Mind the Gap – Gender and Computer Science Conferences

In September I presented my first publication together with my colleague Antonio Fiscarelli at the Human Choice and Computer (HCC13) conference. Our conference paper entitled “Mind the Gap: Gender and Computer Science Conferences” was published by Springer in the proceedings of 13th IFIF TC 9 International Conference on Human Choice and Computers, held at the 24th IFIF World Computer Congress, 19-21 September, 2018. We reworked and expanded my master thesis on Visualising Gender Balance. Ten computer science conferences and the digital humanities conference compared. First we defined different research areas in computer science using topic modelling and a clustering algorithm developed by my colleague. Next we studied the differences between disciplinary and interdisciplinary silos and which impact this had on career length, number of publications and collaboration.

Abstract

Computer science research areas are often arbitrarily defined by researchers themselves based on their own opinions or on conference rankings. First, we aim to create an automated and objective way of classifying conferences in computer science research areas based on topic modelling. We then study the topic relatedness of our computer science areas to identify isolated disciplinary silos and clusters that display more interdisciplinarity and collaboration. Furthermore, we compare the career length, publication growth rate and collaboration patterns for men and women in these research areas.

Link to the paper: https://www.springerprofessional.de/en/mind-the-gap-gender-and-computer-science-conferences/16109370.

Link to research gate: https://www.researchgate.net/publication/327225072_Mind_the_Gap_Gender_and_Computer_Science_Conferences_13th_IFIP_TC_9_International_Conference_on_Human_Choice_and_Computers_HCC13_2018_Held_at_the_24th_IFIP_World_Computer_Congress_WCC_2018_Poznan_Pola

This post originally appeared on: https://dhh.uni.lu/2018/11/22/publication-mind-the-gap-gender-and-computer-science-conferences/.

Digitising the Analogue

In collaboration with my colleague, Eva Andersen.

Some time ago it suddenly hit us that it has been a while since we read an academic book or article in a physical format. This for the simple reason that we can retrace digital information much quicker — i.e. annotating a PDF, extracting highlighted text automatically with ZotFile or quickly looking up a specific word in a pdf document. A second realisation was that as PhD students in the field of digital history and hermeneutics these “small” digital aspects of a scholars life are almost never highlighted.

When we talk about the digital turn, most people think about large-scale digitisation of historical sources and the use of computational tools and techniques such as social network analysis, text mining, natural language processing, linked open data as well as the still often used Excel or Access file. Digital humanists will not be surprised by this brief selection of methods and tools which are also often critically analysed and discussed during conferences and are slowly being teached more and more in the classroom. Websites such as the programming historian offer other interesting examples and practical explanations of different techniques.

However, as PhD students we are constantly confronted by a quite trivial but important gap between the analogue (rapidly scribbling down ideas about our research or information about sources on a stray piece of paper) and the digital world (the earlier mentioned tools and computer folders full of digitised sources and literature). In our research process we use elements of both worlds and sometimes these clash, warranting researchers to think about how to handle this. For example, reading a pdf on your computer is easy for when you quickly want to find a specific paragraph. But what with the ‘analogue technique’ of writing down remarks in the sidelines of an article or circling keywords? This is not easily accomplished on a screen. Different options such as adding ‘balloon comments’ to a PDF or inserting text boxes come to mind, but are not the same as handwritten observations.

We briefly want to reflect on how we try to bridge this gap. The previous months we have tested some tablets that could close the gap between the analogue and the digital, namely the iPad Pro accompanied with the Apple Pencil and the E-ink tablet reMarkable. Below you find some of our remarks focused on the researchers environment. We will also say a bit more about Zotero and its add-on Zotfile that could be quite useful.

Reflections on iPad and reMarkable.

The iPad Pro offered several applications for note taking, such as the free and pre-installed Apple Notes, or paying GoodNotes and Nebo. Furthermore, Nebo also makes handwriting recognition possible and can translate written notes into “typed” notes. Another plus for the iPad are the PDF viewers that work easily with the Apple pencil, allowing different colours for notes as well as supporting the extraction of text to Zotero. This ‘annotation extraction’ is made possible by a plug-in for Zotero called Zotfile. Furthermore it allows you to gather your PDF’s in a folder on your computer — if you use a service like Dropbox your PDF’s even get updated regardless of which device you use — rather than making use of the limited Zotero storage space. Depending on the applications you use on the iPad, pages can be put into different folders and all your thoughts are in one place so that you do not need to search for separate notes and post-its on your desk. In most applications you can also select different layouts to write on, although this is also possible on the reMarkable, but more on that device later. Finally, linking, syncing and using existing folder structures within applications such as Dropbox is easy. One major downside of the iPad Pro, is the backlit screen which can strain your eyes after a while. The surface to draw on is also less ‘paperlike’ because it’s made of glass — although ‘paper like’ screen protectors exist — and the iPad uses more battery than the reMarkable. Eva mostly used the iPad to scribble notes for remembering a specific name or footnote numbers to go back to after reading, or for creating to do lists. The iPad made it possible for her to gather all sorts of notes such as the number of hours she worked or ideas she had in one place.

