Public Lab Research note


Sensor Journalism: A Reflection on the Opportunities and Challenges it Brings

by deirdrem | February 24, 2016 19:03 | 60 views | 0 comments | #12746 | 60 views | 0 comments | #12746 24 Feb 19:03

Read more: publiclab.org/n/12746


The technological advances of the 21st century are changing the types of stories journalists cover and how they find information to report on those stories. In our Data Visualization class, we have studied the history of data journalism, explored the tools available for journalists to organize and present data and discussed the importance of transparency in sharing data. A sector of data journalism that is becoming increasingly popular is “sensor journalism”. As the Tow Center for Digital Journalism report states in their article on Sensor Journalism, “The increasing ubiquity of sensors, their increasing capability and accessibility are on the supply side, while investigative reporters, computer aided reporters and journalist/technologists are on the demand side.” In guest lecturer Lily Bui’s presentation, she defined sensor journalism as “Generating or collecting data from sensors, then using that data to tell a story.”

In class, we practiced sensor journalism firsthand by building a conductivity circuit. We used a kit given to us by Don Blair of Public Lab to test the conductivity of water. The water samples were supposed to be from the following categories: tap water, bottled water, river water and “dirty water” (water from snow or a puddle). When pressing the top of our device to the water, a noise would be emitted; depending on the pitch of the noise was how conductive the water sample was. A high pitch indicated high conductivity, a lower pitch meant lower conductivity. In my groups findings, we found the puddle water to be the most conductive, while tap water was the least.

One of the most interesting things for me was to see how different bottled water samples tested; a Dasani sample sounded much lower than a Pellegrino or Evian sample. The group I was in brought in Smartwater for our bottled water sample and based on the low sound emitted from our sensor, we had some questions about how many of the “smart minerals” are really in their water.

From my own experience, the novelty of being handed a “techie” device as a journalist was thrilling. The opportunity to put technology in the hands of people who might otherwise not have access to it is one of the most optimistic aspects of the future of sensor journalism. Accessible technology being brought to a larger audience yields more positive than negatives and in a journalistic lens, has great potential for stories and honest reporting.

Afterwards, Blair discussed some of the pitfalls his team had with creating a device like this. A simple device like this could be easily reproduced and distributed, but in terms of actual reporting has little capability. Different variables affect the way the noises were emitted, and thus cloud the results. Firstly, the locations of which the class took their water samples were all vastly different. Even being from the same source (for example, the Charles River); we all got our samples from different places within the Charles River. Blair also noted how temperature of the water can affect the pitch of the device. So in order to get honest results, they would need to change the temperature of the water or find a way so the device could change the temperature of the water to make it uniform for all the samples.This is not an easy feat for a water bottle cap. Another obstacle was translating those sounds into a quantifiable format. The pitches differed and that meant something, but sound itself is subjective. Blair had said there was a device that could measure the conductivity in a numerical format, but then eliminates the sound portion of the device. As a class we discussed what the ideal conductivity circuit would have in measuring results: both the noise and the numerical format.

In a Newsgeist 2015 talk, John Keefe of WYNC demonstrates a circuit similar to ours which is used to test conductivity in water. He and Don Blair added to this circuit and created a device (“the riffle”) that stored the conductivity information on an SD card. They were able to successfully test their product in a classroom setting when they took it to West Virginia University. Students were able to build the riffle themselves and test the conductivity of a nearby river. Some of the riffles were even able to send results live by text message. Overall, the experiment was a success and Keefe highlights the fact that journalists can use this and entire communities can come together and participate.But an aspect of the experiment Keefe warns against is starting with a sensor and seeking out a specific type of story in order to use that sensor. The proper order of storytelling for sensor journalism, in his opinion, is starting with a story and finding a sensor to help tell that story.

Which brings me to one of my major concerns regarding the future of sensor journalism: reporters seeking out their stories with an agenda and using technology to support that agenda. As we have discussed many times in class, just because data is quantitative does not mean it is exempt from bias. In fact, often times it is fraught with bias, in terms of what information is included or withheld. Yet bias in data journalism often goes unchallenged because many assume an illusion of objectivity. This could be worsened with sensor journalism, where information is collected from computers and seen as untouchable because it is gathered that way.

I think the attitude should be similar to that of how journalist’s are approaching data journalism. We should be embracing the technology and the capabilities it gives us, but also giving the results the same critique. In the Tow Report, they outlined some strategic recommendations for journalists when approaching sensor journalism. Identifying sensor sources for specific beats, articulating a hypothesis before sensing, working with experts on complex stories and understanding the entire process for your story’s sensor data are some of the most valuable recommendations I took from the report. A quote that has stuck with me from a Kate Crawford video we watched in class is: “Data is always embedded in a context and in understanding that context we can begin to see its limits and its biases.” When using sensor journalism, understanding the context is crucial; whether it’s from an expert in the subject to find out what that information means or simply understanding the sensor by which you’re collecting the information. Lastly, the most important recommendation the report listed was “combine sensing with traditional reporting.” The same rigorous process old school journalism puts on fact checking and ethical reporting should be carried into this new wave of sensor journalism.I think something for newsrooms to remember is that sensor journalism is an investment. It will take a lot of work to report on these stories right, but as the stories listed in the Case Studies show, it is a worthwhile investment.


0 Comments

Login to comment.

Public Lab is open for anyone and will always be free. By signing up you'll join a diverse group of community researchers and tap into a lot of grassroots expertise.

Sign up