DIG 210 – Responders – Group 5 – Week 5

In class on Tuesday, we examined the importance of raw data in our day and age. A few students expressed that they regarded data as rather useless in making accurate predictions about a user’s psychological profile or behavior, whereas others were convinced that data has highly predictive potential itself, but its usefulness depends almost entirely on its interpretation in an appropriate context. In other words, data about the hair color of an individual student at Davidson College has no standalone relevance, but somebody intends to dye his or her hair in the most popular color in order to fit in more seamless way, such specific data becomes important in this particular situation.  In other instances, the data provided was not recent or relevant, and seemed to skew the interpretation.  While these insights are useful, they are only as accurate as the data provided to them.

Aside from this rather silly and purely illustrative example, data enables us humans to answer the most complex questions about our surroundings if and when we ask the contextually appropriate questions. Reciprocally, the underlying power of data analytics and statistics empowers those in possession of data. Our class discussion focused on this aspect for the majority of the time due to its contemporary relevance considering the recent Equifax data breach and eye-opening NSA scandal in 2013. It appeared to be rather easy for most students to point out the looming threats ranging from mass surveillance to data-adjusted insurance premiums. However, we are curious why the perceived dangers of big data analytics seem to far outweigh the perceived possible benefits. Hence, the question remains whether our public perception of data has been skewed by Western pop culture (e.g., “1984” by George Orwell, “The Circle” by Dave Eggers and countless futuristic Sci-Fi movies) or if it really poses a threat to our societal and democratic well-being. It is a question that requires immediate attention because only effective data privacy policies will ultimately protect us when everything has been converted into ones and zeros.”

Data Culture- Responders, group 1

I like the analogy that Group 4 used in their post to explain cookies.  I have seen the word “cookies” and have been given the option to disable or enable cookies on my computer before but never knew what they were.  What stuck out to me even more in the Do Not Track mini series was the insane number of third parties that monitor your viewing history.  There were 63 third-party trackers for the Wall Street Journal!

 

In addition, I agree with Group 4’s analysis of the quote from Dataveillance and Counterveillance.  It may seem like cookies are helpful and allow you to shop and Google search more easily; however, I would feel a lot more comfortable without these cookies and with more privacy.  A quote that I found interesting from this reading is when Google said that they were “recognizing your browser, not you.”  But when Google is using 57 different signals to personalize each person’s search, how different are you from your browser?

 

Group 4’s final question posed, “if it is just for the government to arrest an undocumented immigrant based on social media insights, especially conversations that cookies have recorded” is a very interesting and thought-provoking one. On the one hand, undocumented immigrants have committed a crime, so it is reasonable for the government to use any information available to them to pursue an arrest. However, if the government does decide to use such information, one must seriously question the boundaries to the government’s access to data.

DIG 210 – Week 4 – Group 5 – Observers

Patrick – Computers brought to DIG 210 class on 9/14

I was curious to see the distribution of computers that students were using during class.  Personally, I am a Mac user, who has developed a preference for operating in Windows while programming or using Microsoft Excel.  Though I have no basis of comparison, I was curious to see if students in a class focused on Data had a stronger preference towards non Mac computers.  

My results were found simply by counting the number of computers of each brand that students had during class on 9/14.  14 of the 20 students attending class brought Mac computers with them, while there was one student with each of the following brands: Asus, Dell, Toshiba.  The remaining 3 students did not bring computers to class.  70% of the students in class had Mac computers, which is an overwhelming majority.  Even in a data focused class, Mac appears to be the most popular choice.

Seth – What can the clothes we wear tell us about the weather outside?

I recorded what each student in class on Thursday (9/14) was wearing. Either shorts, pants or a dress and either a short sleeve shirt or a long sleeve shirt/jacket.

This past week I decided to observe how people decided to dress as the weather begins to turn to lower fall temperatures. On (9/14) during our class it was approximately 70° F and one of the first cooler days of the season. I was curious if people would start to wear warmer clothes and what percentage of people would wear pants and jackets vs. shorts and t-shirts. Furthermore, in this period of time when the weather cools off from the much hotter months I was curious of the probability if someone wore shorts, would they wear a short sleeve shirt or a long sleeve shirt etc.

