Privacy and HCI for Personal Genomics

While considering HCI perspective on personal genomics, we partnered with the Personal Genome Project. The nature of the Personal Genome Project (PGP) as an open source online database of personal genomic information raises important questions regarding a participant’s privacy and willingness to share their information publicly. The PGPHCI team, led by Orit Shaer, in collaboration with Dr. Oded Nov (NYU) and with Dr. Darakhshan Mir (Wellesley College) investigates privacy and sharing in the context of personal genomics.

Screen Shot 2014-08-19 at 4.34.00 PM

While conducting an intensive literature review of the field, student researcher Claire Cerda, led the team to discuss how people have unique attitudes and behaviors when it comes to maintaining their privacy and security. Such behaviors include clearing cookies from a browser before logging off of the computer, or covering the keypad when entering a pin number connected to a debit card. Some people are very concerned about having their credit card information stolen when they pay for products online, but some are more trusting in the system. These attitudes and behaviors may vary based on a person’s technical skill or a person’s age, for example. In order to better understand user’s attitudes and behaviors regarding privacy, the team implemented a privacy index developed by psychology Professor Tom Buchanan of the University of Wesminster. Buchanan’s privacy index builds upon the work of the well-respected scholar Alan Westin, who was one of the first to study privacy and develop a way of measuring people’s feelings and behavioral patterns. Buchanan included more technologically relevant questions about the internet and online personal security that was not fully available during Westin’s time. The Buchanan index is made up of three separate scales: a privacy concern scale, a technical protection scale, and a general caution scale. The privacy concern scale measures a person’s attitude and the technical practice and general caution scales measure a person’s behavior about privacy and security. The PGPHCI team tested the scale among six individuals. Check out the results of our pilot study below. The graph presents a score between 0-4 on three privacy dimensions for each of our 6 pilot participants.

Screen Shot 2014-08-19 at 4.35.56 PM 

From the pilot test, the team was able to assess the effectiveness of the index and also understand its capabilities in an online survey.

The team also explored the risks and benefits of sharing personal genomic information among different circles of people. For example, the risks and benefits of sharing personal genomic information with family, friends, scientists, or on social media. The team will test whether making users aware of the risks and benefits of sharing will positively influence the amount of data they are willing to share. The study will continue in the next upcoming months, and the PGPHCI team looks forward to presenting its work at future conferences. Stay tuned for more information…

Summer Program Ending but zSpace Team Still Going On

Basic functionality has been implemented!

Basic functionality has been implemented!

With Wellesley’s Science Center Summer Research Program coming to a close, the zSpace team members are taking a moment to asses the current status of their two projects as well as reflect on the trials and tribulations encountered over the summer.

Following a brief stint of time where the fNIRS machine was not working due to a timer issue, Cassie and Jasmine have resumed running user studies. The issue that caused a pause in the this project is one both team members had taken note of before and has been a recurring problem throughout the summer. Unfortunately the timer is an integral part of the fNIRS and while an attempt was made to fix it early in the course of our research, the fix has proven just as unreliable as the machine. During the course of the summer the fNIRS developed a history of not working on Mondays, erratic signal strength due to poor headband (what holds the sensors to a user’s forehead) construction, and overvoltage signals for no apparent reason. Getting the finicky fNIRS to work in conjunction with zSpace proved to be a trial in patience this summer. While sometimes finicky as well, the zSpace proved to be more reliable than the fNIRS though we still ran into a few problems, the biggest of which were unexpected program shutdowns in the middle of user studies. Despite the setbacks and issues with both systems, the zSpace team is still on track to submit a paper to CHI 2015. In the mean time, we’ve created a poster to present some of the current findings of the brain study in the summer research program’s poster session this week.

A lot of progress has been made with MoClo Web Planner since the zSpace team’s last post as is evident by the image included in this blog post. The web version of the surface application has nearly full basic functionality implemented. A new custom theme has been implemented, draggable panels allow users to switch between levels, all parts in Level 0  (stored in the different tabs) can be dragged into, and only into, the correct boxes in Level 1, Level 1 has a combine feature that puts the parts together in different permutations, and any combined part sequence can be dragged into Level 2. Or in salad terms as Cassie is the one that has been figuring out the functionality: In Level 0 a user can select the parts of the salad that they have on hand and place the different food items in the correct boxes in Level 1. The button in Level 1 shows the user all of the possible salads that they can make, and then any number of those salads that seem appealing in Level 1 can be dragged into Level 2. The issues we’ve run into while creating MoClo Web Planner largely stem from learning jQuery on the fly and odd properties of various constricts we’ve implemented for functionality purposes (most notable of these is getting z-index to work in a way that would allow users to drag parts from one panel to another and have parts being visible above the new panel while in the process of dragging). Today the team met with our Boston University collaborators and discussed where we should now focus our energies as well as what functionality of the original MoClo Planner is most important to them.

Even though the program draws to a close, the zSpace team isn’t done. Both Cassie and Jasmine will be continuing to work on the brain study and MoClo Web Planner through August and likely into the school year.

Wrapping Up

Itglass’s the last week of summer research, and the Google Glass Team is placing finishing touches to their project. Grace and Lily thanked their collaborators at Boston University for completing the week long user studies with Glass. Back at Wellesley, the pair is excited to have finished conducting eight user studies with biology summer researchers.

With all three Google Glasses safely returned home to the HCI Lab, Grace and Lily are polishing up their protocol application. Grace has just incorporated voice commands, so when users nod their heads down, a microphone will be activated and respond to the commands, “next,” “previous,” and “cancel.” Meanwhile, Lily has also successfully added a bookmark feature, which enables users to mark their place in the protocol. With an upwards swipe, the background color changes from black to cyan, offering a visual cue of the step the user last completed.

