Category: emerging technologies
I was at a small gathering last week, called the Digital Learning Research Network. It was hosted at Stanford and it aimed to explore the messiness of digital learning. This was not representative of Silicon Valley’s uncritical love affair with technology. Many colleagues wrote reflections about it: Catherine Cronin, Kristen Eshleman, Josh Kim, Jonathan Rees, Tim Klapdor, Alyson Indrunas, Adam Croom, Whitney Kilgore, Matt Crosslin, Laura Gogia, Patrice Torcivia, and Lee Skallerup Bessette (to name a few). When was the last time you were at a small conference, other than the ones focusing on blogging, and this many people took time after the event to blog about it?
The messiness of digital learning isn’t a new development. It is something that educational technology evangelists ignore, but as a researcher who has an affinity for qualitative data, and as one who is increasingly using data mining techniques on open social data, I can tell you that mess is the norm and not the exception. I’m not the only one.
For me, the conference questioned educational technology but looked to it for empowerment. It critiqued universities but saw them as places to create a more just and equitable society. It brought attention to the US-centric conversations happening in this space, but recognized that we can learn from one another. It sought research, but did not seek to emulate research-focused conferences. It allowed Dave to share his thoughts but called him out on it when it was time to stop. ;)
I see the conference as the start of a longer and larger conversation. Many of us are doing research in this space and many were missing. Let’s expand the conversation.
Recently, I had the privilege of organizing a workshop for the Faculty of Humanities and Social Sciences at Athabasca University. The goal was to help the organization work through what they might need to do to put in practice a new strategic plan which calls for student-centered and open digital learning. I used the slides below to assist faculty, instructors, and instructional designers translate theory into practice.
An interesting article this morning from Jeff Young at the Chronicle of Higher Education notes:
One of the obstacles to bringing “adaptive learning” to college classrooms is that professors, administrators, and even those who make adaptive-learning systems don’t always agree on what that buzzword means. That was a major theme of a daylong Adaptive Learning Summit held here on Tuesday. Several people interviewed at the summit, held by the education-innovation group National Education Initiative, noted that part of the problem is a proliferation of companies that make big promises based on making their technologies adaptive, yet all use the term slightly differently.
I would counter that the big (and unsubstantiated) promises are a greater problem than the buzzwords, but the lack of clarity on what these concepts refer to are an issue, too.
The introductory sentences from Online Learning: Emerging Technologies and Emerging Practices (the second edition of the Emerging Technologies in Distance Education book I edited, which is forthcoming in 2016), make a similar argument:
Many of these (new) approaches to education and scholarship can be categorized as either emerging technologies (e.g., automated grading applications within MOOCs) or emerging practices (e.g., sharing instructional materials online under licenses that allow recipients to reuse them freely).
The terms “emerging technologies” and “emerging practices” however, are catchall phrases that are often misused and haphazardly defined. As Siemens (2008, ¶ 1) argues, “terms like ‘emergence,’ ‘adaptive systems,’ ‘self-organizing systems,’ and others are often tossed about with such casualness and authority as to suggest the speaker(s) fully understand what they mean.”
A clearer and more uniform understanding of emergence and of the characteristics of emerging technologies and practices will enable researchers to examine these topics under a common framework and practitioners to better anticipate potential challenges and impacts that may arise from their integration into learning environments.
At AERA this week, Amy Collier, Emily Schneider, and I will be presenting a paper that makes a series of arguments regarding learner activities and experiences in MOOCs in relation to clickstream-based MOOC research. One of the implications of our work is the following: learners’ participation and experiences in these courses resist binary and monolithic interpretations as they appear to be mediated by a digital-analog continuum as well as a social-individual continuum. In other words, learning and participation in MOOCs are both distributed and individually-socially negotiated. The following visual (which provides some hints on our results) makes this point clearer:
* and since the work of peer reviewers often goes unrecognized, let it be known, that this insight was prompted by a comment from one anonymous reviewer. So, whoever you are, thank you for your input.
I joined Audrey Watters, Philipp Schmidt, Stephen Downes, and Jeremy Friedberg in Toronto last week, to give a talk at Digital Learning Reimagined, an event hosted and organized by Ryerson University’s Chang School. I presented some of our latest research, and tried to highlight research findings and big ideas in 15 minutes. Below are my slides and a draft of my talk.
Welcome everyone! It’s a pleasure and an honor to be here. Even though I’m the person giving this talk, I’d like to acknowledge my collaborators. A lot of the work that I am going to present is collaborative and it wouldn’t have been possible without such amazing colleagues. These are: Royce Kimmons from the University of Idaho, Amy Collier and Emily Schneider from Stanford University, and Peter Shepherdson from the University of Zurich. The Canada Research Chairs program, the National Science Foundation and Royal Roads University have funded this work.