I found that 65% of the class wore shorts, 25% wore pants, and 10% wore a dress. I also discovered that 50% of the class wore both short sleeves and shorts, while 20% wore long sleeves and pants. These were the two highest combinations, however, some people also mixed what they were wearing (shorts and long sleeve, pants and short sleeve etc.) Also I found that if someone wore shorts, the probability they would wear a short sleeve shirt, was 77%. And if they wore pants there was a 20% chance that they would wear short sleeves. The findings from this experiment coincide with my hypothesis that warmer clothes will appear with the cooler outside temperature. It was also interesting to see that people either wore both pants and long sleeves or shorts and short sleeves, which could definitely coincide with fashion and other factors.

I think it would also be interesting if I made this observation every class period and graphically showed the change in clothes worn to class as well as the outdoor temperature side by side. This would obviously take more than one week but nonetheless could be an interesting representation of average temperature over time and its effect on the clothes we wear.

Charlotte – DIG 210 use of filler words in class discussion, by gender

In my study, I focused on the use of “filler words”, such as ‘like’, ‘um’, ‘uh’, ‘er’, ‘okay’, ‘right’, ‘you know’, and so on, during Tuesday’s class discussion of the week’s reading. I was interested in observing the use of these words as we never notice our own, personal use of them, but we do tend to become aware of their repetition in our language once somebody tells us to look out for them.

In terms of methodology, I split my observations up into male and female groups. I was interested in seeing if one gender used more filler words than the other. From my observations, women used 3.14 filler words per comment in class discussion and men used 3.26 filler words.  However, our DIG 210 class is not split equally between genders. Tuesday’s class consisted of 14 men (n = 14) and five women (n = 5). Therefore, it would be unwise to find a conclusion is my data as the subject, or sample size (n), groups were not the same and a fair comparison cannot be drawn.

I think this topic would be very interesting to extrapolate upon in multiple classes across Davidson departments. In particular, I wonder if filler words are more common in specific departments and whether men or women speak more frequently in class discussions at Davidson.

Daniel – DIG 210 Student Hair Color

In our study regarding the DIG 210 students’ hair colors, we found that 64% of students had either brown (n = 6) or blonde hair (n = 6), while students with red hair represented a clear minority with only 10% of the class. As a result, our class differs from other sample groups such as the one that was recently the research subject at Rice University (2016), which showcased an overwhelming majority of 48% (n = 286) as brown-haired while only 21% (n= 127) had blonde hair.

How can it be the case that these two data sets of US student classrooms allow for completely different inferences? On a qualitative level, a variety of reasons ranging from geography to social demographics could be responsible for the differences in probabilities. However, the real statistical illegitimacy that is embedded in these data sets originates from the limited number of observations that prevents the Law of Large Numbers to eliminate selection bias to a sufficient extent. With so few observations, the inferences that can be made about the hair color demographics of Davidson College’s general student body will be inherently flawed.

 

Data Culture — Responders Group 4

In the Do Not Track mini-documentary series, Brett Gaylor, explains how cookies record our online behavior and every time we log into a site, click on something, or browse, the cookies will ultimately be able to build more complex insights into who we are, what we like, and what we may want to see or do next.  If you are having a hard time understanding cookies, think about it as the black parasite in Spider Man 3, that latches on to Peter Parker and fashions a black Spiderman suit.  The more Peter wears the black suit, the more the black parasite can adapt and understand its host, and thus ultimately exploit Peter.  

In Dataveillance and Countervailance, Rita Raley explains, “there are basic steps one can take to delete cookies, but it seems unnecessary to do so because they do not interfere with everyday computer use; in fact, some of them are functionally necessary and the end result is that one encounters advertisements that may be of interest.” So despite cookies’ ability to spy and record, many do not care and are not paranoid. Although our group has come to the consensus that this offers conveniences when it comes to online shopping and finding the best discount packages, there are some major concerns to this level of surveillance sophistication.   

In the same reading, Raley writes, “For every system of disciplinary power, as Anthony Giddens puts it, there is a “countervailing” response from those in precarious, subordinate, or marginal positions, which is to say that dataveillance and countervailance must be seen as inextricably connected” (131).  As of right now, the Trump administration is placing a significant emphasis on deporting undocumented immigrants, and emotions have been heightened with the repeal of Obama’s DACA act.  Ethics and opinions aside, one question we would like to bring up is if it is just for the government to arrest an undocumented immigrant based on social media insights, especially conversations that cookies have recorded.  

