InfoTypography work being presented at SIGGRAPH this week

This week I will be presenting at ACM SIGGRAPH (the top conference in graphics and one of the most prestigious in Computer Science) the work that Johannes Lang and I have done in Infotypography. We looked at how people perceive typographic parameters. This could support the use of these parameters to represent information.

Here is a teaser of the talk (30 seconds):

Here is a fuller, 15 min presentation ():

Here are two links to the paper:
Local copy

Here is a website with resources for using it:

For those of you who are in Vancouver attending SIGGRAPH, the presentation is:

When: Thursday August 11, 2022
Where: East Building, Ballroom C
Session: Perception

How do people think about Constraint Problems?

A paper on our work about how people think about constraint programming has just been published in anticipation of the prestigious Conference on Principles and Practice of Constraint Programming (CP 2022). The paper is accessible through this address.

The work takes place at the intersection of Human-Computer Interaction and Constraint Programming.

We analyzed how non-experts tried to solve constrained programming problems.

Image showing a few examples of how people solved common constraint programming problems, such as scheduling.

This is work with Ruth Hoffmann, Xu Zhu, and Özgür Akgün, from the School of Computer Science at the University of St Andrews (my previous main institution).

This follows our previous work on how people visually represent discrete contraint problems (at the IEEE Transations on Visualization and Computer Graphics).

Dr Hoffmann has also presented at ModRef a synthesis on both pieces of work.

Finally, the SACHI group posted a blog post with a bit more information.

Student Paper on Representations and Transformations

Adam Binks (A PhD Student at the University of St Andrews) and myself (Miguel Nacenta) just published an article in the International Journal of Human Computer Studies) about our recent research.

Adam built a new tool to support people when transforming complex networks of thoughts and ideas into text prose. The tool itself (called Write Reason) and a short description of what we found when we studied essay writers using the tool is accessible through the WriteReason site.

One of the most interesting things about our findings is that there is a lot of interesting stuff going on in between the creation of different representations of the ideas. In other words, transformations between representations are much more complex, interesting and important than we thought before.

This post is also in the VIXI website.

How to talk to your supervisor about what you have read

This is a post in a series of posts regarding student-supervisor communication.

Audience: you are a PhD or MSc or project student, and you meet your supervisor regularly. In the previous meeting, your supervisor asked you to do research in some topic area, perhaps by a keyword (e.g., please do some search in the area of “provenance in User Interfaces”), or perhaps through another paper (e.g., follow the references in this other paper).

Naturally, you have done the work, and you have read (there will be another post on what this means in the future) a bunch of papers. Now you come to your supervisor, eager to demonstrate that you have done what you were expected to do, and indeed, you have found and read some interesting papers.

What NOT to do

Perhaps more important than the “do’s”, consider the things that you should avoid. Unless your supervisor has explicitly asked for any of these, you should NOT:

  • Detail the contents of a paper in general (your supervisor would generally not be interested in the full content of the paper. If they already know it what is the point on repeating pretty much everything on it? If they are not familiar with it, they probably will not be interested in the full content, but only in how it relates to your own project; your supervisor has probably very little time for this, and communication about previous work usually needs to be done in an efficient way.
  • Read a summary that you have written (if you have written something about a paper, your supervisor can probably read it much faster than you can read it aloud – if you really need them to read your summary, and they agree with this, send it in advance of your meeting).
  • Forget, loose or misplace the paper (your supervisor might want to come back to it, it is your responsibility to be able to retrieve that paper quite quickly if you need to).
  • Insist on going over all the information that you have retrieved without paying attention to your supervisor’s reception of it (your supervisor). Do not discuss things that you do not think are important.

What to do

These are all items that you might not know are important, but are often key for good communication about research.

