In this post, I outline 4 crucial areas for Quant UX applications (I'll call them "3.5 skills" for reasons to be noted). Then I address a reader's question about how to show the combination of skills during an application process.
The general areas are covered in depth in dedicated chapters of the Quant UX book, and I won't repeat all of that. Instead, I add reflections on the interaction among them in an application.
The 3.5 Skills for Quant UX
There are 4 skills areas to consider in Quant UX, in the following priority order:
Research Design
Statistics
UX experience
Coding (... we'll call this 0.5 skills for reasons below)
Before I dive into the skills, here's a graphic that (I hope) speaks for itself. (Bonus: see the Appendix below for the R code that made this diagram.)
As you can see in the diagram, I put Quant UX at the intersection of four skill sets. Here is a brief discussion of each skill set.
Research design means skill in planning, running, analyzing, presenting, and following up on research to answer a business need. Basically, it means that a UXR should be trusted to come up with a research plan and ensure it is useful for the product team. The breadth and depth of projects vary according to one's career stage. (See Chapters 4 and 14 in the Quant UX book.)
Research design is #1 because Quant UX is a research position, and strong skills in planning and conducting research are essential, without exception.
Statistics is the technical aspect of acquiring and analyzing data. For Quant UX, the approximate level of required statistical knowledge is applied and fluent understanding of the uses and misuses of linear regression models; plus some other area of personal depth (such as data visualization, or special models). By "applied and fluent," I mean an understanding of how to do regression properly and how it goes wrong in practice. Again, this would be relative to a career stage. (See Chapter 5 in the Quant UX book.)
Statistics is #2 for Quant UX because it is where research design meets data to deliver user understanding.
UX experience means that you can map the preceding skills — research design and statistics — to the UX world. That means understanding UX stakeholder questions, how to work with human subjects data, understanding the issues of reliability and validity for human data, and delivering results to UX audiences such as designers.
UX experience is #3 because, at the beginning of a career, it may be an aspirational and largely undeveloped skill. Some positions do not require prior work experience in UX. Still, you always will be expected to discuss research in a way that would fit well in a UX organization. That means focusing on a user-centered approach to the research problems, data, and deliverables. (See Chapters 2, 3, and 11 in the Quant UX book.) Put differently, as long as your thinking is user-centered, you can learn the details of UX engagement on the job.
Coding is the ability to use a general-purpose programming language, such as R or Python, to accomplish various tasks of acquiring, formatting, and otherwise dealing with data, and to do statistical modeling and data visualization. I've said more about programming for Quant UX in a separate post, and will simply note that although coding is not required in every position, it is highly advantageous in all positions. (See Chapter 6 in the Quant UX book.)
Coding scores #4 here, and counts for 0.5 skills, simply because it is not always required for Quant UX and Quant UXR positions, even if it is advantageous.
The Skills Intersection ... and Interviewing
A reader emailed and asked me (I paraphrase):
I have the skills for Quant UX but across different projects and contexts. How do I present them together when interviewing?
The short answer is this: you don't need to show all of the skills combined on a project. As long as you have each skill independently, that's enough. Or, to say it differently, you don't have to be a Quant UXR to land a Quant UXR job.
To expand on that, a simplified view of the typical Quant UXR interview process is that you will need to do two things:
Present research that shows you are a good researcher, are passionate about your work, and can explain it and answer questions from an intelligent audience.
Do well in individual interviews that probe into your skills in statistics, research design, communication abilities, UX engagement, and (maybe) coding.
Neither one of those requires a demonstration of research that combines all of the skills for Quant UX. For example:
Academic candidates may lack the UX experience aspect ... but could show solid aptitude for UX it during interviews
Data analysts may lack a project with coding ... and yet may have that experience from other sources, such as hobbies or coursework
Qualitative UXRs may not have a project that demonstrates statistical modeling ... and yet may pass the required knowledge from academic work.
In fact, the lack of combined experience may serve as a great explanation of why you are interested in Quant UX!
Consider this answer to why you're interested in Quant UX:
I have these skills but have not been able to combine them. That's why I'm interested! And here's how I would do that in Quant UX ..."
That is a very convincing case — as long as you can back it up with at least basic fluency in each of the 3.5 skills areas.
We have more to say in the book. And stay tuned here for more posts to come!
Appendix: R code for the Venn Diagram
Two-circle and three-circle Venn diagrams are easy to make. But how about 4-way Venn diagrams, like the one in the post above?
R can help. Here's the code!
But first, I'll say this: a huge part of programming consists of finding examples — books, online, from colleagues, from yourself — and then adapting them. This example uses code adapted from @HazelJay on StackOverflow (thanks!)
# based on https://stackoverflow.com/questions/71006727/venn-diagram-4-sets-with-r-problem-with-percentages
library(ggvenn) # install if needed
dat.a <- sample(x = c(TRUE, TRUE, FALSE), size = 1000, replace = TRUE)
dat.b <- sample(x = c(TRUE,FALSE, FALSE), size = 1000, replace = TRUE)
dat.c <- sample(x = c(TRUE, FALSE, FALSE, FALSE), size = 1000, replace = TRUE)
dat.d <- sample(x = c(TRUE, TRUE, TRUE, FALSE), size = 1000, replace = TRUE)
df <- tibble(values = c(1:1000),
Coding=dat.a, Statistics=dat.b,
`Research Design`=dat.c, UX=dat.d)
ggvenn(df,
fill_color = c("lightblue", "grey60", "grey90", "lightpink"),
show_percentage = FALSE,
stroke_color = "white",
digits = 1,
text_size = 0)
This R code draws a 4-ellipse Venn diagram as shown next. (For the diagram presented at the start of this post, I manually added annotations for the intersections.)
Are you convinced to learn R? LOL, I hope so! Thanks for reading :)