Data Culture- Observers group 1

Using survey monkey, our group asked a series of questions about music preferences.  These questions included- 1) What type of music do you listen to?  2) How many songs do you think you listen to in one day?  3) Do you play any instruments?  and 4) Which medium do you usually use to listen to songs?  We found that the most popular genre of music in the class was hip-hop- 53.33% of the class responded that they listen to hip-hop.  In addition, the most popular response to question two, how many songs do you listen to in one day, was 0-10.  40% of students responded this way.  80% of our classmates do not play an instrument, and lastly, the most popular medium for listening to songs was Spotify, with 60% of the class using Spotify to listen to music.

 

Data Culture Group 4 — Surveillance: the Double-Edged Sword

Group three wrote about how Finn argues, “that the progress in forensic science over the years has helped law enforcement identification, but emphasizes that the mold of identification could manipulate how the country views law enforcement at the local, state, and national levels.”  Data surveillance innovations and advancements present a double-edged sword.  With improvements come more accurate targeted insights into your sample size to deduce trends rather than having to rely on speculation.  One such example is reverting back to the notorious Jack the Ripper case where technology and forensic science was virtually obsolete.  There was very little to go off on because surveillance technology was basically in the dark ages: little to no DNA matching, security footage, or other data that could pair down the suspects.  To bring it back into the 21^st century and into Davidson, NC, every time we swipe our catcards into a building or dorm a time log is created with our name and ID number.  If a crime in or around the building you swiped into occurs the police would reach out to investigate.  In this way, data surveillance technology advancement offer positive benefits to society as resources for law enforcement to use to solve crime. 

On the other hand, capturing and analyzing large data sets to find trends could lead to precarious territory such as the ethical dilemma the US is facing now with racial profiling.  It seems that the trend is with more data, stereotypes have a greater chance of succumbing to being upheld with “substantiated evidence,” even though the data could be subvert to subjectivity from a myriad of selective practices.  We agree with group 3’s assertion that, “these advances need to be checked by academics, professionals, and common citizens to inspire trust in law enforcement officials and new technology.”

Data Culture — Observers Group 4

Using Survey Monkey, our group surveyed the class asking for their honest answers pertaining to what website they spend the most time on, and how much time they spend on that site per day.  Facebook was the most popular website for our class by 26% with the average time being 1.2 hours.  We then compared our class’ preferences with the same questions Alexa.com gathered for their statistical breakdown of the 65 million active Amazon Prime users.  Alexa.com reports that Google is the most used website, at 3,560,046 linked sites and people spend an average of 7:50 minutes on Google, but the length of observation (ie measured per visit, per day, per month, etc) remains omitted.  On Alexa.com, Facebook came in at number three with 7,600,185 linked sites and an average of 10:09 minutes spent on the site, again we are unsure about the specifics in regards to time measurement.

From our class observation, we can deduce that social media occupies our peers’ website traffic, followed by Youtube (1:40 hours) and then Netflix (1:20 hours).  We are skeptical about these results because students hypothetically should be doing research or completing homework which usually requires Googling something.  We have concluded that people often disregard Google as a site itself and think of it as just a passageway to other content.  If content is being reached through Google, at what point is Google excluded from the usage equation? Furthermore, college students are 18-22 years old on average, representing a cross section of the population.  Complicit with other research, it makes sense that social media and entertainment websites are occupying our time the most.  Interestingly, Facebook is still very popular among Davidson students even though it was ranked third on Alexa.com and has been decreasing in popularity for some time, much to Zuckerberg’s chagrin.

DIG 210 Group 3 Readers

      The central theme in this week’s reading is the use of data to distill the human condition to a few data points. In Capturing the Criminal Image: From Mug Shot to Surveillance Society, we are presented with the history of collecting crime statistics. Here, Finn argues that the progress in forensic science over the years has helped law enforcement identification, but emphasizes that the mold of identification could manipulate how the country views law enforcement at the local, state, and national levels. Though Finn does not believe that we have reached the point of no return. He concludes that the advances in crime-based data representation are a necessary evolution from the limited criminal investigation tactics utilized in the 19th and 20th centuries. At the same time, Finn stresses that these advances need to be checked by academics, professionals, and common citizens to inspire trust in law enforcement officials and new technology. If this data collection continues unchecked, a sense of mistrust might develop between the people and law enforcement agencies.