I want to start my talk by telling a story.
This castle that you see here is one of the most recognizable parts of Royal Roads University (RRU). But, don’t let the castle fool you. RRU was created in 1985. It’s purpose was to serve the needs of a changing society by serving working professionals through graduate digital education and multidisciplinary degrees. It has grown since 1985. It has matured, developed a social learning model that is now infused in all courses, developed new areas of focus, forged global partnerships, and continues to explore how to improve what it does through pedagogical and technological approaches.
Why am I sharing this short story about RRU?
Because this story, minus the specific details, is a common story. It’s also a Ryerson story, a story that is played out at the University of Southern New Hampshire, a story that Open Universities around that world have gone through. It is a story that repeats itself over and over for years and years.
What is the essence of the story?
It is often assumed that universities have been static, unchanging since the dawn of time. The short story I shared illustrates that universities are, and have always been, part of the society that houses them, and as societies change, universities change to reflect those societies. The economic, sociocultural, and technological pressures that universities are facing are sizable, and for better or for worse, usually for both, there’s a continuous re-imagination of education throughout time. Throughout time. Universities have always been changing.
As universities are changing and exploring different ways to offer education, faculty, researchers, and administrators engage in a number of practices that I like to describe as emerging. Emerging practices & emerging technologies are those that are not necessarily new, not yet fully researched, but appear promising.
Online learning and openness are example of emerging practices.
Online learning has a long history. But it also has a new history, with the development of multimedia platforms, media that can be embedded across platforms, syndication technologies that enable learners to use their own platforms for learning and so on. So, even though some of the problems that online learners are facing in contenmporary situations are not new (eg dropout), learners abilities’ to congregate in online communities is expanded through newer technologies and that poses different sorts of challenges and opportunities.
Another emerging practice is openness. Openness refers to liberal policies for the use, re-use, adaptation, and redistribution of content. Openness is also a value: It refers to adopting an ethos of transparency with regards to access to information. And this ethos ranges from academics publishing their work in open formats, to teaching open courses, to creating open textbooks. And it doesn’t stop at individual academics or institutions. In 2014 the Premiers of Alberta, British Columbia, and Saskatchewan signed a Memorandum of Understanding to facilitate creation, sharing, and use of Open Educational Resources. In the same year, SSHRC, NSERC, and CIHR have drafted a tri-agency open access policy to improve access to and dissemination of research results (NSERC, 2014);
There is a growing interest in and exploration of online learning and openness, practices which are still emerging. Next, I will share four recent results from our research into these practices that I believe are interesting to consider because they reveal the tensions that exist when dealing with emerging topics.
First, research into online learning is becoming more interdisciplinary
Interdisciplinary research into online learning means that individuals from a diverse range of disciplines, not just education, are interested in making sense of online learning. It is hoped that more research into online learning and more research from multidisciplinary groups will help us learn more about online learning and about learning in general.
We have evidence to show that research into online learning is becoming more interdisciplinary. I won’t bore you with the statistics, but we measure diversity in published research using a nifty measure and found that the period 2013-2014 can be described as more interdisciplinary than the period 2008-2012.
This is a positive trend, but before I explain its significance, let me explain to you how I view technology.
My perspective on online learning centers around the idea that technology is socially shaped . That means that technology always embeds its developers’ worldviews, beliefs, and assumptions into its design and the activities it supports and encourages.
What does this mean for interdisciplinarity? This means that we have both an opportunity and a challenge.
Our opportunity: to use our respective expertise to improve education.
Our challenge: to actually do interdisciplinary thinking and to go into the study and design of future educational systems with an open mind and the realization that our own personal experiences of education may not be generalizable. A lot of educational technology is produced by people of privilege and to develop educational technology that matters and makes societal difference, we need diversity in thinking and experience.
Our second finding refers to the increasing desire to collect, mine, and analyze data trails to make inferences about human behavior and learning. This practice is often referred to as learning analytics and educational data mining. This practice is a reflection of a larger societal trend toward big data analytics. The idea is that by looking at what people do online one can understand how to improve education.
A couple of things that researchers discovered for example are:
Data trails. Nearly everything that learners do online is tracked. Can we understand learners and improve learning by analyzing their data trails?
While these approaches can help us explain what people do, they often don’t tell us why they do they things they do nor how they actually experience online education.
My colleagues and I are interviewing MOOC students to learn about their experiences in MOOCs.
I am now going to tell you about our third result. We find that learners schedule their learning, use of resources, and participation to fit their daily life. This is in stark contrast to the idea of undergraduate education situated at a university and happening at particular time periods.
One retired individual in Panama that we interviewed works on his class early in the morning every day. Why does he do that? He does that because at that time his daughter is asleep. She is homeschooled and once she wakes up she needs access to the 1 computer that they have in the household to do her own schoolwork. In this case a lack of resources necessitates this scheduling.
