User research from a psychological perspective
This website was created mainly with scientific research in mind, but I'm putting some thoughts on the transition to applied research here. These are mainly based on my experiences working in user research (UR) / user experience (UX) research. The following sections describe some differences and similarities.What does UR look like?
Research in UR is typically done for a paying client, within a project with a specific scope and a set duration (it was mostly six weeks for me, but can vary a lot). Projects are likely split up into "sprints" of one or two weeks with a particular sub-goal (people call that "Agile" working, even though, en passant, it seems to be the polar opposite of the original concept). You're likely a member of a team, with a project leader organising the work. There'll be multiple moments where you, with varying levels of formality, report on your progress or results, to the team or to the client. As a rough rule, you're likely working on one project at a time (or providing a specific service at specific times within multiple projects), and when it's done it's done. The project will likely end with a client presentation and a slide deck, designed to be very easily readable. While of course you'll want personal and career development, it's a job you do for an organisation, and the research is research you do to serve a project and, hopefully, the client.That's of course different from the day-to-day of academic research, especially in terms of the focus on project structure. As a postdoc, you're still doing work for someone, but it's much more flexible and you might have more ownership, or at least the sense of it. It's not a job in the sense UR is - you'll be working (or should be) towards having an academic career, and the postdoc is supposed to be a place to let you do that (while obviously, and with less uncertainty, benefitting the person you're working for). You're very much aiming for personal, future-focused outcomes - authorship on publications, grants, an academic network, possibly teaching experience, and skills. You'll by definition be on a fixed contract and might well not be staying at the same organisation afterwards. As faculty, things are completely different again, and research is one part of a complex job. I've seen it be primarily delegated by principal investigators who have played the funding game successfully, and focus on securing more and more of that funding. For me, as someone unblessed by grant agencies but working at a department that was quite generous in terms of research time, I kept doing research hands-on, together with collaborators and students.
So, I'd say the main difference in the step to UR in terms of the structure of the job is working within short-term projects - put very simply, projects in which there's a client, a deadline, and a team, and in which the specific content of the research isn't relevant to you personally.
Research aims
Following on from the point above, the "why" of research is also very different. In UR, the research aims are determined by the client. This is a critical element of good UR - finding out what the client wants/needs to know. This is likely explicitly included as an early aim of the project, involving interviews and group sessions with stakeholders. This is also critical to get buy-in and support for doing the research, and avoiding unhappy clients who don't believe they're getting value from research at the end. Coming from an academic background, I would mix this process of alignment and identifying research aims with reading relevant academic literature and government or industry reports. This can give a priori ideas of what might be useful to clients, although there still needs to be agreement and shared understanding.Research questions and aims in UR broadly concern understanding how people experience some service or product, but - specifically "understanding" in such a way as to be useful to the client, ideally to help them make that service or product better, which could mean: more profitable, requiring less calls to helpdesks, more accessible to more people, etc. So, the aim of research isn't for the researcher to understand things better! That might be practically necessary, but it's ultimately all about the actual downstream effects of what you deliver to the client. A little more specifically, UR aims could involve how people use a service or product, in granular steps; where things go wrong in that process; what the different types of users are; what users really need from a service, and whether they're getting that; what attitudes could affect usage; what skills or knowledge employees need to be effective; how organisational changes could influence morale; and so on.
Additionally, there are the "hidden aims" of user research, which aren't about research. This can be about gaining buy-in from stakeholders, increasing confidence of clients, signalling values, and so on.
In comparison, academic research is, idealistically, driven by the educated curiosity of a researcher deeply immersed in a particular field of scientific study. This research happens on the cutting edge of human knowledge, in order to expand it and progress science. Less idealistically, research is done for a range of less high-minded reasons - to get papers in high-prestige journals, to support getting the next grant, to build one's profile by claiming some insight or method, to fight bitter battles of egos, or to satisfy university management optimizing for competitive incentives (or at least claiming to).
Finally, the position of falsification is vastly different. Popper was clear that falsification is normative, not descriptive, and in that sense there's maybe less difference than there should be, but in a consultancy context the idea of opening yourself up to falsification will be extremely alien to a lot of people. In academia, people at least pay lip service to the concept, and it does inform how you do research - you have to have that vulnerability to discomfirmation in science; you have to value reality telling you "no". In industry, people's aim may well be to avoid the risk of falsification, obviously at the cost of research being worthless. Good URs have a struggle here - you need to be able to be the bearer of unwelcome news, and stakeholder-manage your way through. The difference is that "unwelcome news" is going to be defined relative to what a client wants to hear rather than a predicted result not occurring - UR can be so exploratory in nature that it's hard to be wrong in that sense.
