What is experimental research in psychology like day-to-day?

What is it like to do a PhD in something like social or cognitive psychology and be researching those sorts of problem full-time?

For context, prior to continuing, please read Background on the author.

My day-to-day

I work from home 99% of the time and I don't work all the time.
Often, during summer months, the university is pretty quiet so I may not have any studies running. This might be a time when everything I work on is either in a preparatory phase or is on hold until the September semester begins.

Realpolitik: The Most Important, Least Urgent Task

I try to spend as much of my time as I can writing papers.

Note that I said "writing" papers, not "reading" papers.
A dirty little secret is that very few academics actually read papers.
Okay, that's not perfectly true, but it is close!

Professors read papers for two primary reasons:

1. they were asked to review the paper, and
2. they're trying to get into a new area or trying to learn a new method.

Otherwise, established academics tend to skim titles and read relevant abstracts, but it is much less common to actually read a whole paper.

Graduate students and undergrads are in position #2: getting into a new area or learning new methods.
I am pretty late in my graduate studies so I've already read quite a few papers. At this point, I may still read papers to fill in knowledge gaps, but I typically read papers because I'm reviewing (position #1) or I want to make a claim in a paper I'm writing so I want to check the literature. It isn't that I am only looking to confirm: that would be biased. It is that I am reasoning my way through the manuscript I'm working on, then I realize that I am starting to make a claim without a source. At that point, I need to look for sources that discuss the topic, whether they support what I thought or undermine it completely. Discovering papers that indicate that I may be wrong are just as valuable, if not more valuable, because then I can re-think the situation or add nuance to my view.

More on writing papers and writing grants below.

Data Collection, Research Assistants, and Mentorship

I usually have a study or two running.
By "running", I mean data is getting collected. People sign up to participate and go through the study, whether this takes place in the lab, in a neuroimaging environment, or in the comfort of their own home as they fill out an online survey.

When studies run, I have research assistants (RAs) run them.
By "run", I mean that RAs do some or all of the following: recruit participants, get consent, explain tasks to the participant, answer participant questions, provide compensation to participants, and engage in various administrative tasks as needed.

The last time I ran a participant myself was several years ago.
At this stage, the purpose of my personally running participants is to train new RAs by showing them how it works: I use the "watch one, do one, teach one" approach to training. I also design my experiments to be easy for new RAs to run: most of my experiments are highly automated. I was an RA for several years during my own undergrad and I remember what it was like. I have plenty of experience running participants through studies, which helps me avoid data-collection pain-points during the design phase.

I have built an efficient structure for my RAs.
Over the years, I promoted some of my RAs to a Senior RA position, which puts them in charge of a whole study and gives them they responsibility and experience of training additional RAs as needs be. With this structure in place under my supervision, I have had no need to commute to my on-campus office in years. Sometimes I need to go to campus to set up new experiments and make sure they're working on the lab computers, but this is pretty infrequent (about once every 6 months). Technology allows data to be accessible over the internet.

This structure might not be available to everyone, including non-psych grad students.
My understanding is that grad students in other sciences often do the hands-on work themselves (e.g. mixing chemicals, preparing dishes, etc.). In psychology, some grad students still run all their participants, but this takes up a lot of their time. To my mind, taking this approach in psychology is very inefficient since one can get undergrads to do all the data collection. As a grad student, I have higher-level tasks I want to work on. Plus, undergrads need opportunities to try being an RA: they want to see what research is really like and collecting data is part of that experience!

Sometimes, I may ask an RA to do a literature review of an area of their interest if we want to run a study in that area and I'm not familiar with it. For example, creativity was not my field so, when I designed a creativity study, I had an undergrad search the literature to make sure my idea was novel and that it made sense to run the study.

Day-To-Day Tasks

In my day-to-day, I design and program experiments, analyze data, and write papers and grants. In this section, I describe what all that actually means.

Design

"Design" includes coming up with an idea, imagining a broad conception of how I might test that idea, then developing the specific ways I could test it.
The idea might start as broad as, "I wonder if ..." or "Nobody seems to be asking whether ...". This phase could involve some time reading the literature or I could take a "first principles approach". This phase often culminates in a literature search for relevant measures, i.e. questionnaires or behavioural tasks.

Design also involves considering feasibility, such as cost, skills, tools, and timelines.
Everyone has limited time and grad students usually have even more limited budgets. There is no point designing a two million-dollar study when your budget is "free". There's also not point designing a study that requires instruments to which I don't have access or analytical methods that nobody on my team understands. Timelines also matter: I would love to run decades-long longitudinal research, but that is not feasible on a graduate student timeline. Feasibility considerations provide necessary reality-checks, despite their sometimes revealing that the reality is limiting to the point of abandoning the idea, at least for the time being.

Design also involves considering what could go wrong.
I don't mean ethically, though that is also a factor to consider during feasibility. I mean what if the results are unexpected? As a scientist, I am curious, not biased toward specific results. As such, I design studies that probe reality and provide results that offer insights, whatever happens to be true. I do not design studies trying to "show" one specific statistically significant result lest they be considered "a failed study". Ideally, studies should not have "fail" conditions beyond technical errors (everyone makes mistakes!). Studies should be informative, whatever the results.

Design also considers relevance to the literature and "what will this look like as a paper".
I get a minimum of one paper out of each study and preferably three papers if I've been clever about the design. I'm not talking about "salami publishing". I'm talking about running a complex study that tests multiple ideas or contains several angles relevant to different audiences. These studies turn into multiple interesting papers rather than one massive unreadable paper.

