Dear Stats 252 students,

The second tab in spreadsheet SDRSurvey_Compiled_Students.xls has the 106 anonymized responses to the survey on the social data revolution. Please share on this wiki below one insight next to your name by noon, Monday May 18th. (Feel free to contribute other ideas to questions you are curious about. The names are assigned at random to the columns, see the first sheet, the letter next to your name corresponds to the column. And if your name is not listed, please pick a question you are interested in).

If you believe your insight is cool, please put one star * in front, up to three starts if you want everybody to see it.

Thanks, Andreas

Table of Contents

1. How did you first hear about the swine flu outbreak?

ASW: C2C makes up 20%. Would you have expected more or less?
Adji, Tirto [3]: From the table below, internet news is obviously the most dominant form of communications here. People without time to catch up with sporadic news through the day through newspapers or television or other medium, can hit up on these sites or get the latest breaking news delivered to them.
internet news
another person
news (unknown)
can't recall


Murphy,Robert Cornelius [3]: The majority of students found out about the flu outbreak through traditional authoritative news sources. Specifically 64% discovered the news through directly viewing or listenting to sources such as CNN, NY Times, and NPR. Of these, a significant portion (28%) used “old-fashioned” means to access their news source (Radio, TV). It is interesting that such traditional sources persist through new technologies. While computers, the internet, and iPhones have changed the way in which people access their preferred news sources, they apparently have not changed the sources themselves. Blogs and social media sites were among the sources students got their news from, but they made up a very small percentage of the whole. A question now is whether this small percentage is indicative that these social sites are insufficient for getting such news or if instead this small percentage is the start of a trend. It would be interesting to compare these results to the results of where students (and others) find out about other types of information (social, entertainment – movies/music/sports, etc).

Erdem,Tevfik Burak: As it can be seen from the numbers above, more than 40% of people learn about breaking news via the internet and it is followed by TV with %21 of people. This is something expected as most people have access to the internet during day time while they may have limited access to TV, radio or friends. To me, the most interesting result of this question is the respectively high percentage of people who learn from someone else first. This shows how word of mouth is still important for communication even in a kind of breaking news that takes a huge coverage on internet and TV. I guess, the percentage of word of mouth would be much higher if this was a news which did not have as much media coverage as swine flu. For instance, the launch of a new product such as Kindle 2. Therefore, word of mouth is still important and companies should take this into account while deciding about their marketing campaigns. Another interesting result can be the very low percentage of people who first hear about the flu via email as emails from specific groups or friends are usually good news sources for people. I think, this is because the swine flu news had huge coverage on media immediately due to its importance. So I guess people did not first hear via email but most probably followed the updates through emails they received from groups, friends etc.

2. What was your biggest surprise ever on a social network?

ASW: The lightweight interactions reduce propagation times. While for news, it is still mainstream media, for rumors and fads and stories about friendships and breakups, it is the propagation through C2C and C2W. Not only in our lifetimethe has the world gotten connected (as in plumbing), connections also transcend the present, reconnecting with the past, and the data we create living forever on human scale, making data immortal. As in the Glass Menagerie: "I didn't go to the moon. I went much further--for time is the longest distance between two places".
Hamra El Badaoui,Bilal [3]:
Most responses focus on three issues:
a) How it became very easy to get back/keep in touch with almost everyone one got to know during his life.
b) How people became very open about sharing personal and private information on the network for things they would not share in real life. (there were also some quotes about certain misfortunes because of that)
c) How fast information circulates on the network (viral effect) and how the reported/stated opinions and activities let us discover a side of our friends which was hidden in real life. <ASW: fast growth of rumors>
Social networks have a major impact on the way we live and see our lives. They've changed the mindset of the people and the environment we live in.

Pary Harikrishanan, Sowmyalakshmi: [4] The common sentiments running through the replies are surprise at how willing people are to share personal information on a public forum, and how easy is has become to regain contact with friends from the past. In combination, these two phenomena beget surprise at how much people have changed or result in surprising facts about friends coming to light, particularly through facebook newsfeeds.
The two-sides-of-the-coin factor is very much evident - some say that they get news faster through these networks then the usual channels, while it can also engender annoyance when facts about close friends are revealed here rather then more personally. Some have been put off by being inundated with trivial updates while some have found crucial career opportunities through these networks.
It seems that people are still adapting their tolerance levels for receiving so much personal information, but they realize that the real advantage presented by these networks to sustain connections is not to be taken for granted.