Background and context: Not all papers are created equal or are going to have the same importance. If you only list the title (and yes, the title is important), you are missing many of the ways in which your supervisor might help you filter out the really relevant papers from those which are not. More specifically, this is information that is likely to be useful:

  • Year (of publication). This information is key. The paper might be precisely about your topic of interest, but too old to have been important. Perhaps the opposite: the paper is very old and incredibly relevant, but has been ignored. Or the paper just came up last month and your supervisor is not aware of it (and should be!). Making all these judgments is much more difficult without the date.
  • Authors. It is important for you to start recognizing the people who have work a lot in that particular subarea, and their collaborators and students. Your supervisor might use that information to remember about other important papers by the same author, or to assign more weight to that paper because they trust the author (or even know it personally). Put the authors next to the title if you are showing only a subset of the paper. Sometimes knowing which institutions the authors come from can be useful. For example, if they come from a place that is well-known for research in the area.
  • Venue. Although it is not a perfect correlation, a paper published in a reputable and competitive venue is likely to be more important, since it denotes interest by the community, and some degree of rigour. A paper in a leading venue which is ignored in a literature review is also likely to cause more trouble in the review cycle, whereas most reviewers will easily forgive (or forget) a poster in an unknown or disreputable venue.
  • Format/length. Scientific publications come in many shapes and forms. For example, a valuable piece of scientific writing can be a textbook, a chapter in a textbook, a chapter in an edited book, a survey article, a long journal article, a paper, a short paper, a poster, a position paper, or even a blog or a tweet. It does not help if your notes, summary or what you are showing or describing to your supervisor hides what kind of publication it is. Your supervisor (and yourself) can use this information to determine the importance and value of that work.

Note that giving some kind of impression of the above should take very little time. You do not have to list or read aloud the whole list of authors, but having it out there so that your supervisor can take a glance at it can be very useful.

Content: An article or piece of scientific literature typically contains much information, more than you can efficiently convey within a meeting. If you have to summarize any part of the work, because you think it is relevant, this does not mean necessarily that this is going to be a piece of text. There are often different kinds of elements that might be way more effective at summarizing a paper. For example:

  • A diagram included in the paper.
  • An image of the system/technique that describes the core of the paper (or the part that is most related to your work).
  • A short fragment of video.
  • A formula.
  • A table (with results or a taxonomy)
  • A chart showing results (e.g., a bar chart).
  • A couple of sentences describing results of conclusions (yes, this is text, and is legitimate).

Note that any of the elements above might be incredibly difficult to understand in isolation, but this is why you are having a meeting: a quick few sentences can help make any of these items really effective.

Relationship: Probably the most important reason why your supervisor wants you to read a piece of scientific literature is because it relates to your work. Very often this relationship remains implicit. You should be able to describe as succinctly and accurately as possible how what you have read relates to one or more of the following (depending on the stage of your research):

  • The fundamental issues, techniques or technologies upon which your work builds upon.
  • One or more of your research questions.
  • The specific contribution that you are hoping to make in your research.
  • What you have designed/built/proposed already.
  • The methodologies that you have applied.

By making explicit the relationship between your work and what you have read, you are also communicating a number of things that are important for the supervisory relationship and signalling your level of maturity in research. For example, your supervisor will learn to what extent you are able to relate your work to the close and fairly distant (perhaps only abstractly related) pieces of literature, and whether you really understand the contributions of existing work. This is a kind of syntopical reading or deep reading that is very important for academic work.

Please let me know if you have found this article useful, or if you have concrete examples, or possible improvements that you think I can use to improve it.

Postgraduate Supervision in the COVID19 Era (for Supervisors and Students)

I moved to the University of Victoria (Canada) in January, so I have since been working with my St Andrews PhD students remotely who, for a number of reasons, preferred not to move continents to Victoria with me. This means that we are quite used at the moment to work remotely.

Remote PGR supervision has its challenges, yet I believe it is important to acknowledge that the new constraints imposed by the virus, quarantines, social distancing etc. add an additional layer of complexity to the whole situation. I also believe that postgraduate students are likely to suffer the extraordinary situation particularly hard because with the additional burden to move to online teaching, supervisor’s energy and time will be in shorter supply. Here are just a few of my thoughts regarding remote supervision (first), and regarding supervising during quarantine (second).

