Author Archives: Lana Yarosh

About Lana Yarosh

Svetlana “Lana” Yarosh is an Assistant Professor in the Computer Science & Engineering Department at University of Minnesota. Her research in HCI focuses on embodied interaction in social computing systems. Lana is currently most proud of being one of the inaugural recipients of the NSF CRII award, of her best papers at CHI 2013 and CSWC 2014, and of receiving the Fran Allen IBM Fellowship. Lana has two Bachelors of Science from University of Maryland (in Computer Science and Psychology), a Ph.D. in Human-Centered Computing from Georgia Institute of Technology, and two years of industry research experience with AT&T Labs Research.

What’s In a Word

Posted on by
Reply
9409141159_66fbabd750_z

There were a lot of sticky notes involved in figuring out my next research agenda. Luckily, I got great help at DSST!

Recently, I had an amazing opportunity to attend the DSST (Digital Societies and Social Technologies) Summer Institute in College Park, MD. DSST brings together junior (e.g., postdoc, pre-tenure faculty) and established researchers from fields like Computer Science, HCI, Social Informatics, Information Science, Sociology, Anthropology, STS, and more. While there, I participated in workshops, got more familiar with methods used in other fields, and got a lot of great tips on articulating my research agenda (this is frequently a struggle for me because I have such broad research interests).

In small groups, we mapped out the research space of DSST, focusing on how we fit into the picture of a research area that encompasses so many different communities and venues. Interdisciplinarity is not for the weak!

There is so much I could write about this event, but there is one particular issue that resonated in multiple presentations, workshops, and offline discussions — the issue of vocabulary. It seems to me, that the biggest problems in interdisciplinary work arise not when we don’t know the words the other person is using but when we use the same words in different way. Here are a few examples:

  •  Problem: In Computer Science, this term often means “research challenge,” as in “I’m working on the problem of how to connect parents and children who live apart.” However, in some Social Science domains the word “problem” may be reserved for situations that are broken or non-normative. In other words, “connecting parents and children” is not a problem to solve — I might be supporting families in facing the challenge of separation, but I’m not fixing something that is broken. In some ways, using the word “problem” takes agency away from the people we am trying to support, instead positioning the designer as “the fixer.” Usually, this is not what people who build systems actually mean. This is a loaded term and might be better avoided or replaced with “challenge.”
  • Theory: In Computer Science, this term often refers to the study of abstract constructs like Algorithms and Data Structures and frequently developed through mathematical proofs. In the Social Sciences, there are many types of theories serving different functions (my favorite reading on this is Halverston’s “What Does CSCW Need to DO with Theories?”). Theories are frequently heuristics used to understand and discuss empirical data and may or may not need to have strong predictive power. My personal observation is that theorists of both sorts have keen minds and a love for an elegant explanation and can become great friends once the vocabulary issues can be resolved.
  • System: In Computer Science, “system” usually refers to a set of software and hardware that performs a specific function. At DSST, it was more frequently expanded and used to refer not just to the software and hardware but also to the people, practices, and ecosystems around the use of that particular tool.
  • Ethnography: I’ve heard builders use this terms to refer to any sort of formative work that involves observing and talking to users. Not surprisingly, people who are trained as ethnographers might see this as an over-simplification. It’s hard to put the same label on a single week of observation versus two years of being embedded in a particular setting. Perhaps, rather than using the loaded term, it might make more sense to refer to specific methods used such as “two weeks of participatory observation” and “contextualized interviews.”
  • Social Computing: This is a slippery terms, because we are all still trying to figure out what it means. This area is at the intersection of social science and computational systems, but what is included or excluded? As I spoke to people at the workshop, many equated Social Computing with either large-scale social network analysis (e.g., “we looked at 3 mill Tweets”) or Crowdsourcing (e.g., “we leveraged the crowd to do citizen science”). I was happy to provide a third possibility that I believe should be included in the definition: mediated communication for supporting one-to-one and small-group relationships. I hope that we continue to include things like videochat, ShareTable, haptic connectedness devices, small online support groups, etc. in our definitions and investigations of Social Computing.

In industry, most of our work happens in interdisciplinary groups. This kind of reflection has helped me understand why at times I have had trouble explaining my work or outlook to somebody from a different background, like formal methods or software engineering. DSST reminded me the importance of establishing a common ground and a common vocabulary in any work that brings together diverse disciplines.

Availabowls and Other Uses of Bowl-and-Pebble Systems

Posted on by
1

I’ve always been kind of obsessed with interacting with abstract digital ideas using simple physical metaphors. One of my recent projects in this space has recently gotten some press coverage from Fast Company, so I thought that I’d share and offer a bit more detail.