Paniagua, Pablo [3]: To many, the most surprising thing was reconnecting with old friends. Words such as friend, friends, old or school appear prominently. Others share personal stories and many comment on privacy and the amount of information people are willing to share. What is indisputable is that no matter what was their biggest surprise, it was most probably on Facebook. The size of the word compared to MySpace or LinkedIn is very telling.

3. What's the coolest thing you have seen on the web in the last month? Why did it delight you?

Pribula,Alexis Juri
It is hard to group the answers into meaningful classes so I am listing the ones that seemed either very cool or frequent (indeed, absolutely subjective):
  1. twitter search/updates on swine flu
  2., <ASW Caterina Fake's new company, asking questions, and helping people with decisions> Launch Jun 15> c.f. >
  3. Susan Boyle's rise to fame
  4., a new way to search for recipies
  5., a new way to manage your finances, analyze your credit/debit card transactions
  7., connecting graphical designers and people who need them
  8. '6th sense' tech demo at TED
  9., an image search engine
  10., find out what the best room in a hotel is <ASW Yay! Finally>
  11., sql-like syntax to search the web
  12. :)
  13., python code to analyze your email
  14. Yahoo Pipes
  15., sweet music discovery website (with basically free music)

Singh, Aditya


The tag cloud (with common words edited out, part manually, part electronically) shows that, as expected, google, facebook, youtube and twitter generated plenty of buzz; being responsible for creating or delivering data/information that people saw and found interesting. This cloud tag ends up answering the question "Where did you see the coolest thing on the web?" rather than "What is the coolest thing you have seen?" and does a terrible job at "Why did it delight you?".

Sinha, Shakti
Grossman,Nathan Stuart

4. Which blogs do you follow? Why those, and what does it mean for you to follow a blog?

Andrei,Victor Alexandru

Long tail behavior in blogs
Following blogs is similar to shopping for music. You follow the blog in accordance to your interest. Few of us like blogs that talk about surfing, playing poker or baking bread (areas of smaller common interest), and at the same time more of us will be interested in technology or sports (areas of larger common interest). As with music, people who decide to follow blogs make two types of “purchases”: social and individualistic. The social choice is the choice made in the area of common interest (in our case most people who wrote the survey followed blogs about technology – techcrunch, gigaom, etc.), while the individualistic choice is the choice made in the area of a particular interest (some people followed blogs about surfing, friends and family, or Stanford FML). In this view I think that the blog has revolutionized the architecture of the web. Until the blog, only popular websites made it on the Internet. Blogs have created the long tail in the website industry the same way the Internet did for music or books.

Before having seen the results of this survey, I personally expected that due to the variety of topics in the blog industry there would be very few “block busters” because I thought blogs were websites that had really targeted content. Surprisingly, I realized that you could map blogs to interests. It then follows logically that people will have some interests that are common within larger populations, generating thus “blockbuster” blogs.

Tronson, Andrew

The chart below shows the most popular blogs from the survey responses. The size of the horizontal bar corresponds to the number of people who cited a particular blog. It is important to note that only blogs which received 2 or more votes are included in this chart. Also note that individuals who stated that they follow “Technology blogs” or “Business blogs” are not included because they did not list specific sites.

From this chart, we notice that the most popular blogs are areas of common interest among survey respondents, namely technology and business. We also observe a long tail of blogs cited by relatively few people, which correspond to areas of personal interest. In fact, there are over 120 more blogs which received a single vote, but are not included in the chart. These quantitative results reinforce the ideas summarized by Victor, above.