One individual that we interviewed moved from the UK to the USA to be with her partner. She is currently waiting for her work permit, driver’s license, and so on, and she was enrolled in multiple MOOCs at the same time. She would work on her courses on Monday because she just “wanted them out of the way,” and so she would work on these courses straight throughout the day.
The fourth and final finding that I have for you today, is that MOOC platforms to date have not offered learners the ability to keep notes, so that particular activity, by virtue of being unsupported by the platform goes undetected when researchers only look at data trails.
Unsurprisingly, learners keep notes. A number of students that we talked to described that they keep notes on paper, frequently keeping a notebook for particular courses and returning to them during exams or during times that they needed them. Learners of course also keep notes in digital format. Usually in word documents, but again documents are dedicated to particular courses, but sometimes they are dedicated to particular topics across courses.
To give you an example, of how we believe this activity could be supported in the future and how we believe innovations can contribute to learning, we recommend designers support this practice by pedagogical innovations such as scaffolding notetaking, but also by technological innovations, by developing online systems for notetaking. What is important here is that such systems should support learning by being interoperable, by allow learners full and unrestricted access to their notes, supporting them to be able to import & export their notes between platforms. Such a design is in line with emerging ideas in the field which call for learners to own their data.
Thank you for being a great audience. I am really excited to hear the speakers that follow me, as I am sure you are!
A visualization of my talk, created by Giulia Forsythe
The Networked Scholars course starts in two weeks, on October 20th, with 2 options for participation.
1. Through the Canvas Network.
2. Through personal blogs and twitter accounts, syndicated, via the…. drumroll…. Networked Scholars Syndication hub. With special thanks to Alan Levine who has been helping a number of people implement this design, all readings and activities will be publicly-available, and this site syndicates blog/twitter feeds used as discussion/reflection spaces. The official Twitter hashtag for the course is #scholar14. If you are interested in this option, feel free to head over to the syndication hub, and connect your blog to the site!
Image: STS-131 Discovery Launch
I will be visiting my colleagues at the University of Edinburgh in mid-June to give a seminar on MOOCs, automation, artificial intelligence and pedagogical agents. This is a free event organized by the Moray House School of Education at the U of Edinburgh and supported by the Digital Cultures and Education research group and DigitalHSS. Please feel free to join us face-to-face or online (Date: 18 June 2014; Time: 1-3pm) by registering here.
This seminar will bring together some of my current and past research. A lot of my work in the past examined learners’ experiences with conversational and (semi)intelligent agents. In that research, we discovered that the experience of interacting with intelligent technologies was engrossing (pdf). Yet, learners often verbally abused the pedagogical agents (pdf). We also discovered that appearance (pdf) may be a significant mediating factor in learning. Importanly, this research indicated that “learners both humanized the agents and expected them to abide by social norms, but also identified the agents as programmed tools, resisting and rejecting their lifelike behaviors.”
A lot of my current work examines experiences with open online courses and online social networks, but what exactly does pedagogical agents and MOOCs have to do with each other? Ideas associated with Artificial Intelligence are present in both the emergence of xMOOCs (EdX, Udacity, and Coursera emanated from AI labs) and certain practices associated with them – e.g., see Balfour (2013) on automated essay scoring. Audrey Watters highlighted these issues in the past. While I haven’t yet seen discussions on the integration of lifelike characters and pedagogical agents in MOOCs, the use of lifelike robots for education and the role of the faculty member in MOOCs are areas of debate and investigation in both the popular press and the scholarly literature. The quest to automate instruction has a long history, and lives within the sociocultural context of particular time periods. For example, the Second World War found US soldiers and cilvilians unprepared for the war effort, and audiovisual devices were extensively used to efficiently train individuals at a massive scale. Nowadays, similar efforts at achieving scale and efficiencies reflect problems, issues, and cultural beliefs of our time.
I’m working on my presentation, but if you have any questions or thoughts to share, I’d love to hear them!
MITx and HarvardX deserve huge congratulations for making data associated with a number of their MOOCs publicly available. Four months ago, I wrote that the “community would benefit from access to the data that HarvardX and MITx have, as other individuals/groups could run additional analyses. Granted, I imagine this might require quite a lot of effort, not least in the development of procedures for data sharing.” It seems that the researchers at MITx and HarvardX have tackled the issues involved to make the data available, and have developed thoughtful procedures to ensure de-identification. While some of the steps taken may limit analyses (e.g., the de-identification process document notes that “rows with 60 or more forum posts were deleted,” thus eliminating highly active users), this is a big step in the right direction and it should be celebrated.
Now… can we have some qualitative data? If any institutions are interested in making those available, I’d love talk to you, give you input, and work with you toward that goal.