A bowl with physical "pebbles" can be a more intuitive way of managing complex settings.

A bowl with physical “pebbles” can be a more intuitive way of managing complex settings.

The issue is that there are lot of complex settings with many facets that are a pain to manage by using checkboxes in a settings panel. I use privacy as an example in a system I call “Availabowls.” Currently, the granularity allowed for specifying availability is often a bit all-or-none, but there may be a lot of nuances in how you want to manage access to self. For example:

  • I may want to be seen as available to some people/groups and not others
  • I may want to be contactable via some technologies and not others
  • I may want to specify a level of busyness between “available” and “not available” and let the sender decide if they’re issue is important enough to disrupt me

It would be a nightmare to manage each of these settings in a control panel every time I come home. Literally, I have nightmares where I’m swarmed by radio boxes and checkbox panels! Instead, I imagine that a physical token can be used to represent a specific setting. For example, pebbles represent one unit of “busyness,” small blocks represent people or groups of people, plastic tokens represent communication media. Now, to set my privacy settings, I just have to transfer objects between a green bowl (on) and a red bowl (off) when I get home. There’s more about all this in the Fast Company article.

But really, I imagine that a physical pebble-and-bowl system of this sort could be useful for other complex settings. For example, pebbles may represent separate elements of my security system, making it easy to dis/arm everything but also easy to just leave specific elements disarmed (e.g., the back door while I’m having a family BBQ). Or they may represent different eco-friendly subroutines in the house such as ones that turn off lights in un-used rooms, control the house temperature, outdoor sprinklers, etc., again allowing me to only pick the things that I really want to have on right now.

Advantages? I think this is easier to deal with than checkbox panels. It’s easier to have a pebble-and-bowl system “live” where you would most likely be making these changes (e.g., by the front door) than a computer or a tablet. It doesn’t feel like a computer, so it won’t freak out grandma. It’s glanceable or even potentially eyes-free if each pebble has a unique shape.

Disadvantages? There’s only one physical pebble-and-bowl object and I see no elegant way to be able to sync states with this object if you wanted to make changes to your settings remotely or if you wanted to have multiples of these in a big house.

Is this something that appeals to others? Or am I unique in my checkbox-panel-phobia?

Cross-Cultural Parenting Advice

Posted on by
3

At CHI 2013, I attended and helped organize a workshop on Designing for Diverse Families. I’ve been meaning to blog about it for the past two months and recently reconnecting with one of the other organizers has given me the push that I needed. I want to tell you about one discussion that we are hoping to turn into an actual project. I also want to invite both workshop participants and readers who find this to be a compelling idea to potentially join in on this project.

The goal of the workshop was to better understand our own assumptions about what constitutes a “family,” understand the gaps in our own work, and see how we can be more inclusive as a community. The variety of projects presented really highlighted how different families can be in terms of structure, practices, values, and culture. While this diversity can make it very difficult to design for families, it is also an incredible resource. At a time when many parents experience a great deal of anxiety about “parenting right” and face conflicting seemingly authoritative sources on doing it this way or that way, it can be valuable to show that there are many ways to be a parent. Things that are assumed to be true in one culture, can be unheard of in another, but nonetheless children grow up and succeed. My favorite idea that came out this workshop is that getting a perspective from a different culture on specific parenting questions can be of benefit to many families. Basically, we want a parent to be able to ask “how do other families do this?” when faced with everyday worries such as when to introduce particular foods, how to sleep train a child, how to deal with an unruly teenager, etc. Presenting these answers in the context of their originating geography can help bring home the idea that there is no single right way of dealing with most of these questions, but rather many approaches that can be successful.

So, a few of us would actually like to make this happen! Here’s the basic sketch of a plan:

  • We populate the initial list of questions and answers using Yahoo answers
  • We continue getting answers to common questions by selectively using Mechanical Turk, tracking the geographic locations of the Turkers
  • We allow users to ask questions of the community and see a selection of answers from different parts of the world displayed on a map
  • Users are encouraged to give their own answer to another question when they see answers to their own

Obviously, it’s fairly rough at this stage, which is why I’m reaching out to y’all. So for the parents out there, would this be useful for you? For the builders out there, would you be interested in getting involved in making this happen? For the researchers out there, would you be interested in getting on board or just giving advice? If you are interested in being a co-investigator, please fill out this survey to let us know what you’d like to do and we can start getting a plan together!

Continue reading →