Sumbaly,Roshan Rajesh
- I could categorize people into two categories. The first group of people read loads of blogs primarily focusing on technology or blogs from their family/friends. The technology blog readers also showed a lot of inclination towards startup related blogs. [Michael Arrington would be a happy man since Techcrunch was found on most lists]. The primary motivation to read up on these blogs was to stay updated with respect to either technology or their friend's lives. The 2nd group of people can't find time to read up blogs because of their tough schedules. A simple count of the survey shows around 24% people are interested in reading blogs related to their family, 48% love technology blogs and the rest 28% either don't find any blogs interesting or have tough schedules.
Heng, Serene

Soundararaj,Sai Prashanth

In my opinion, people follow blogs to be up-to-date on things that are close to their interests, since blogs tend to provide a great breath of information. Individual interests of people follow general trends and fall in definite categories. Technocrati's top 100 blogs have most falling in the following main categories: Technology, News, Culture, Art and Design, Fun. The chart below shows the breakup of the responses in the popular blog categories. Tech blogs, like Technocrati, Engadgets,Mashable, etc are the most popular, followed by miscellaneous blogs that are close to people's interests, e.g. stuff people like, post secret.
Top blog categories

5. How many people do you follow on Twitter? Why those?

Assudani,Tripti Bharat
Around 50% of the class does not use Twitter. For the ones who do use Twitter,the following pie-chart shows who the class follows on Twitter:

Who do people follow on Twitter?

Of all users who use Twitter, the people/accounts they follow fall into multiple categories. Some people follow only friends while others follow a combination of friends,news/media websites,celebrities and technical websites and blogs. As the pie-graph shows, friends is the most popular category of people followed(34% of votes). This is followed by news and media websites and people use it to stay up-to-date with world news and news primarily related to politics,sports and the economy. Celebrities and technical sites/blogs are tied at number 3.The class has people enthusiastic about following updates of celebrities,leaders and icons(Barack Obama, Lance Armstrong etc.).A number of people are interested in technical blogs/news, entrepreneurship and business and follow both people and organizations to connect and be informed. As evident from the graph, only a minority of people are using Twitter to be in touch with their families and colleagues(people you meet often on a day-to-day basis and interact with closely either in person/or via other means).

Aisen,Daniel Charles
Below is a histogram of the number of people students follow on Twitter. Notice that nearly half of all students do not use twitter at all and over three quarters follow 20 or fewer users. Very few students in the class are extremely active on Twitter, and based on the comments it seems that most students prefer Facebook to Twitter.


Cutler,Blake Robertson
I wanted to see if there is a difference in the purchasing habits of Twitter users and non-Twitter users. To begin answering this question, I created a binary variable that classified whether a classmate uses Twitter. In figure 1 (below), I created two histograms that illustrate number of purchased items students discovered through a friend. The histograms indicate and the summary statistics (figure 2) verify that Twitter users are more likely to discover purchases through a friend. In addition to having a larger mean, Twitter users had a larger standard deviation in the quantity of their friend-discovered purchases.

Because this data simply addresses correlation, we cannot infer causality. Either students who use Twitter are more likely to purchase items through a friend or Twitter encourages students to buy more items through friends.

Figure 1.

Figure 2.

Figure 3. Note: if the number of followers was not defined, those users were not included in this chart.

Hong, Junghuan
Robertson,Harry James Mayflower

6. Of the most recent 10 items you bought, how many did you discover through a friend? (Not counting trips to the supermarket for food or similar.)

Dumbacher,Brian Arthur: Below is a histogram of the number of the last 10 items purchased that were discovered through a friend. Uncertain responses were not included, and if a student gave a range of numbers, I took the average and rounded up. As you can see, more than half of the students thought that only four or fewer of the last 10 items they purchased were discovered through a friend. Furthermore, the value occurring most frequently is 0. This came as a bit of a surprise to me, but it appears many of us do independent research and discover products on our own.
Crawford, Erika
I found these results to be very interesting, since I personally expected to see respondents saying they'd discovered a higher number of their recent purchases through a friend than the results show. However, I think that the more interesting information was brushed aside by this particular question -- after all, this question seems (to me) to clearly have been intended to show the importance of word of mouth and recommendations/insights of those closest to us in making our purchasing decisions. Yet, the bar graph above shows the exact opposite of this. The more interesting -- and useful -- information to have had would have been, for example, what types of items we buy after discovering them through our friends. Are they expensive items, such as vacations, cars, homes, and appliances? Or more inexpensive purchases, like toiletries, entertainment, or clothing? Of the few people who did provide such information, the items purchased post-recommendation usually were in the less-costly category. (However, we purchase fewer expensive items in general, so the likelihood of one of our past ten purchases being one of these more expensive items is smaller. It would have been valuable, to answer this question, to have asked how many $1000+ items purchased in the past year had been discovered through a friend/family member.) Also, an equally important question would be the extent to which the discoveries/tags of "random" people -- individuals we do not know -- affect our purchasing power, since we can optimize search result pages for this (ie Listmania lists on Amazon).

