I’m a post-doctoral researcher in Human Computer Interaction (HCI), and here is my parents’ answer when somebody asks them what I do for a living: “she is doing computer science”.

Although this is not not completely wrong, it is still… a bit vague. And that’s entirely my fault, because I’ve never been patient enough to tell them what I actually do. In fact, explaining to our (not so tech-savvy) friends or relatives what we actually do as HCI researchers can be, let’s be honest (and polite), quite painful.

So here is my attempt to describe the job of a HCI researcher. This post is mainly intended to:

  • parents, siblings, friends, partners that are willing to dedicate 20 minutes of their time to try and understand what their child, sibling, friend or partner is actually doing when/if they go to work;
  • masters’ students who are considering doing a PhD in HCI but are not so sure what is it all about;
  • new PhD students in HCI who are a bit lost and are not sure what they should do with their time or what their co-workers are actually doing.

I thought about discussing what topics HCI researchers work on, and what’s the point of HCI research, but the article was already too long so it would probably be for a next post. In the meantime, this is what I’m going to talk about:

Let’s start by stating the obvious: HCI researchers, after all, are mainly researchers. HCI is just their field of expertise. So what does a researcher do? First thing first, a researcher does not only do research. Researchers’ activities mainly fall into two categories : teaching and teaching-related tasks; research and research-related tasks. What differentiates HCI researchers from others is what they do when they are not teaching, reading or writing. Or, in other words, when they actually do research.

Teaching

Most researchers working in the public sector are also teachers. The official title depends on the countries (click here for the Wikipedia list), and on the rank of the teacher. Examples include teaching assistant, lecturer, associate professor and professor. Some researchers, such a the majority of CNRS researchers in France, are not entitled to any teaching duty.

The amount of teaching varies a lot, depending on the researchers’ status and their country of residence. For example, in France, a regular PhD student in Computer Science will have 64 hours of lectures a year while a “Maître de Conférence” is entitled to 128 hours of lectures, which equals to 192 of “lab hours”. In Slovenia, things are also different. At my research lab, my colleagues who are PhD candidates have to teach around 300 lab hours while as a teacher assistant with a PhD, I am entitled to around 180 hours of lectures (6 hours a week). The amount of teaching that researchers are supposed to do can also differ from the amount of teaching they actually do - it is not rare that teachers to additional hours because there isn’t enough staff in the department.

But one should not forget that teaching is not just about giving lectures in front of students. Teaching also includes: preparing the lectures, preparing and correcting the exams, entering the grades, answering students’ emails, learning about educational software or pedagogical approaches, etc. Teaching also involve activities related to the department, such as attending meetings to elaborate the new syllabus or promoting the department programs at various fairs.

Each researcher will spend a different amount of time on each of these activities: some will prepare their Powerpoint in no time at all, others will spend hours polishing it; some will never answer any emails, others will not only answer emails but also meet up with students to give additional explanations.

Research

When researchers are not teaching, they try to do some research. However, and as for teaching, doing research is not only about doing research :) Research also involves a lot of peripheral activities that can be very time-consuming and can leave little to no time at all for actual research.

Helping the lab and the research group.

At the local level, these activities can include: attending a seminar; communicating about research results through articles or interviews for the laboratory newsletter, attending meetings organized by the laboratory to address important issues (e.g. do we need to hire a new researcher) or more trivial ones (e.g. shall the lab pay for the new coffee machine ?).

A consequent amount of time is also dedicated to research funding. In order to get money for their projects (e.g. to buy equipment or to be able to go to a conference), researchers have to write project proposals, and, if they got some money, they usually also have to write reports that describe how their research is progressing and how they are using the money that they received. Finding funds for a research project often involves lots of meetings with potential partners, and lots of administrative tasks. Once a researcher get some funds, it is time to spend it wisely! Easy peasy? Not really… Researchers who need to buy new equipment may spend hours trying to figure out what is the best equipment and then even more hours to get a certain amount of quotes from different companies.

Depending on how high the researcher is in the hierarchy, these activities can take a few hours only, or most of the researcher’s time. A new and candid PhD student will very unlikely be involved in so many and time-consuming activities, except for the fun part (attending the conferences), while the director of a lab or of a research group will have to answer tons of emails every day.

image-title-here

Source: PhD Comics

Helping the community.

At a global level, these research-related tasks are often referred to as community services. This may sound like a weird expression, but it is actually quite self-explanatory. Researchers working in the same field pertain to a community. This community exist through collaborations, publications being read and distributed among researchers, networks, but also conferences, workshops and so on. Within the HCI community, there is one community that is particularly big and strong: the CHI community - CHI being the most important conference that everybody MUST attend to really consider himself/herself a true HCI researcher :) (yes, I am sarcastic).

Community services may include: helping with the organization of a conference; being a member of a committee and attending meetings; setting up a new local chapter, which will later organize seminars to help researchers share their knowledge and grow their network; being responsible for the proceedings of a journal, etc. The services can be made for regional, national or international communities.

Reviewing articles is an other important activity. The process is as follows: a group of researchers finally manage to put together an article after weeks/months/years of research. They sent it to a conference or a journal. The members of the editorial board will assign other researchers to read and comment this article. The point is to decide whether the article is worth publishing at the said venue. These reviewers are selected based on their notoriety and/or levels or expertise. As a researcher, you can sign in to some systems to say that you are willing to review, or you can receive an email reading something like: “Dear Researcher, would you like to review this article ? We need your expertise.” If you accept to review the article, your job is to read it and write down a summary that will highlight the pros and cons of the article. Based on this, you make a recommendation on whether this article should be accepted or not. Sounds simple, right?

You can have a look at this page to see some examples of good and bad reviews.