Remote Supervision

I think it is important to acknowledge that remote supervision is harder than in-person supervision. For a number of reasons:

  1. It is harder to read the mood of people through a remote connection, even if there is video (this cuts both ways… I know that my students can sometimes tell that I’m in a bit of a mood). Without the ability to judge mood and attitude accurately it is harder to make the right calls regarding next steps, tone etc.
  2. I find that it is just that bit harder to be compassionate, empathetic and understanding when you are separated by a long distance and, perhaps more importantly, by a few hours in our daily cycle. I meet my students in the morning, since I’m 8 or 9 hours behind. This means that my students are mostly at the end of their days when I’m still full of energy; this has a significant effect.
  3. The serendipitous and opportunistic contacts that one has at work are important for students; if you cross them in the hallway, they know you are here, you know they are here. A quick question or two minute talk, or even a sentence might not only give each other acknowledgement of each other’s needs (“yes, I’ll get to read your chapter soon! I have not forgotten!”), but it might also save a significant amount of time if they don’t have to wait until the next meeting to get a little bit of help.
  4. The opportunistic encounters with other members of staff are also important for the students and for the supervisor in order to support the students. For example, that little bit of awareness of the member of the Systems Support Team who tells you that certain service will have to be down soon, or that a certain machine needs replacement.
  5. It simply is harder to critique, point out issues, discuss ideas that have a graphical substrate when you are connecting over video conference. I cannot count the number of times that I had to sketch something on paper and awkwardly held it in front of the camera, trying to point at different areas of the sketch, or the number of times that I tried to scroll on a document which is really just a shared desktop from the other side. This is a real issue in Computer Supported Collaborative Work that has not really been solved yet in the mainstream, despite the efforts of many in the CSCW community, including my mentors Saul Greenberg and Carl Gutwin.

Supervision in COVID19 times

Everything in the previous section applies to the current quarantine situation, sometimes amplified. And some aspects are new:

  1. Both supervisors and students are anxious, even if they are often not ready to admit it to themselves, much less to others. Being a PhD student is more often than not a difficult thing, with the justifiable stress and anxiety of academia and trying to prove oneself. Now imagine how this feels when you add the uncertainty of not being able to run experiments, have access to your usual materials, or having to go through extra hoops to just be able to do what is expected of you. Anxiety and extra work can be particularly damaging for deep work.
  2. Loneliness can be a real issue. Most people rely on looser or tighter communities for support. Postgrads who work remotely tend to find it much harder to find the motivation and energy to sustain focus and work without a group of peers. Self-isolation and social distancing is much more likely to wipe out these sources of peer-support than many other practices, probably because the value and perception of these networks is easy to underestimate.
  3. Some people’s life circumstances are being affected more than others. Being away from family, having to care for dependents, or even trying to help out the communities one belongs to (e.g., elderly residents in the same apartment building) can take a very significant toll on people’s time and energy.

Strategies and tools worth thinking about (for remote, COVID or both)

I have never been a perfect supervisor, but I try hard to do right by my students. Despite knowing that I can get much better at this, I believe that there are some obvious and not so obvious things that can work:

  1. Remember that we are all, above anything else, people. With our vulnerabilities and needs, good and bad moments. It is surprisingly easy to forget to be human, and caring and compassionate and just adopt a role as the conscientious student or the wise and driven supervisor as we interact with each other. Although it sounds obvious, make sure to start every contact with a sincere effort to relate, connect and care. This applies to both students and supervisors.
  2. I believe that it helps to know that it is normal to not be as productive remotely as you imagine you have been in the past. This is even more important within the current unprecedented circumstances. Procrastination and lack of focus is not just a matter of pure will (no matter how many times we convince ourselves that we can just “power through”) and achieving sufficient progress in research work is a constant challenge that we fight everyday, a problem never definitively solved. Therefore, if you did not have a good day, forgive yourself a bit, practice a bit of self-compassion, and reassure yourself that it will get better tomorrow.
  3. It helps to know that for most institutions, the top allegiance is to their students. Despite many people’s cynicism, I have found this to be consistently true throughout my experience (UVic is my 8th institution, not counting Microsoft Research). As a student you might be feeling very insecure about deadlines, regulations and funding. It is good to have some pressure to make sure that the work keeps moving on but, in my experience, an overwhelming majority of people in academia understand Hofstadter’s law, and institutions tend to know to be flexible when they need to be. You still have to do your best, but you will get some leeway.
  4. Listen extra hard. This applies to both sides of the supervision relationship. It is particularly hard to listen and connect when one is stressed and overloaded, so it is particularly important to remember to listen in these circumstances, and give each other the opportunity to share the challenges, problems and feelings arising.
  5. Communicate to your supervisor. They can help you. One of the harder tasks for a supervisor is to assess the state of mind of many of the students. As a student, it is tempting to try to present a flawless appearance of invulnerability and performance, but it is much better to actually share the challenge and the feelings. This, of course, relies on the trust invested in the relationship, and it flows in both directions. Although difficult to do, I think it pays off to share the challenges and vulnerabilities that we experience as supervisors as well.
  6. For supervisors: It helps to have great colleagues (who can be proxies) and are physically closer to the students. This is more relevant to the more general issue of remote supervision than to the current quarantine situation, but it can help as well. I have been lucky to have good collaborators still in St Andrews (call out tothe  amazing Juan Ye, Ozgur Agkun, Alice Toniolo, Uta Hinrichs and Aaron Quigley) who are good points of contact for the students when they need to. They can also supply a bit of that co-located experience that is so reassuring.
  7. It is helpful for both student and supervisor to acknowledge that working remotely requires additional “meta-work” to work smoothly. That is, it takes time and work to figure out how to work together more efficiently. This can take the form of extra thinking about how to best schedule meetings, figuring out substitutes for support, or even spending some time learning and finding new tools and technology.
  8. Some new tools are actually quite amazing, and can help (and they keep evolving!) Although I personally find some of the new communication trends such as Slack quite overwhelming and difficult to keep up on top of (too many “channels” of information to keep track of, when considered on top of e-mail, which is still king for me), some other tools can be fantastic. Sometimes one has to even remind itself to use these tools, even if we know that they will work better. My two favourite examples right now are miro, a collaborative diagramming tool that works well when used synchronously with a source of audio on the side (typically Skype for me), and opening a shared Google Doc when brainstorming ideas, which I found can be transformative.