Jones,Matt Knight
This seemed like a tough question for a lot of people, as a lot either didn't remember what they had bought recently or weren't sure if they had heard about it through a friend. This seems to be indicative of a few things: First, there's not an easy way to look at purchase history. It's possible to look at a credit card bill, or do online analysis with a system like, but these fall short for cash purchases, don't include itemization, and of course don't include information about how you discovered a product. As consumers become more interested in the way they discover goods and services, having a platform for tracking this may be more important. It also seemed that it was difficult to define whether recently-purchased items were discovered through a friend. In general, interactions involved friends, but the exact product wasn't necessarily recommended by that friend. There's no disputing that friends influence what we buy, but it's hard to tell from these data to what extent or in what specific ways.
Segal,Aleksandr V

7. Do you pay for any online content? If so, for what?

Kim,Nam H:
Here is a very crude data visualization using a tag cloud.
We see that entertainment is the biggest online content that people pay for. I created a rough growth graph on annual itunes songs sold.

Another entertainment that we seem to be missing from this is games. Online gaming industry has grown significantly, but because this is a Stanford class, there aren't that many people paying for it. So I decided to contribute by giving some data on the most popular monthly-subscription based MMORPG, World of Warcraft.


World of Warcraft, just within 4 years, has gotten over 11 million subscribers in the world, which translates to over 165 million dollars each month. Although this growth isn't as large as itunes sales growth, I believe that this is another type of internet content that a significant amount of people pay for.

Eng,Carlin Paul:


This tag cloud does a decent job of visualizing the responses to this question. It excludes several of the most common words, including "no", which was by far the most common response. Along those lines, "free" was also a common word. Amongst services that people do pay for, music (including itunes) is the clear frontrunner, with news sources like the Wall Street Journal and The Economist coming close behind. Some people consider paying for Internet service or data plans on a phone as paying for online content.

Beekman, Jonathan David
An interesting insight from the data: People in general believe that content on the web should be free. Because there are so many alternatives people have regarding finding free content, paying for content, unless incredibly specialized - is a hard sell. That content that people purchased online included content that is specialized but also distributed and freely available on the web. This leads me to believe that people also are willing to pay for convenience. This was borne out by the many people subscribing to news sites and research tools. Both likely have a significant "convenience factor" in that they aggregate content that is considered to be pre-vetted in a single place.

Huang,Jennifer J
Shah, Hemal

8. Some people share in great detail on Facebook, Twitter etc what they are doing. Has this changed your notion of friendship?

Hui, Katrina
Here is a graph depicting the answers that most people had in response to the the question "Has Facebook/Twitter" changed your notion of friendship.Most people seem to believe that Facebook has not changed their idea of friendship and believe true meaningful friendships to be based on personal interactions, rather than online status updates or wall posts. Around half of those who responded "yes" generally find that they are able to stay in touch with people with ease or learn more about people through their updates or profiles. People generally agreed that they were able to stay in contact with people who liver far from them with more ease. They also believed that their relationships with their acquaintances strengthened by learning more about them through Facebook but that their relationships with close friends remained the same. The other half of people appear to hold the belief that the overall standard of friendship has been lowered and that friendship has been a broader term applied to a lot of superficial relationships. There were also a portion of people who offered responses that did not quite answer the question, but rather commented on Facebook/Twitter's effects on friendship.