 

   In the Lyon reading, we see how societal anxieties against the super-panopticon, the ability to asymmetrically survey large numbers of people (essentially the entire populace), in novels like George Orwell’s 1984, and The Handmaid’s Tale. If you can think of a dystopian novel, it likely resembles Foucault’s version of a super-panopticon. In these dystopian societies, the powerful creates an environment where a person believes they are constantly surveilled, and therefore must comply with the hegemony in every decision. To us, this anxiety is quite obvious: we are not perfect beings, and therefore we are bound to make a mistake that will put us in the cross hairs of the powerful. Like our fictional counterparts, do we feel the cold stare of the government on our necks at all moments? Or do we truly feel free to do whatever we deem right?

201 Group 6

In the BBC article titled How Supermarkets Tempt You to Spend More Money?, we learn the trade secrets behind how large corporations can market to consumers in the most effective way. We find that there is a lot of strategy behind where each individual item is placed in a store. This phenomenon exists not only in physical stores; we believe that it is most prevalent online. For example, when shopping Amazon, many times there are suggests that the company makes to supplement your order (i.e. a charger with your calculator purchase). We believe that the personalization behind these suggestions online is more powerful than that of a tangible store, as all data is accessible and malleable in real-time. In comparison to physical shopping, it would be as if the entire store reconfigured itself after every single item that one puts in one’s cart. The question is not whether or not companies can effectively target a consumer to their perceived needs, but it is whether or not it is ethical to do so?

Big corporations are not the only entities that benefit from big data. In Tufte’s Envisioning Information, we see data visualization being used for scientific knowledge. Big data is a catalyst for individuals to navigate an ever-changing complex global society. For example, one of the diagrams in Envisioning Information depicts a train map of China. This diagram shows that data is important and useful to everyone. Without data that is easily accessible, there becomes a lack of knowledge that is detrimental to mankind.

Group 3 Observers: Phones, Computers, Watches, and Shoes.

Saad Farooq’s Study on Phones:

Method:

I noticed who had his phone out either on the desk or in their hand, during  the class period.

Intentions:

I intended to see how the this affects student participation in class.

Generally comparing how having the phone out or phone not out could differ a person’s level of concentration. Does the phone become a distraction for most students?

Jason Feldman’s Study on Computer Presence:

Day	With Computer	Without Computer
Tuesday	16	8
Thursday	18	6

Collected these data by observing who had their computer out on their desk during class. The idea was to see how much this number changes from class to class, whether the number of computers steadily goes up or down, etc… I wanted to see how it may affect participation, but there were issues with the ability to keep track of participation and what was the best way to actually measure it.

Andrew Feld’s Study on Watches: 

As someone who wears a watch somewhat religiously, I was curious to see how many other people within our class (including Owen) wear a watch. I defined a watch as any item worn on the wrist, in which the item’s main function is to keep time and has an assumed lack of electromagnetic communicative ability. To study this, I arrived early to class Tuesday and Thursday (August 29^th and 31^st) and used a counter application (called Counter+) to count every person who entered the class wearing a watch from my arrival until Owen signaled the beginning of class by taking attendance.

My question would best be answered in terms of a proportion, therefore, I counted the number of students present each time Owen called a name and the student confirmed their attendance. I kept up with the total number of students using a simple tally-mark system. My intentions for these data are to cast light on the possible redundancy of watches in a society saturated with mobile devices that can keep more accurate and precise time than their simpler counterparts. Also, it is my hope that this data will either start or add evidence to the debate about how we define objects, as technological innovation expands the utility of objects.

Nick Fantuzzi’s Study on Shoe Brands:

Method:

 

Survey question sent to the entire class via email.

 

Context/ Breakdown:

Last week, I chose to observe shoe brands to gain more information on brand preference within our class. Since I only tracked these brands for two days, the results are relatively inconclusive, but I did make a see an abnormality related to the method. For 8/29, I sent out the survey in class and then a follow up email the next day. I received 22 responses out of a possible 26 (25 classmates + Owen). For 8/31, I sent out the initial survey, but did not use a follow up. I only received 13 responses for this day. This could be an anomaly, but I did find the disparity interesting.