The team has also created a lovely poster for the summer program’s poster session, and they are eager to present their findings at the end of the week. 

“Next” Steps for Google Glass

The Google Glass team has solidified an ambitious To Do list for the last two weeks. As user studies continue during the daytime, Grace and Lily are researching different areas for their app development and beginning a research paper.

nextprev

When the Wellesley biologists are not engaging with the Google Glass, Grace and Lily are working to add features to the app. Grace is still focusing on embedding voice commands. By the end of the two weeks, she hopes to achieve a simple command of “next” and “previous” to navigate through the protocols. Meanwhile, Lily plans to add bookmark feature that will enable users to mark their current progress in the experiment. Afterwards, she wants to start designing an accompanying program that allows users to upload their protocols directly into the app.

At Boston University, the third user has just started his week-long user study and will continue to send us feedback. By the end of this week, the Google Glass team will have finished all user studies with the BU researchers.

 

Web Application of MoClo Planner

MoClo Planner Web
During the zSpace team’s previous endeavor to finish 60 use r studies in 10 days, the fNIRS device failed so the team is now pursuing a different project. While waiting for progress on fixing the machine, we’ve started web development with jQuery on MoClo Planner, a modular cloning planner used by synthetic biologists to plan their experiments.  MoClo Planner is a project originally done in 2012 by past summer researchers in Wellesley’s HCI lab. During its creation, the goal of MoClo Planner was to provide a touch compatible interface that allowed for easier information gathering and construct designing in synthetic biology labs. It contains a biological part library that can be filtered and searched through, access to a data sheet and relevant publications from Pubmed for each part, and a layered workspace (the featured  image displays a prototype of the layered workspaces in the web version of MoClo Planner).

Neither Jasmine nor Cassie have much experience using jQuery so there is a learning curve to their development process but they chose to remake the application with jQueryUI, a library that has features like drag/drop and resizability. Their goal is to make the application as similar to the version of MoClo developed for the Microsoft Surface as possible. The web version was highly requested by our collaborators at BU and International Genetically Engineered Machine (iGem) Foundation.

So far, we have made resizable panels that function in the same way as the ones in the surface application. Our short term goals are to make drag and dropping between panels possible. During our process, we are trying to bring the high school interns from the Wellesley-Framingham Outreach Summer Program, Shristi and Savanah, on board with our project to give them a wider base of experience in HCI.

When Life Gives You Google Glasses, Conduct More User Studies

photo 1
The Google Glass team has just begun conducting another wave of user studies. The arrival of a third Google Glass (grey) enabled Grace and Lily to kick off a new study with Wellesley’s own synthetic biology students. They ensured that each Wellesley student also works in a wet lab environment and uses protocols regularly. During a lunch meeting they held, Grace and Lily invited 11 Wellesley biology students to play around with their protocol application and to munch on Boloco burritos. As the Google Glass team explained its purpose and goals for the study, the biology researchers chimed in with their feedback and initial suggestions for further enhancements.

From this collaborative discussion, Grace and Lily noticed that some of the most common suggestions included features to bookmark current steps and to input unknown values through voice commands. The team also learned that lab goggles are potentially problematic, as Google Glass does not fit well over or under them. Companies like XOne seem to be addressing such issues by designing safety goggles with Glass-like features. Fortunately, only one of the Wellesley users is required to wear goggles, and Lily and Grace plan to find her a substitute for the study.

Jusphoto 2t this morning, Grace and Lily handed Google Glasses over to two Wellesley students for a three-day study.  The biology researchers quickly grew accustomed to their customized protocol apps and immediately started using them. In the meantime, the first week-long user from Boston University is turning over the Google Glass to a second week-long user.  

Excitement fills the air, and data is flowing up their stream.

Stereoscopic Vision and zSpace

zspace2

In the process of exploring new interactions, Wellesley’s HCI lab has a team of students, Cassie Hoef and Jasmine Davis, working with zSpace. The zSpace is a holographic rendering machine capable of displaying objects in 3D using stereoscopic vision. It features: head tracking glasses, a stylus that offers 6 degrees of freedom, and haptic feedback.

In collaboration with Erin Solovey (Drexel), we use fNIRS brain sensing as a supplemental measurement to more traditional performance metrics to investigate the benefits of such techniques for spatial problem solving.

fNIRS is a non-invasive device which sits on the user’s forehead (as shown in the photos above). It shines near-infrared light through the skull and into the brain to measure the amount of oxygenated hemoglobin. The measurement can indicate the amount of workload a user is experiencing which might be a useful indication of how helpful the modes of interacting with the zSpace are.

In order to test this we are using a series of conditions and activities in the interactive stereoscopic 3D environment provided by zSpace to determine if a fNIRS device is a viable option in the collection of supplemental data in an HCI setting. Additional features of zSpace like haptic feedback and stereo vision are toggled between users to assess the fNIRS data’s ability to differentiate between different conditions. During the study the user completes puzzles in our zPuzzle application while wearing the fNIRS device. The levels change in difficulty, so the workload on the brain should change and ideally the change will be detectable between levels.

Some preliminary assessment of the data suggests that there is the possibility of a difference in frustration levels, physical workload, and subjective effort.  The brain data shows a higher level of oxy-hemoglobin during the stereo-haptic condition when compared to the stereo-no-haptic condition. We hope further studies will support and strengthen these initial suggestions.

These days, the zSpace team primarily spends their time scheduling and executing user studies in order to collect more data. In the next few weeks, we will be executing 60 more user studies and beginning data analysis.