Cardaci,Tina Marie

The class was split regarding whether Facebook and Twitter have changed the notion of friendship. Interestingly, the split seemed to be centered around whether students found the news feed to be useful. To some, friends' status updates were relatively meaningless and often unnecessary. They would prefer not to hear about the bizarre activities of mere acquaintances. To others, however, updates revolutionized their idea of friendship. These individuals feel they now discover things about their friends they otherwise wouldn't. They benefit from staying up-to-date on what's going on in their acquaintances' lives. Regardless of whether students feel that social networks have changed their notion of friendship, we can all agree that it has changed the way we interact. The Social Data Revolution has impacted us all, even if we find status updates to be a nuisance.

Although face-to-face interactions with friends generally remain unchanged, Facebook has added a new dimension to friendship. We may be 1000 miles away from Jacky, but we still manage to find out he’s having computer troubles. Or we might have not seen Christine all year, but we are still kept up-to-date on her latest crush on a robot. Facebook might not offer the most meaningful information about what’s going on in friends’ lives, but it does offer something. For people we’ve lost touch with, it reminds us that they still exist and keeps us interested--at least sometimes--with what’s going on in their lives. For friends we see every day, it gives us perspective on what they think about when we’re not around. Overall, Facebook has made it easier to stay in touch with more and more people. Keeping in touch with 700 friends is not an easy task, but social networking sites have allowed us to maintain such a network in a relatively effortless way.

9. In what areas do you want to discover new people (for jobs, dates, …)

Krishnakumar, C.V
It was really amazing to note the different interests that our peers have in meeting new people. The responses given were parsed and a TF analysis was carried out to describe the most interesting areas that people would like to find new people in. The text of the responses was taken and its words were normalized (using case-folding and Porter’s Stemmer ) and the words of interest were extracted from them. The different interests that the students have are summarized in the three segments as follows:
Categories of Interests in Meeting New People

Professional Motives form a majority as expected from a class of students. The relationship building follows at 25% and hobbies comes 3rd.
Constituents of Relationship-Building Measures:

This chart shows the categories of relationship building that seemed to interest the students of the class, in response to the question. People liked to make new friends as indicated by the 32% support it has. In addition, dating seemed to find favor, coming second at 27%. Family, colleagues, neighbors were other factors with whom the students of the class seemed eager to connect with. As an aside, dating seemed to quite popular motive behind using the social networks. It constituted around 27% ( representing that out of 100 people who had given at least one relationship-building motive, 27 were likely to opt for dating) and was just next to finding new friends.

Distribution of Hobbies:

The most popular hobbies and extracurricular activities mentioned in the response are summarized in the chart. It was interesting to know that among all the people who mentioned any hobby, nearly a third (30%) were very interested in Music. Other activities included traveling, golf, skiing and biking and their relative importance to the students of the class is roughly summarized in the percentage distributions given by the chart.

Professional Interests:

As evident from the pie-chart, most students wanted to network for jobs ( comprising of 23%). The primary motives revolved around obtaining business connections, expert consultants, partners to support entrepreneurial ventures. Certain students were also inclined to learn from others and share their knowledge and expertise with others.

Ma,Eric Shubo

I went through each of the responses for the question and labeled on 4 of the prevailing, general categories: Career Oriented, Meeting New People, Continuing Relationships, and Dating. There were a variety of specific comments within each category, but I noticed that for career-based people, individuals were primarily looking for people to work with on startups or networking to find jobs. Under the Meeting New People Category, a lot of people were looking to share hobbies (face-to-face activities), meet new people to discuss various interests, or joining online communities to be able to discuss similar interests (without face-to-face interactions). For continuing relationships, people mentioned using social networks to keep in touch with old friends and see what acquaintances were up to rather than meeting new friends. Finally, the last category were just people mentioning they'd use the networks for Dating. Below, I've attached a histogram, showing that Career dominated responses. The percentages don't add up to 100% because many people listed multiple choices.

What I found to be especially interesting was that although some people sounded quite open to meeting new people, there was the opposite spectrum with many individuals who would never want to meet new people online. I wonder if the people who said they wanted to meet other with similar interests would in the future or have in the past actually take the initiative to meet people in real life after networking with them online. It seemed like everybody was comfortable with using social networking for career reasons, and many who didn't list it as an option explicitly stating that they weren't interested because they already had a job.