Some of you might have different tips and ideas. I’m sure I have missed a lot here, and I can learn a lot more. Even better, some of you might disagree with me. In any of these cases, I’ll be very happy if you let me know through a comment or by e-mail.

Book Light Review: Object-Oriented Ontology: A New Theory of Everything by Graham Harman.

This is a first blog review in a series for books that I finish reading for professional reasons, usually within the context of my research. I call them “light reviews” because they are not meant to be comprehensive and because the book might fall outside my direct domain of expertise (information visualization, Human-Computer Interaction, Infotypography).

[link to publisher’s site]


Why I got it/read it?

The title looked promising; I’m currently interested in how people model the world, and both ontologies and theories are good examples of forms of knowledge that represent the world for the human mind. Also, I cannot generally resist buying these kind of books at one of my favourite bookstores in St Andrews: Topping and Co. (I’ll miss you!).

What did I learn?

Object-Oriented Ontology belongs to a current trend of philosophical work that calls itself “realist”. It is quite contemporary and seems very influential in other areas such as Architecture and History. OOO relates to Latour’s philosophy (I’m generally a fan), and his Actor-Network Theory. OOO addresses the issues of truth, reality (not the same thing as truth), knowledge, causation, and perception.

A key part of the intent of this theory is to remove the human/mind/cognition as a central actor of any theory of the universe. It tries to create a theory that does not depend on humans or any other thinking being to explain reality and the world. Objects relate to objects, and humans, the mind, etc. are just other types of objects. In OOO if the proverbial tree falls without anyone noticing it, definitely still fell. At the same time, in a direct line from Ortega y Gasset’s philosophy, it denies the possibility of directly knowing real objects, since these are not directly accessible (even by other real objects), and instead only relate to other real objects through other kinds of objects (sensory objects), and their qualities (real qualities and sensory qualities). From the relationships between these four types of things, OOO derives, initially quite unintuitively, but also quite convincingly (at least for me) the concepts of space, time, and a full system that explains how do we actually get to access objects (surprise: through metaphor), what we can learn about them, and also new forms of analysis.

Objects in OOO are a very generic term that includes things that are not material, or even imagined etc. For example, a commercial company is an object, Spain is an object, a dream is an object and, of course, material objects are also objects. Combination of objects are also objects. Objects not being directly accessible is provided as an explanation of why a description, formal description, or collection of properties of a real object is never going to replace the real object itself (this is connected to Ortega’s perspectivism).

I found particularly useful the developed terms of undermining and overmining. This has to do with abstraction and compositability. Other existing theories are undermining if they try to reduce objects to its components, and overminining if they try to explain everything based only on the properties that it has. OOO provides a way of defining objects as a balance between overmining and undermining.

What did I like about the book?

The book is actually quite accessible to non-philosophers, and is a good read. It is also quite good at relating OOO to existing contemporary and non-contemporary theories and philosophies, which makes me feel a bit more confident about my knowledge and gives me a useful context within the realm of theory.