The most obvious advantage of the reMarkable was it’s e-ink technology that made it easy to read several articles without straining your eyes. However, reading digitised sources is not always that convenient because of the scan quality. ReMarkable also gathers all your thoughts in one place which can be sorted into different notebooks or different folders . We both found that the paper-like tablet writes smoothly and also offers different layouts to write on, without using much battery. Unfortunately marked text from other PDF viewers is not copied onto the files (PDF or EPUB) when transferring it to the reMarkable. However, annotations made on the reMarkable do show up when you open the file with a PDF viewer afterwards. Although annotation extraction into Zotero is not possible. Changing the order of pages cannot be done on the tablet itself, but a tool such as PDFsam can fix that afterwards. Files can be up- and downloaded either via a USB-cable or using the reMarkable Desktop application that can automatically synchronize. For highlighting Eva missed colour, although for Sytze the three shades of grey were sufficient. Some drawings do look a little pixelated on another screen and the files cannot be paired with existing applications such as Dropbox. Sytze used the reMarkable for designing observation sheets and conducting a user experiment with the pre-designed sheets as a template on the tablet. Although it took some time exporting the design onto a laptop to duplicate the pages using PDFsam and upload the template onto the reMarkable, it meant she only carried one lightweight tablet and did not have to fuss with paper or attract the attention of the participants during the observation.

Several tools could help researchers achieve a “Digital hybrid workflow” with applications such as Dropbox, Zotero or Evernote or devices such as an iPad Pro or reMarkable. Although all these technologies have their positive and negative sides, these tools may not work for everyone in the same way. Thinkering with tools to facilitate your research can help you establish a digital workflow, which is as important as exploring the digital methods used.

This post originally appeared on the C2DH website.

The User Experience design method

When academics or developers create a website or an application, they usually start from the back end and only then focus on the design or front end. In this blogpost I would like to argue that the design process and user analysis specifically should come first. The methods outlined below were discussed during the workshop on the introduction to user experience design and evaluation methods by dr. Carine Lallemand and dr. Vincent Koenig.

During the introduction to user experience design and evaluation workshop, we dove straight into practice and had to redesign … a kids toothbrush. The purpose of the exercise was not to design the most innovative or gorgeous toothbrush, but to question an existing design based on the misconception that a kids toothbrush is just a smaller version of an adult tooth brush. Did anyone ever truly observe the difficulties kids face when brushing their teeth? Perhaps that is a better starting point for toothbrush manufacturers than their own experience. In order to design a useful toothbrush, or website, we have to understand the audience, describe the target group and study the real users.

Once we define the scope of the project and plan the user experience (UX) design process (phase I), we should start exploring (phase II) user needs through observation, interviews or surveys. The main purpose of observations is to collect qualitative information about the natural behaviour of a user. It allows us to understand the material, cultural, social and organisational context. The researcher should also confront the hypothesis and expectations to the real world, and confirm, reject or revise any assumptions. Observation techniques include tracking, fly on the wall, shadowing, contextual inquiry and undercover agent. For each user test, a designer should follow five steps:

  1. Define the research question and objectives
  2. Choose the method or technique
  3. Design and test the material
  4. Recruit participants
  5. Organise logistics

Observing users does not provide a designer any understanding of the motivations and attitudes of users. In an interview participants can freely express their opinions, while their experiences are easy to communicate to stakeholders and designers to start idea generation. Interviews take different shapes and forms, including a structured, semi-structured, exploratory or contextual format and can take place before, during or after the observation. Observations and interviews are both qualitative and rather time consuming methods, whereas questionnaires can also provide quantitative data and reach a larger number of participants. Digital questionnaires have the added advantage of saving time by automating the analysis. Before conducting questionnaires, the coding scheme and administration should be defined.

For my research on the history of the design and use of computing devices I am interested in the material objects themselves and how people perceive computers from the past. A fly on the wall study of how museum visitors react to these machines can provide insight into their experience. I would need an observation sheet and an audio recorder but I do not want to scare visitors by taking unauthorised photographs or filming them without consent. Such a user test requires permission from the Ethics Review Panel and the museum itself. If people notice that I am observing them, I should give them my contact information should they have questions or wish to be excluded.

Once the initial user analysis is finished, the ideation process (phase III) can start. For example, experience maps can describe the dynamics of a user experience before during and after usage and is based on aggregated or raw data from the previous analysis. Furthermore, designers can add both positive and negative emotions of users to an experience map or a customer journey map. During the course we used two typologies to identify both positive (Desmet, 2012) and negative emotions (Fokkinga). Besides card-based ideation, designers often rely on storyboards to visually display a scenario of use.

In the final stages of the UX design process (phase IV and V), several prototypes will be discussed and tested before the development of the website or application. Prototypes can be categorised as low or high fidelity, where low fidelity prototypes could be simple wireframe drawings and high fidelity prototypes might be simulated digitally and include more functionalities. During the workshop we brainstormed about a paper prototype for a single ‘screen’ or function and had to come up with six alternative layouts within ten minutes. Afterwards we presented our ideas to the group and received positive and negative feedback and created a final paper prototype including the best features. After every design iteration, prototypes should be tested by users either via observation of a think aloud exercise, or contextual interviews followed by tried and tested questionnaires. The workshop hosted its very own user test in the usability lab so that we could practice observing a user evaluation. The exercise demonstrated why we have to pre-test our material and practice.

The workshop provided us with several tools and practical guidelines for conducting user tests. We also touched upon the ethical aspects of user testing, which I did not outline here. The new General Data Protection Regulation from the European Union means designers and researchers have to reconsider which data they collect and ensure informed consent from participants. Nevertheless, user testing remains important if we want our designs to be usable, relevant and exciting.

This post originally appeared on the C2DH website.