Methods
When I read books on UR, there was a massive overlap with methods and concepts from psychology; that makes sense of course, given the topic. But one of the "moats" UR has built, protecting itself as a form of expertise, is the specific naming and forms of methods. There's a new language to learn, with terms such as user journeys, user needs, usability, personas, or pain points. There's a kind of ironic stumbling block here coming from an academic background, in that there's far less to the methods than you expect, and that actually trips you up. The concepts aren't much more than just the basic ideas, as spoken, with some prevailing formats people expect you to use; you primarily just need an example to follow. Sticky notes come into things a lot, either physical ones stuck on walls or via special software primarily designed to let people use virtual sticky notes online. That's not to say there's no expertise to be built around UR methods, but it's not like the body of knowledge of scientific research methods. Given that you already have training in psychological methods, what's new is more about the context of using the methods - how do different stakeholders like to be engaged? What provides understandable outputs? What evokes confidence? How can you give research to developers in a way that's concretely usable to them? What can be done within typical project limitations?In terms of data collection, typical UR methods are (semi-structured) interviews, focus groups, surveys, and various kinds of workshops; possibly combined with analytics data or secondary data. Sample sizes tend to be extremely small, apparently based on generalising from the recommendations of authoritative figures (which I suspect were originally only intended for very basic usability tests). Analysis tends to be qualitative - I've seen both "capital-Q" holistic interpretivism and "standard" thematic analysis with a more data-driven mindset that aims to constrain interpretation depending on the analysis step. I've tried AI-based analyses and they seem quite feasible at generating reasonable output, at the cost of not giving the same level of (possibly actionable) detailed insight to the researcher. Surveys tend to be very simple, with open text questions playing a relatively large role; validated scales are used sometimes but certainly not necessarily. People will often use Likert scale questions they generated themselves, that make sense to them. My impression is that statistical analyses are typically very minimal or absent, beyond very basic descriptors; many URs don't have that background. There seems a widespread belief that "why" questions can only be answered by qualitative methods.
Workshops are perhaps the biggest methodological difference with psychological research. These are very activity-based, having the group of participants be literally (if possibly also virtually) hands-on. Sticky notes traditionally play an absolutely critical role in this. The workshop involves possibly quite creative game-like activities - placing sticky notes of various colour near visual stimuli (perhaps representing axes of binary opinions), writing ideas or responses on sticky notes, moving sticky notes around so you get groups of similar answers, placing sticky notes on other sticky notes in a sticky-notesception... The aim in terms of research narrowly conceived, very roughly speaking, tends to involve either getting a broad overview of feelings on a particular topic or extracting subject matter expertise or contextual information from the participants. The latter type of aim aligns with the notorious "lend me your watch and I'll tell you the time" style of consulting. In principal, of course, more versus less sophisticated uses could be made of workshops, as with other methods. However, workshops are perhaps really most aligned with the "hidden aims" of user research noted above, around stakeholder management.
The methods in academic psychological research are clearly more complex and varied, and subjected to extensive critical research as topics in themselves. The amount of methodological education and training in even bog-standard undergraduate courses is considerable - you have multiple years of study with at least one module per semester on something like interpersonal skills, formal interview skills, basic data analysis, statistical techniques like ANOVA and regression, qualitative methods, and applications to research projects. And then you might have additional clinically-focused classes, such as introductions to psychotherapy including role-play exercises or training neuropsychological assessment. Coming from an academic background, you can lose sight of the fact that those basics are not generally a normal part of people's educational background. This can present difficulties in terms of getting buy-in when planning research for a project, possibly more from your own team than from clients.
Experimental designs and implicit measures are perhaps the most obvious types of method missing from typical UR but potentially relevant. These would, however, require a very different skill set or resourcing to use; psychophysiology even more so of course. There could yet be a cross-over for such measures (specialist agencies focused on them do exist) but it would be a very different kind of UR. However, I've found that richer surveys, using validated scales and basic statistical methods like regression analysis and dimension reduction/clustering methods, are sometimes feasible to introduce and highly appreciated by clients (especially after the results have been presented, which makes the value of the possibly unfamiliar approach concrete).
Outputs
The output of user research on a project will tend to be part of a slide deck. This will usually be corporate style - think bullet points, varied visualisations, and random pictures. There'll also be a presentation, which may use the same slide deck. This will tend to be highly simplified and should be focused on an immediate "so what?" in terms of the client's specific needs (although you will need your own details at hand to back yourself up if challenged). I have also written more detailed research reports, but less often and you need a sufficiently research-minded audience for there to be a point. Your employer may also want a case study in a given format, primarily to use as evidence of experience when bidding for work. There may also be some use of research by comms or marketing teams for their purposes.Evidently, academic outputs are vastly different - the peer-reviewed academic paper and poster or oral presentations to expert peers at conferences. The aim is to convey detailed, novel, generalisable findings and their relevance to the scientific field, i.e., more research.
A final difference in terms of outputs is that, with UR project work, the output is likely the end of the research. You'd have to be lucky to even repeat research on a similar topic. In science, iteration is the norm. The oputput becomes the input of further research - why did this happen? Why didn't X1 work? Will X2 work? Does this happen under different circumstances? Was my interpretation of the previous findings correct? I think that iteration is significant factor underlying differences - UR has to (claim to) get it right in one go, in the sense of having one chance to provide value to the client.