What did I have trouble with?

The mechanisms of “vicarious causation”, in which objects affect other objects through the creation of intermediate objects was difficult for me to digest. I’m not sure if this is a weakness of the theory, the book, or just something that is difficult to grasp.

How it can help (with my research)?

OOO seems like a useful lens to look and understand the role of humans in the larger systems where they inhabit, without having to reduce everything to the mechanistic, reductionistic or materialistic views of contemporary hard sciences. It might be particularly useful to explain interaction and the interface (although I haven’t worked this out yet), and it might enable a nice way to reconcile results and theories from the social and humanities side of the spectrum with more formal views from math, physics, data science and cognitive science.

Open questions and other thoughts.

In some ways, the inescrutability of real objects or “objects in themselves” remind me of the identity functions (self referential) in category theory (from Maths).

Worth reading (1-5)?

For it’s potential to be useful (to me and the field), how it made me think and see things in a different light, and how it helped me learn about OOO and other bits of contemporary philosophy, I think this deserves a 5 (highly recommendable)

Moving to Victoria, British Columbia, Canada

It has been a great 9 years in St Andrews, Scotland. Thanks Scotland for having been so welcoming, and thanks to the University of St Andrews for having supported my academic career. I will miss St Andrews and the East Neuk of Fife very much!

Although I keep working with my St Andrews friends/colleagues, and plan to do this for a long time, the University of Victoria (UVic), in the West Coast of Canada is now my new academic (and family) home. The city of Victoria is also a great place to live, with amazing views, nature and art, a growing Tech startup scene.

Aerial view of the University of Victoria

The University of Victoria is laid out around a circle (the Ring) and very close to the water.

If you are student and are interested in my research topics, I am actively searching for students, mostly in the following categories:

  • UVic honours project students
  • UVic independent study projects
  • future PhD students from UVic, Canada and overseas

I will soon refresh the research pages in this website; in the meantime, the most reliable indicator of my current research directions are my most recent papers. If you are interested in exploring possibilities working with me in Visualization, Problem Representation, Infotypography, Visual Language Interfaces or Human-Computer Interaction, just drop me a line at nacenta at uvic dot ca.

The world’s population in 1880 people per pixel (or 4878 digits)

A few years ago I came up with the idea of FatFonts, a special kind of digits that encode quantity both in the shape of the digit (as in regular numbers) and in the amount of ink or black pixels (the area of the glyph is proportional to the number it represents). I then worked with Uta Hinrichs and Sheelagh Carpendale to develop the idea and publish a paper.

Numbers in Cubica FatFonts

The numbers 19, 28, 37, 46, 55, 64, 73, 82 and 91 represented in Cubica FatFont.

Although a quirky idea, FatFonts seem to have a bunch of usages… for example, they are convenient when you want to provide a table of numbers that is also a graphical representation. This allows the viewer (or the reader) to very quickly capture the overal distribution, but also to go in and read the specific number, which they can then use to compare to other numbers (in the FatFonts table or in their heads).

FatFonts are great in maps, and that is why Uta and I set out to create a poster that would give a picture of one of the most pressing issues of our time: world population. Thanks to SICSA (and our wonderful helpers Carson, Jed, and Michael), we got the time, money and support to develop the idea. The result is a poster that represents the population of the world using FatFonts.

An overview of the FatFonts poster of the world population.

An overview of the FatFonts poster of the world population.

The poster is made using an equal area projection of the world, and it represents data collected by CIESIN and others. Each grid in the main map, which represents an area equivalent to 200 by 200 km has a 2-level digit FatFont digit in it. That way we can know, with a precision of 100,000 people, how many humans live there. Naturally, the precision is as good as the data (and these are projections using 2005 data, the newest available), but it gives you a really good idea of where people really are. In fact, the map is so mesmerising that I have learnt a lot from it by just spending a lot of time looking at it. It is not only the distribution, but also the numbers. Obviously I am biased, but I strongly believe that seeing the numbers gives you a lot more than just representing density with colours, in part because colour scales are very arbitrary.

Since the number of dark pixels of a FatFont digit is proportional to the number that we are representing, we can calculate how many people each black pixel represents. For an A1 poster in the main area of the map at 600 pixels per inch, each pixel represents approximately 1880 people! FatFonts with the orange background are up by an order of magnitude, so there the ink of a pixel represents approx 18,000 people.

