Guide to Building Data-Driven Organizations in the Public Sector

Collecting Group Data

Collecting Group Data (Team 5)

Topic Overview

In Chapter 3 of Reality Mining and Chapter 5 of Social Physics we see examples of how we can learn from small groups. In Social Physics, Pentland worked with employers and their employees to conduct short-term experiments in order to improve the productivity of the workplace. He and his team were able to draw out key learnings from the data collected and discovered important links between behavior and social learning opportunities. Their data showed that the pattern of ideas flowing is central to driving productivity. To take this a step further, their research showed that face to face engagement and exploration were the largest factors influencing productivity and creative output. (Pentland, 2014) In other words, the more time people spending time interacting with each other, the more ideas could be generated, discussed, and their colleagues would be in the loop, thus resulting in higher efficiency and productivity.

In Reality Mining, Eagle and Greene discuss the need to expand collecting data from small groups outside the world of academia and opportunities to scale it to address wider community needs using sensors and technology. They acknowledge that small group data collection is often the most difficult type of information to gather because participants are often wary of sharing information that is trackable back to the individual. In small group settings, there is heightened concern about anonymity. The authors’ central belief is that incentives should be used more by those wanting to collect group data and that given the right incentive, data could be shared freely. (Eagle and Greene, 2014)

DISCUSSION

There is no doubt about the benefit that can come from capturing small group data as these two texts illustrate. However, how realistic is it that nonprofits or government agencies can collect and use such data? Despite the relative ease of collecting the data now, compared to a decade ago, organizations must have the ability to capture the right data that is most meaningful to their work and the expertise to use it to affect the delivery of their programs or services. Further, organizations must have a strategy to use the data and processes to interact and learn from the data. (Barton and Court, 2012) (Duhigg, 2016) Using big data can feel overwhelming for these reasons. Additionally, agencies must commit the resources needed to extract and interpret the date and promote the validity of the investment. Far too often this type of research is considered nonessential and bypassed due to funding or staffing shortages, or simply because we are all too busy. If data and statistics are to be the foundation for decision making, then it must be treated as an essential tool, not a secondary accessory.

Perhaps a better and more manageable role for nonprofits and government agencies is collecting targeted data on a few key indicators and performing micro tests of focused hypotheses. This would allow them to more easily track, manage and use the data to answer a key research question. For example, an advocacy organization could test what subject lines drive click-throughs to increase the number of advocates taking action via email. A government agency like the Motor Vehicle Division could test alternative strategies to find out what best reduces customer wait times. Some ideas might include creating an express lane, having staff specialize in a particular function and having them focus on those tasks, having staff keep track of frequently asked questions and communicating the answers, thus cutting down the amount of time needed when being helped. It seems like the idea of using big data is good in theory when you have the ability and technology to use it well. For both nonprofit and government settings, using the data in a more strategic, targeted approach could be more manageable and practical.

Chapter Summaries

Reality Mining: Using Big Data to Engineer a Better World Chapter 3: Gathering Data from Small Heterogeneous Groups

This chapter explores the potential of collecting data from small groups of people. It discusses the challenges, costs, and benefits of a number of different types of technologies that collect data from small groups. Detailed examples from multiple fields are given, including a look at smart conference badges that track movement and engagement of conference attendees, companies mining data from their employee activity to create a more productive work environment, neighborhoods collecting data to help solve social problems such as reducing pollution and determining the best bike paths, and audio surveillance systems that enable media companies to have a real-time look at how people consume multiple types of media. The authors cite concerns throughout the chapter about maintaining user privacy, legality of data collection, the expense of collecting data, and having the technology available to gather the data efficiently. Despite these concerns, the authors believe that the biggest challenge of group data collection is finding the right incentives. They argue that if the benefits of the data collection outweigh the costs, individuals would be more willing to share their data and companies who want that data should compensate people appropriately.

Social Physics: How Social Networks Can Make Us Smarter Chapter 5: Collective Intelligence-How Patterns of Interaction translate into Collective Intelligence

In Chapter 5 of Social Physics, researchers detail their findings regarding networks and exploration. Their findings indicate that the pattern of idea flow was the single most indicative element of a successful group. When groups featured characteristics such as equal participation, turn taking, and high engagement with other members, the group was likely to have high performance. Researchers summarized that when these group conditions exist, it will likely result in a large volume of ideas being generated and filtered in a group process with a majority consensus being gained. When a network functions in this capacity, research indicates they have a higher efficiency and yield a more creative output. These groups had a higher level of collective intelligence than group that did not function in this format.

In order to develop the volume of ideas, the members of the group must actively engage in ‘exploration’ whereby they venture outside of the network to gather unique data and return to the network to add it to the collective stockpile of information. Once this “idea dump” has been completed, the network reconvenes as a unit to filter and evaluate it. Once completed, any ideas or information that are deemed valuable or actionable by the majority are put into practice.

Key Take-Aways (for Yellowdig)

Collecting Group Data

Discussion Questions

In one case, a company subsidized the cell phone equipment and service for people participating in a study that tracked their media consumption. Their phones were enabled to listen to everything that user experienced and took short snippets of sound and matched those against a database to see what movies, songs, TV shows, or other media a person may be listening or watching. For you, if a company offered to pay your cell phone bill, would you give them an all-access pass to listen to your life? Describe your answer.

Follow Up: This was published before the mass introduction of Amazon’s popular “Alexa” gadget. Do the recent incidents around Alexa’s unauthorized recordings and subsequent invasion of privacy influence your answer? In regards to the previously mentioned cell phone research, do assurance that conversations and other non-research related recordings are to be deleted and discarded appease you?

In another case, the authors talked about companies monitoring their employee’s activities, including phone calls, activity on their computers, emails, recording of phone calls, movement around the office or text messages. Companies do this to better understand the social networks of their employees, connect staff with similar needs or interests, detect potential fraud, track productivity, and other things. Would you be comfortable working for an employer who is watching your every move? What benefits or incentives would your employer have to provide to make you comfortable with this?

In one case study, the author noted the productivity of a team increased when the members of the team were given a collective break, as opposed to staggered break times, allowing them to have more face-to-face interactions. The conclusion was that more engagement translated to better efficiency as a team. Do you find this to be true in your own work setting? Is there a point of too much “collectiveness”?

Does the research showing high functioning networks are a result of equality, turn taking, and engagement surprise you? Is this something we all learned in kindergarten but have to relearn in adulthood?

References