Finally, research!

When researchers are done with all of this teaching, administrative tasks and community services, they can work on their research. Finally! To me, research consists of three main activities: reading, doing, and writing. Some researchers follow these steps in a very linear way, other tend to do everything at the same time. Some love programming, others much prefer writing, very few like to spend hours reading.

Reading

So what is it all about? As researchers, our aim is to produce and disseminate knowledge. To produce knowledge, we must first know what knowledge already exist. This is the “reading” part. Let’s give an example. I am actually working on interactive web documentaries. When I started my post-doc, I had never worked on this topic, so I didn’t know what knowledge already existed on this topic. I therefore read a lot to gain some knowledge, e.g. I read articles published by other research laboratories or popular websites. I also watched several web documentaries and read Wikipedia pages about topics I was not familiar with. After reading all of this, I had a clearer idea of what were the actual “gaps” in the literature. For example, I found out that the interactivity of web documentary has not been studied much, i.e. we do not really know whether people really enjoy clicking on different parts of a website to access some videos or if it actually annoys them so much that they close their browser because of this (this is a bit simplified, but that’s the general idea).

There are thousands of thousands of articles out there, published by thousands of researchers in thousands of venues. Reading without a specific purpose in mind can be extremely time-consuming. However, reading on a regular basis is still very important: any article can prove useful at some point, new articles are published everyday and one should stay up-do-date, reading papers can inspire new ideas and reading articles is the best way to learn more about one topic.

A good thing to realise how many articles are out there is to go to Google Scholar and to type in some keywords. For example “smartphone food” lead to… 168 000 results!

Doing stuff

After/while reading, a researcher will come up with some ideas on how to “create” knowledge, e.g by starting addressing a problem that hasn’t be resolved yet. This part of doing research can be extremely diverse, especially in HCI. There is no one single procedure to follow, no single methodology.

How researchers will address a problem depends on their background, skills, interest, time they are willing to spend on that project, fundings and equipments they have, etc. This can include: conducting interviews, developing a new software and evaluating how it works, developing a new theory, designing a new algorithm, analyzing large data sets, comparing different ways of interacting with a smartphone or a computer (e.g. vocal commands vs gestures), building a new system (e.g. an invisible drone)… the list is endless!

Researchers will spend hours implementing their ideas. Of course, the first idea is rarely the best one. So researchers will spend even more hours trying out new ideas, discussing their ideas with their peers, redefining their original research question, implementing new ideas and testing them out. And so on and so on, until they reach a point where they think their idea, and its actual implementation (be it a theoretical framework or a very futurist technology), is actually a worthy “contribution” - it adds knowledge to the existing knowledge. It might sound simple, but this is the most engaging and demanding part of research.

According to J. Wobbrock and J. Kientz, seven types of HCI contributions exist. I’ll focus on three common types:

  • Artifacts. This is, I believe, the most popular type of contribution within the HCI community. The idea is to design a new system, tool, piece of software, technique, you-name-it in order to better understand how this new you-name-it could help people achieve a task more efficiently, improve accessibility, change the way people consider/use technology, facilitate communication, etc. The purpose of this new artifact can vary a lot and be perceived quite differently: artifacts that I might find not so useful might be seen as real game-changers by other researchers :)

  • Empirical contributions are based on data. Data may come from experiments, interviews, sensors, observations. Anything. Thanks to these data and through hours of data analysis, researchers try to answer one question. For example, if the question is “is it easier to read white subtitles on a black background rather than yellow subtitles”, researchers will collect data (e.g. by measuring how many seconds people need to read a sentence) and then analyze it to conclude something like: “people need less time to read yellow subtitles than white subtitles”. Conducting experiments is therefore one of the key activities of many HCI researchers: whether in the lab or in the wild, experiments allow them to empirically measure how people interact with systems.

  • Surveys are also called literature reviews. They summarize existing research by comparing results, artifact characteristics, methodologies, etc. Their aim is to help structure knowledge on what topic “with the goal of exposing trends and gaps”. Compared to artifacts and empirical contributions, surveys may seem easier to do as they only require reading, analysing and writing, but they can also be very time-consuming. A good survey will help other researchers identify what knowledge already exist, what questions haven’t been addressed yet, what are the best tools and techniques that have been used and investigated so far, etc.

Writing

To me, that’s the most interesting part. For many researchers that I know, this is the boring part. But writing is really important, because at the end, the article is the main way to disseminate knowledge that researchers somehow managed to create. HCI researchers usually publish 10-pages long articles in conferences or 20 to 30-pages long articles in journals. Regardless of the venue, one article usually contains the following information:

  • what is the research about?
  • what was the literature saying, and how does the proposed research contribute to the literature?
  • how was research conducted and what were the results?
  • what are the key findings and limitations of the research work?

Writing an article may sound easy, but it is actually demanding. Each word is important, and each sentence is important. Researchers have to be sure that any other researchers will be able to understand their work. Figures also play a major role and designing good figures can also take a lot of time. Adding references to other articles is also quite boring and time-consuming.

Writing an article is not a linear process: I do not know anyone who is able to write an article from A to Z without jumping back and forth between the sections, editing, adding or removing paragraphs. Also, and because researchers usually write articles collaboratively, articles go from one researcher to another, leading to different versions of the articles with comments, suggestions, etc. which must all be taken into account. Writing help to make sure that an idea is clear, or that a methodology is sound. If I am not able to write down a correct and interesting research question, they are high chances that I am actually not sure what the research question is. In highschool, one of my teachers used to remind us of one citation from Boileau, whenever we had to write essays:

“Whatever is well conceived is clearly said, And the words to say it flow with ease.”

This couldn’t be more true when it comes to writing research articles.