The South Eastern Mediterranean population is concentrated in the Nile delta (Egypt) and Palestine and Israel.

The South Eastern Mediterranean population is concentrated in the Nile delta (Egypt) and Palestine and Israel.

We partnered with Axis maps, who make wonderful typographic maps of cities, and we are selling them here. All the profits will be reinvested in research (e.g., helping pay research internships for students). We think that they are a wonderful present and that they are really fun to look at and discuss.

To give you a better feeling of the map, and because we like to try our new stuff, we have taken some Lytro images of the poster that you can explore in this gallery.

Books for a good PhD start

A research career is a complex career. It involves many skills and knowledge that are not necessarily related to the specific topic that you choose to investigate.

In my experience, students just before of at the beginning of their PhD research (at the beginning of their research careers) are often quite disconnected from the actual skills and background that will make them successful. This is why I try to supply my students with some of the knowledge that, sooner or later, they will need to apply. To help with this, I have selected four books that all my students get at the beginning of their PhDs (to read in their free time). Here is the selection, and why I selected each book. Note: I’d love you to share other books you think are valuable at this stage (use the comments below).


content1. The Craft of Research (Booth, Colomb, Williams)

Science/Engineering students often think that writing is the boring part of the job. Most of them realise that they have to do it, and some might even know that they have to do it well to be successful. However, telling a student that they have to get better at writing is not the best approach. In the best case, they already know that they have to do it, and in the worse, they might start hating it.

Instead, I like to consider writing (of academic papers and reports) as thinking on paper. It is often not until I have written the last bit of a paper (e.g., the discussion section) that I fully understand the research that I have done, the implications, and the value of it. Of course, the research is mostly already in your mind (and in your code, data, etc), but putting it on paper takes you to the next step: you can communicate it to the world and, perhaps most importantly, to yourself.

And this is what this book is about: setting up questions, understanding the problem, structuring a solution, all mediated through writing. A particular favorite of mine is the bit about making arguments; being able to make a claim and support it with evidence in a convincing way is one of those things that students think they know, but only learn after supervision, much experience and, perhaps, reading this book.

Cover of the book "The Elements of Style" 2. The Elements of Style (Strunk & White)

Once you know why you are writing, you need to know how. Although some authors think that this book might not be as good as everybody else thinks it is, it takes many students out of some of the worst habits in writing, namely:

  • Writing to look smart.
  • Writing without thinking of the reader (e.g., long sentences).
  • Writing to fill in the space (lack of brevity).

Although the grammar advice might be somewhat antiquated and not always completely correct, the rest of the book, in particular the parts about style, helped me significantly improve my writing (although I certainly don’t claim mastery!). I think this book is particularly useful for students who are not native speakers of English and who come from traditions where clarity and brevity is not as central as in the English speaking scientific community (I’m from Spain, and I’m in shock most of the time I have to read or review a thesis in Spanish).

The most important point of the book might be summarised in a quote attributed to Blaise Pascal: If I had more time, I would have written a shorter letter. Well, a student should have the time, so the text should be shorter while keeping the crucial information. It takes time and effort, but readers (and markers) will be happier, the world will waste less paper, and the paper/dissertation will be more likely to be read and used by others.

Cover of the book "The A PhD is not enough!" 3. A PhD is not enough! (Feibelman)

Very often students lack context. They might know that they want to do research, they might even know that they like research. But what else is involved? Why would a PhD be useful? What does it get you? Most importantly, what does it NOT get you?

This book might be a bit harsh to start on (sometimes reality is a bit hard), but it provides a nice glimpse on the world of research and highlights much of what really becomes the focus of what you do as a researcher and academic. The bad news is that there is a lot more of politics, strategy, and marketing in this job than what we all expect when we start. The good news is that you can be prepared for it, and might even get to enjoy some of those bits. In any case, and from my personal opinion, being in research is awesome, but it is better to be ready for what it requires from you.

Note: there are other similar/related books about research and academia that are worth mentioning and reading (e.g., this, this, and this), but perhaps not strictly necessary at the beginning of a PhD).

index 4. Getting Things Done (Allen)

So, what is really required from a PhD? Effective work and perseverance. Most people in academia know that you don’t have to be a genius to get a PhD. Gosh, you don’t even have to get the best or most novel ideas. But your ability to work hard, avoid procrastination, and persevere will determine the chances of being successful in your PhD and of being able to take your career further.

