The Cost of Perfect Dataveillance

For most of the semester, I have typically argued that society’s constantly improving ability to monitor and record data about our daily actions would be beneficial, provided that the information would be more or less equally accessible to everyone.  For example, if car insurance companies had perfect information available to them to evaluate how risky drivers are to insure, they could naturally develop a socially optimal pricing structure (in addition to creating incentives for people to be safer drivers).  However, what I read from pages 101-205 in the Circle began to make me question that assumption.

If we view the Circle as a closed network in which everyone who works for the Circle has more or less perfect access to their electronic information about one another, we begin to see some of the potential problems that could arise in this society with perfect access to information.  One of the biggest problems that arises in this scenario is the fact that the can be large transaction costs associated with collecting the information.  If the opportunity cost of Mae filling out various social media profiles (the revenue she could have generated for the company if she had been working in Customer Experience for the time it took her to fill out the profile) exceeds the “Community Value” that The Circle gets from having this additional information, then obtaining this information was not efficient for the company.

In addition to this, The Circle also creates an adverse incentive for people to use social media to inefficient levels.  The Circle may value the data it obtains very highly, but if the data was created for the sake of creating data (for example, people trying to climb up the popularity rankings), there is no guarantee that the data is accurate and/or useful.  For example, when Mae is doing her best to race to the top of the popularity rankings, the book mentions that she commented 33 times on a single page.  From the context given, it seems fairly safe to assume that her comments were made solely for boosting her ranking, and not because they reflected her actual opinions.  Therefore, returning to the issues of transaction costs that perfect dataveillance could potentially incur, it seems as though The Circle would have been better off if Mae had been working in Customer Experience than commenting on that random page.  The value of these comments ties back in to our previous discussions about what types of communication are considered “good” communication.

One of the best examples of this in The Circle is the concept of “Zinging” someone.  When someone receives 10 “Great Job” Zings for completing a relatively menial task, it’s ultimately a waste of time from the people who sent the Zings.  This wasted time is ultimately the largest cost that was created when The Circle implemented the popularity rankings.

The common theme that ties the past three paragraphs together is that perfect dataveillance is inefficient, because the additional benefit of having access to 100% of information instead of 99.9% is relatively small, but the costs of obtaining that extra 0.1% is relatively high.  While this does not entirely refute the position that I initially believed, that perfect dataveillance is not detrimental to society provided that is it not applied asymmetrically, it definitely reinforces the fact that we should not attempt to achieve perfect dataveillance purely for the sake of perfect dataveillance.

Average Response Time in Class

In class on Thursday, I recorded the duration of every response to each question asked in class.  I did this, because I thought it would be interesting to see if the response time increased as we got more in depth with a particular subject, and then when the subject changed, the response time would drop, and then slowly build back up again. The data indicates that this theory is not correct, because the response times do not appear to have any clearly visible trend. This is likely because as we go more in depth with a topic, the questions do not necessarily require an longer answer. In fact, some of the introductions to new topics might have required the longest answers, because it was necessary to describe a large amount of information to the class.

Summary Statistics (Time Spent Talking in Class)

Proc Means

This indicates that approximately 10 and 1/3 minutes of our class time was spent by students responding to questions.

Plot of Response Times

Response Time

Plot of Difference in Response Times


Plot of Whether there was an Increase or Decrease in Length of Response (1=Increase)

Increase or Decrease



In Case You Ever Wondered Why Cartographers Don’t Go To Davidson

Mark Monmonier’s article “How to Lie with Maps” draws a lot of parallels with the Lev Manovich article that we read in class last week, highlighting the fact that data in and of itself is nowhere near as important as the human ability to analyze and interpret data. Maps are unable to perfectly represent the region that they are showing, because the cartographer is forced to superimpose a 3-d world onto a 2-d sheet of paper. This goes hand in hand with some of the problems that we have when we obtain data, because it may have been filtered in a certain way that we are not aware of, which could completely change the results of the analysis.

Topographic Map

The example that Monmonier uses with maps is that map users usually trust the benevolence of the mapmakers , which may or not be misplaced. Because cartographers are not licensed, just about anyone can make a map, which means that they might not always be accurate. If someone is aware of this, they might look at the map with a more critical perspective, compare it to the world around them, and then note some of the mistakes that may be present in the map. However, if an individual is unaware of this, they may assume the map is correct, and any problems they experience while using the map must be due to their own personal mistakes. The individual who views the map with a more critical eye will eventually be the one who walks away with a better understanding of how the map he is holding relates to his surroundings. Likewise, in our data analysis, we should be sure to investigate all of the assumptions that the data set uses, as well as other relevant background information in order to make our analysis of the data set be thorough, which will eventually allow us to make the most accurate analysis possible.

Image URL: Continue reading In Case You Ever Wondered Why Cartographers Don’t Go To Davidson

Distribution of Mac and Windows Users in DIG 210

DIG 210- Data Collection: Distribution of Technology
5 Windows Computers
7 Mac Computers
1 Windows Tablet
2 Mac Tablets

(Statistics Obtained Prior to Use of Lab Computers)
5 Windows Computers
1 Windows Tablet
2 Mac Tablets
12 Mac Computers

In addition to this, as we learned in class, we have 18 iPhone users (iPhone 5 or later), 8 Android users, and 4 people who use a different type of mobile device.