Chamoun,Emile Elie: The tag cloud pasted below is a very good representation of the interests people have in social networks. Career networking seems to be the most important reason people join social networks as shown by the size of the words career, networking, linkedin, jobs, work and professional. Meeting people with similar interests ((hobbies, share, similar, social,hiking, common) and forming new friendships (finding, friends, friendships) were often cited as the main reason students use social networks.

While it seems reasonable that professional networking, friendships and discovering people with similar interest were the most popular answers among students, It is interesting to see that only 12% of the class mentioned interest in dating. This is in part due to the fact that part of the class is already in a relationship. However, it can also be explained by the fact that most people do not join social networks for the sole purpose of finding dates: once someone joins a network, they might find people they will date but that's not the main reason that attracted them to social networks in the first place.


Johnson,Theresa Lynn
Vasuderan, Prasanna

10. The web has had a huge effect to say the least, e.g., Craigslist has essentially replaced classified ads. Progress has been slower In other areas, such as online dating. What is your view of online dating? What works well? What does not work, why not? Any suggestions?

I thought it would be interesting to take a look at the distribution of two key measures from this data: People's experience with online dating, and people's feeling regarding online dating. I took each text response, made an assessment, and assigned each response a value for experience and a value for feeling. The cross-plot is as follows:

Two-Dimensional Chart of People's Feelings vs. People's Experience Regarding Online Dating

Your Feeling Toward Online Dating
Your Experience
No Opinion
Don't Know
2nd-hand stories
Can't Use
Haven't Tried
Some interesting insights I drew from this table:
  • People's feelings toward online dating are still quite split. However, the majority of people still perceive online dating in a negative light.
  • Very few people have actually tried online dating. This should be qualified by 1) The question did not specifically ask about online dating experience, and 2) We cannot assume these responses are 100% true. In the population of people whom can participate in online dating (Tried+Haven't Tried), those who chose not to participate in online dating outnumber those who do by 4:1.
  • Of the people that tried online dating, most have negative feelings regarding their experience. However this is based on 6 responders.
  • Of the people who have not tried online dating, the majority feel negatively about it.

Sharma,Chetan Veeru Bhotla
The majority of issues with online dating is that of trust. Specifically, users of online dating services can put down any information they want, and skew it in any way to make them perceived as more attractive. Students specifically mentioned the photo, in which the user can pick a photo that can distort someone's true physical appearence. In general, there is more incentive to lie to get an initial interaction than to tell the truth to avoid later conflicts.

Beyond the trust issue, there were some other interesting insights. I was particularly impressed by one comment that said, "Expressiveness in writing does not equate to equal expressive power in real life". It is true, online dating puts heavy weighting on the means of communication used on the site, which is usually writing. There needs to be a way to capture the expressiveness of one's conversational skill from the start, rather than passing through a writing vetting process.

Another interesting note is that some users pointed to the possible consequences of interacting with unreputable sorts. Responders mentioned sexual harassers, rapists, and other horror stories that provide serious disincentives to use the service. It seems that there needs to be some invitation process or other screening method to provide some guarantee that the users are good people. The possibility of an interaction with a sexual criminal, whether it's real or simply perceived, is a serious obstacle to any successful online dating service.

Djian,Pierre Abraham:
A few insights:
1. Yes progress has been slow in online dating. From the sample answers, about 2/3 state that they have never used online dating
2. Need for better matching tools / processes. Out of the people who say they have used online dating, only 3 state that they have had a positive experience. So clearly, there is progress to be made in matching people and improving the success rate of online dating. It feels like it is not a very mature space which still needs to find the right tools to meet the needs of its target users.
3. The most interesting comments / suggestions are around: i)the need for more authenticity in people's profile [need for a form of verification of information], ii)the need to combine online and offline experience, iii)the need to take into account the "people filter" [in real life, your friends will filter the right people for you based on what they know about you and the other person -> need to recreate this online if possible], iv)the need first focus on activities where people could interact anonymously [e.g. discussion forum] to assess their match / similarities on more objective elements, and then only be able to uncover their identity / picture [this way, no storytelling about yourself in the first place]