The last step in design is writing a preregistration on the Open Science Framework where I lay out hypotheses and statistical tests. At this point, the study has gone from being a broad idea to a very specific implementation of that idea.

Program

"Program" means that I implement the study's structure and tools.

Generally, that means either implementing surveys (in Qualtrics) or writing code to run a task (in Python).

I also include writing instructions for participants in this step. The entire flow of the study is implemented, from the first point-of-contact (whether that is an email or a posting) through the study itself and ending with the last point-of-contact via the debriefing letter. All of this detail is required to submit an application for ethics approval.

After ethics approval, I pass the study off to RAs and they do data collection.

Analyze

"Analyze" means the coding and analysis that happens after data is collected.

Typically, this involves coding qualitative data (in Qualtrics), writing code to transform and score quantitative data (in R), and running statistical analyses, usually multilevel-model regressions (in R).

Write Papers

"Write papers" means interpret findings and write papers for publication.
This involves writing all the sections of a paper, reading a few papers to provide context in the introduction and discussion, thinking about how my results fit with the broader literature, how my study is limited, and what future directions there are. This also includes back-and-forth editing with collaborators. Email is often sufficient for a lot of this, but sometimes I'll meet in person with my collaborators to talk through results with them.

I also include preparing figures (in R with ggplot2) in this step.
I often generate quick figures while running analyses, but this is the step where I generate visually appealing figures for manuscripts.

I also include the entire journal selection and submission process.
This includes considering journals to submit to (with JANE), discussing journal options with collaborators, thinking up reviewers to suggest, making accounts and submitting to journals, and dealing with reviews/revisions once they come in. As with writing, email is often sufficient, but meeting in person can sometimes be more efficient, especially with more complex revisions.

Write Grants

"Write grants" means finding grants and other funding to apply to, making an application, polishing it, making an account on their website, finding references, getting someone else to review the grant before submitting, and submitting on time. Sometimes grants also include making a budget.

Grants are about as important as publications, but grants tend to have deadlines whereas publication timelines are often more flexible. It is my understanding that grants become more and more important (and time-consuming) as one continues up the academic ladder (or pyramid).

Teaching Assistantships

Teaching Assistantships (TAing) also takes up time.
TAing is low-importance as far as careers go, but can be relatively urgent as far as deadlines go.

During my degree, I generally TAd one course in the Sept–Dec semester and one more during the Jan–Apr semester. Early in my degree, I TAd quite a few more courses for the money, but, after I got a government grant, I TAd less and less.

I have been especially lucky because I TAd my supervisor's courses.
My supervisor and I have a great relationship and he designs courses well. My TAing involved grading digital assignments and term papers, plus physical tests, midterms, and final exams. I also monitored an online discussion board and I invigilated midterms and exams, which did involve travelling to campus. In later years, my supervisor shifted my responsibilities more toward digital grading and less toward in-person tasks (great to have a great relationship!). I made myself available to students by email and never held physical office hours other than by appointment.

I have heard other graduate students describe having quite a lot more TA work.
Some came up with questions for tests and exams, which could mean they need to read the course content. Some held in-person office hours every week and/or handled general course emails. I didn't have any of those responsibilities so remember that experiences can differ greatly for different people under different professors and lecturers.

Admin and Logistics

Graduate studies involve various administrative tasks, such as putting together a PhD committee or submitting yearly progress reports. These pop up here and there, but these are not daily tasks.

More frequently, I send emails and coordinate collaborations. I take meetings with various people, e.g. my supervisor, my mentees, journalists, podcasters.

In the earlier years of my PhD, I also had to take graduate courses.
That involved attending class, reading papers, writing reflection papers and term papers, presenting, and engaging in-class discussions. Statistics courses had assignments. Everything was easy, but courses were major time-sinks back in those first three years.

As with TAing, courses fall into the category of low-importance as far as careers go, but relatively urgent requirements as far as immediate degree progress goes. Courses cannot be ignored and they eat up quite a bit of time, but they provide a very low ROI for time and effort. The best things that could come out of a course are learning a new method or writing a term paper that turns into a study or even a publishable manuscript (e.g. theoretical work or review). One benefit of a course can be the opportunity to get to know the professor since class sizes tend to be much smaller; this can turn into a partial mentorship relationship or even genuine collaborations on research.

Conferences

I search for and apply to present at conferences.
This generally involves applying, booking flights/accommodations, preparing presentations, flying/travelling, presenting, networking, flying back (sometimes after a mini-vacation at the conference destination), then filling out accounting forms to get reimbursed for the costs of the trip.

For more details, jump to Conference Advice.

Future Day-To-Day as a Professor

Professors have courses to prepare and teach, students to address, committees to sit on, more papers to review, bigger grants to write, a lab full of equipment to set up, grad students to interview and supervise, invitations to give outside lectures, industry collaborations, book chapters, etc.

Plus, you know, life.
My supervisor has a wife and two children, plays competitive ultimate frisbee, AirBnB's an apartment, etc.

Self-Direction

I have designed my experience so I get lots done, get pubs and grants, run many projects, mentor, and still have time to chill. Like I said, I work from home. I wake up around 11:30 am, eat breakfast, dick around for a bit, then figure out what I'm going to work on that day, maybe around 1pm.

When I get tired of working, I stop. I may go to the butcher or go for a walk with a friend who walks his dog. I eat dinner and I'm usually done working around 7pm, but if I'm having fun with something I might pick it up for another hour or two later on. Then I dick around and go to sleep around 2–3 am. I've got a sleep disorder, so this is my sleep schedule, but everything works out because I took the time and care to design it that way.

Index

Return to Start Here

Jump to Compare yourself to others, but don't beat yourself up