiff collects, and basic statistics for that data.
Individual Scores in Choice Models, Part 3: Respondent Segments: https://quantuxblog.com/individual-scores-in-choice-models-part-3-respondent-segments. I examine one way to segment MaxDiff data. But even more, this article is about how segmentation is usually misunderstood and what it actually does.
Individual Scores in Choice Models, Part 4: Inspecting Model Fit with RLH: https://quantuxblog.com/individual-scores-in-choice-models-part-4-inspecting-model-fit-with-rlh. I look at the RLH fit statistic in conjoint analysis and MaxDiff. Using basic math, theory, and code, I show why common heuristics to interpret RLH are misleading. RLH is diagnostic but cannot “filter” respondents.
Method Critiques
Surveys and the “True Score” Mistake: https://quantuxblog.com/surveys-and-the-true-score-mistake. This article explains why surveys are not about finding “the truth”, aka a latent score, in people's heads. Instead surveys are about listening to people. (This article is also one of the ground for my criticism of synthetic data in the next article.)
Synthetic Survey Data? It’s Not Data: https://quantuxblog.com/synthetic-survey-data-its-not-data This article explains why I believe the concept of “synthetic data” is impossibly flawed in multiple ways, including basic logic, statistics, and in regards to the scientific method.
Critical Assessment of the Kano Model: Part 1, Part 2. If you are tempted to use the Kano Model to identify “delighter” products … I don’t recommend it. This pair of articles explains why not, and gives alternatives.
Career Advice
Skills Combination for Quant UX Applications: https://quantuxblog.com/skills-combination-for-quant-ux-applications. I’ve interviewed hundreds of candidates and, for several years, coordinated the standard Quant UX hiring criteria at Google. This article summarizes the skills needed to be a Quant UXR, and is also a good preview of the Quant UX book!
Recommendations for Quant UX Interview "Portfolio" Presentations: https://quantuxblog.com/quant-ux-interview-portfolio-presentations-recommendation. If you are asked to give a Quant UX research presentation — sometimes confusingly called a “portfolio” presentation — this is how to approach it. The short version: give a research presentation that interests you; don’t try to guess some “right” content.
Distinguished Guests!
How to make HEART metrics work in practice: https://quantuxblog.com/how-to-make-heart-metrics-work-in-practice One of the founders of the Quant UX discipline, Kerry Rodden examines the evergreen and useful HEART framework, detailing practical advice to apply it. HEART is my all-time single favorite Quant UX framework, and a memorable heuristic to draw on in many, many research contexts.
Be a T-shaped Quant UXR: How Doing Qualitative Research Made Me a Better Quantitative UX Researcher: https://quantuxblog.com/be-a-t-shaped-quant-uxr-how-doing-qualitative-research-made-me-a-better-quantitative-ux-researcher Kitty Xu, cofounder of the Quant UX Conference, explains why Quant UXRs should also engage personally with qualitative research.
More Reading?
Each article above has pointers to additional articles or references. Or check out:
Happy reading!