Although there is a lot of crap in the self-help and productivity literature, this does not mean that it is better to ignore it all. This book describes my favorite system, and although it is not perfect and I still work really really long hours, it has helped me enormously. This might not be the best system for productivity that there is, or be the best system for everyone, but at least is honest, well explained, and feasible. I’m a fan.

The reality of a PhD is that, if students think they are busy during their undergrads or MSc, the demands on time will only keep increasing. This is certainly true after you have become a doctor. If you don’t like GTD, you better find something else!


Have you come across other books that you think are useful? I’d love to compile a list with your suggestions, and I might even add a book or two to my list!

What and how to log in your experimental HCI software

You have worked hard on your project. You searched the literature, learned about the methods, painstakingly designed an experiment, and have almost finished implementing the software, but… what about the logging?

Most students think that logging is easy. Just write some lines on a text file. A couple of hours on the software should do, right? I don’t think so.

Experience has shown me that logging is extremely error-prone, and that paying little attention to it results in incredible loss of valuable information and time, and that most students do not realize how important doing good logging is.

Before I go on let me qualify a little bit what kind of logs I’m talking about. I’m referring mostly to the kind of experiment that you often see in HCI or Experimental Psychology, were there are many participants, and each participant performs many trials, possibly in multiple conditions. This is usually information that is suitably recorded in a simple format like comma separated values (CSV).

Let me state then my five fundamental laws of logging:

1. Log everything you can

Disk space is cheap, your time is expensive. Probably the most common mistake here is not to record enough information in each record because you think it is redundant and a waste of space. For example, why record today’s date in each record of each trial, if they are all the same? I tend to record the same information in each trial anyway, because it is always easy to discard info, but it takes a long time to recover data from different sources (including the name of the file, the creation date that the OS stores etc.). Don’t assume that you will remember where you are storing all that information when the time for analysis comes. Things that I tend to save in each trial record: userId, all condition values for all factors, the number of the trial with respect to the condition, the number of the trial with respect to the cell, the absolute number of the trial within the experiment (and the phase) and, of course, all the dependent variables.

Perhaps the only caveat to this is that all this recording should not negatively affect the performance of your program or the accuracy of the time measurements. If performance and timing accuracy is important, good strategy is to write first to memory, and only save to disk in between trials (or when timing is not an issue).

2. Make your logs self-contained

Name your variables wisely, and always include names on top of the file. This should be quite explanatory, and most analysis programs (e.g., SPSS) will allow you to name the variables automatically from the file. Handy and convenient. A good complementary practice is to have some comments (or a separate file) that provides an explanation of how each variable is recorded, but this requires discipline to maintain, because the logger program tends to evolve. Best to keep your measures simple.

3. Debug, debug, debug

Never assume that your code is recording properly. Simple visual inspection won’t cut it. I have experienced many problems that only became visible after all the experiments were recorded. The best way to avoid problems here is not only to debug, but also to use your pilots to gather realistic data, and analyze it in the same way that you will analyze the overall results from the finished study. This is not only good for your logs, it is also helpful to avoid possible flaws in your statistics (e.g., I do not believe in a posteriori power analysis).

4. Backup, backup, backup

Don’t trust your hard-drive, don’t trust your experimental software. Within your program, save the data to drive as soon as you can (but take into account the comments in point 1). This will allow you to recover from failures in your software. It is actually kind of nice to code your experimental software so that you can restart it again at any given trial within the session. When the experiment is finished, the first experimenter action should be to verify that the data is in the right place, and perhaps making a copy (or send yourself the data to your gmail account – if your data is properly anonymized, of course).

5. Protect yourself against confusion

If something can go wrong during the experiment, it probably will. It is good practice to save the date hour and second of experimental recording in the name of the file that your program saves. This will help you prevent accidental overwrites. Similarly, try to leave as little as possible human intervention for the actual session. For example, I never trust the experimenter -often myself- to select the right name for multiple files depending on the condition. Have the software do something reasonable for you. The only thing that I often make configurable is the participant identifier, so that I can separate real trials from debugging logs.

Hopefully these might be useful to you some time. Write a message below if you agree/disagree or want to add some more advice!