Last post, I proposed that UX is just specialized Applied Psychology.  If that’s so, what can modern UX learn from Applied Psychology?

Far Side by Gary Larson

A clinical psychologist at work

Stepping back, what is Applied Psychology? A smart-ass answer would be: it is the application of psychology… then if the questioner doesn’t walk away, I would expand on this: Psychology is the study of human behavior; so much of psychology is basic research to build a body of knowledge. Applied psychology uses that knowledge to solve practical problemsClinical psychology, for example, which most people think of when they think “psychology”, is a practice that applies methods and findings shown to be effective in research to help individual patients. I stopped telling people I was a psychologist because they’d respond “Are you gonna mess with my head??”; now when I say I’m an experience designer they just look at me funny and move on.

Each sub-field of psychology has its own research/practice element — looking at the site for the International Association of Applied Psychology (founded in 1920, the first international society within psychology) and Wikipedia, it seems that AP is just an umbrella for a laundry list of specialties. Specialization is the natural tendency of academia; Hugo Münsterberg, the father of AP,  was probably the last generalist — he worked in forensic, clinical, and industrial psych.

So the problems AP looks to solve depend on the specialty–for example, I was writing promotion tests, working in Industrial/Organizational Psych. I also wrote job surveys, used for employee classification, selection, training, etc. The process and methods were based on research going back to military psychology in WW I.  So the problems solved are as broad as human behavior; including the ‘wicked problems’ that designers want to address using ‘design thinking’. The IAAP, for example, works with the United Nations on topics like human trafficking, reducing poverty, empowering youth (to address lots of issues)… thinking about UX then: how do you wireframe that solution?? (Well, first you argue about which is the best wireframing tool….).

How does Applied Psychology approach solving a problem?

Like the sub branches of psychology, UX has distilled generations of research and practice into established methods and principles of designing solutions. For example, if you know you’re making a banking app for the iPhone, you have platform standards, UI patterns, competitor products, and user research and evaluation methods to apply to solve the problem. You don’t have to reinvent the wheel, process-wise, every new project.  Same thing with I/O Psychology, for example; if you’re contracted to help a company fix a toxic workplace, you have all the background in organizational behavior to apply.

If you want to step back to solve a more general problem, you go back to the general process.  Applied Experimental Psychology is what I studied at Virginia Tech (I’d link to the program but they dropped it…), which is essentially the scientific method.

Let’s take an example — the banking app, and forgive my sloppiness here since I’m writing this in an hour and not the year plus it took for my thesis..

First step is the research question and understanding the problem.

In basic research, many of the topics are follow-on to previous research, or questions that arise out of following the topic. In product development, UX is typically handed the requirements from marketing; though some in UX would argue you need the user research to suss out the user needs, hence requirements (hints to applied psych). For my example, I could be interested in a news article that Americans aren’t saving enough, inspired by my own observations. I’d then do a literature review to understand more about the issue, and validate my personal observations. If I find enough data to come up with a testable hypothesis I can move on, otherwise I might need to do more basic research (surveys, interviews).

Second step is to formulate the hypothesis and operationally define the variables.

Lets say I decide to focus on college students (who are the subject of most psychological research anyway), with the idea that they don’t think they have enough money to start saving, and are waiting for the first job. My literature review showed that’s not true, it’s more a process of forming habit, so I’m going to try that route. My informal idea is that if students are prompted to save then they will save more (than those who just have a general goal but no prompt). And to make it a little complex (not generally a good idea), I have two concepts for prompting, one at the beginning of each month, and one at each spending opportunity. Now I have to operationalize my variables — what is a prompt, what is ‘more’ etc (e.g. I decide to measure percent saved vs. total $ since it is independent of $ earned).

Third step is to create and conduct the intervention.

Here is where AP probably diverges from UX–the domain most UXers are working is known: digital, computer software, web pages/sites/applications, mobile applications, etc. In AP, the potential solution is open, so there is a lot of interdisciplinary work both in understanding the question and building the intervention (or experimental prototype). A team of experts is brought together to work the project. My favorite ‘aha moment’ in not pre-supposing the solution was at an early human factors conference — the problem was certifying/training truck drivers. Since I was working computer-based training, the answer was obvious to me; but I was wrong because the solution that worked for them was ‘books on tape’…duh, they’re driving all day, have to watch the road…nobody’s going to sign up for a course off duty.

Anyway, going back to my example, maybe I get college students working for the college (standardize on those getting a paycheck), which still has some volunteer bias but maybe I can work with that randomizing treatments.  Once I get approval for my study from the Institutional Review Board months later (something web sites don’t have to worry about–but should)  I’m running my study.  I randomly assign students into three groups: G1 is asked to put aside whatever amount they want, as soon as they get the paycheck (pay yourself first idea).  G2 is asked to keep a diary of each payment transaction, then for each transaction add an additional amount for savings and put that aside. G3 is control group, they just sit through the initial briefing and hear how great saving is, and what you should typically save. After a month I bring all three groups back and measure the amount they saved, and compute the percentage against their monthly paycheck. Since I’m studying habits, I bring them back another month later to ask for totals, even though they’re not prompted.

So, maybe this is a pilot and I find all the biases and potential confounds, and rework my study — maybe find a programmer to make an app to streamline the prompting and saving. But say I collect and analyze the data to find the ‘per transaction’ prompt works significantly better. This would lead to the last step.

Fnially, Scale up and Follow up

In the last stage I need to think about scaling up the successful intervention, and continually following up to verify it is working and tweak, or propose follow on experiments. For instance, working with the per-transaction example, we could design an app that would ping the user for a ’round up’ amount when it gets notice of a transaction, or a lump ‘pay yourself first’ when it gets a direct deposit.  You could find a bank to partner with, to create a service that does this automatically, which is what IDEO did with Bank of America. Tweaking and improving would look at visualization and gamification (see how you’re meeting or exceeding goals).  You would keep an eye on new research in behavioral economics, cognitive or social psych to suggest improvements or alternatives.

 What have we learned?

There has been a renewed interest in psychology for UX design; so much of what UX can learn from Applied Psychology deals with findings from those fields: social psych, personality, decision making, cognitive, sensory/perception, etc.

Aside from keeping current with the external research;  an Applied Psych perspective highlights following the experimental process, understanding how to operationally define your goals and measures, and how potential confounding variables can mess with you.  Probably my biggest take away is taking the time to understand the problem fully; doing due diligence in your research before jumping in with a solution.

Knowing there is a theoretical basis for the plan-research-design-evaluate cycle I find useful as well. There is art to UX design, but you can apply the science you learned as a kid too.

Leave a Reply

Your email address will not be published. Required fields are marked *