Kapteia, Maurits
Weiksner,George Michael

11. What are examples of data you would share more readily with your online friends than with your real life friends?

Qian, yana
1. A majority of answers said that their online friends are the same as their real life friends.
2. People tend to share their data online in a more conservative manner.
3. These data are shared more readily online due to the ease of sharing: Photos, blog posts, weblinks.
4. Ease of sharing to a lot of people online: Facebook status, Twits, hobbies, travel plans.
Jinadasa,Sampath Deepal:
I read everyone's comments and grouped them into the following 5 categories. The majority of people did not answer the question or did not understand the question. Status updates, hobbies, and minor details about one's daily activities was tied at 21% with media content. Then 15% of the class said that there is nothing they would share more readily online than in person. Finally, 10% felt more comfortable sharing personal information such as emotions or embarrassing stories.

Khare, Indrajit
Wu,Shengxi David

12. What difference does it make to you if someone sends you a virtual rose vs a real rose?

Shi,Xin- 77 out of 96 people (if my counting is correct) who have answered this question vote for a real rose. They don’t think virtual roses can replace the real ones for the following hot reasons: sending a real rose requires more effort to attain whereas a virtual rose only costs a click of a mouse; a virtual rose is fake and meaningless and virtual gifts are nothing but jokes whereas real flowers are with real scent, rarity and beauty;a virtual rose can amuse them, but it cannot overly impress or provide emotionally impact on them.
On the other hand, there are 12 people in the class who are indifferent between a virtual rose vs a real rose. Some hot reasons include: they are both gifts; the choice definitely depends on context--a real rose is more meaningful because the effort required to go buy a rose and send it in person is much greater than clicking send on a facebook gift; on the other hand, a rose on one's profile will get more views than a real rose sitting in her room. So perhaps a rose online will be more significant in terms of boosting one's social status.
Finally, 7 people actually prefer to receive a virtual rose mainly because its low maintenance but carries the same message and publicly visible to all friends is somehow nicer than a real one that is only visible to the receiver herself.

Ng,Chee Wei-The results just go to show that even as more things go virtual, like shopping at the mall evolving into online shopping, it is hard to replace some things like a real rose. Humans do have a preference for tangible objects, especially if they are gifts. There is a much greater significance and emotional response when you receive something that you can hold, smell and bring home. Virtual gifts still have a long way to go before they'll even come close. They are probably appropriate for friends, but definitely not for your significant other.

Narayanan,Aravind - A majority of the answers preferred a real rose. This was because it meant that the sender had put in more effort and time to show how much the person meant to them. On the contrary, a virtual rose often signified laziness on the part of the sender. It was also viewed as less personal, if the action was viewable by their entire network of friends. It was amusing to note that most of the people who said they did not have any preference between the two types were male.
Kim,Joon Yeong

13. Do you use Twitter? If so, is tweeting yourself more important, or reading what others wrote?

Sniadecki,Jennifer Lee: Below is the percentage breakdown of the first question, "Do you use Twitter?". A tag cloud is also displayed for most common sentiments made by the students in the "infrequent" category.
Do you twitter?
Sumithra Jonnalagadda
Zamkow,Robert Maxwell

14. What do you expect you will use Twitter for in half a year?

Wu, Wen:
The answers to this questions can be approximately grouped as 6 main ideas. 39 people showed their expectation on connecting with friends, sharing cool stuff with family members and following interesting people. 31 people said that they what to use twitter as a main source of updating news and information in particular fields they are interested. 10 people showed that they hope they can use twitter to help their business and careers. 24 people confessed that they don't use twitter or don't like it, though they don't mention the reasons. But 3 of them said that they are going to sign up and start to use it. 9 people said they don't expect any change on using twitter.

People have different opinions.
Some people will sign up Twitter in the future and they want to use it in professional way. For example, in financial field, marking.They hope Twitter can have more applications like networking for career growth or job hunting itself. And others want to use it as informal ways. Like sharing things they find interesting, follow their friend's updates. There are still a lot of people donot like Twitter and will not sign up in next half year, they feel twitter are not cater to them, or Twitter just for young people.
Twitter need to do more ads to enhance their number of users.

Kuok, M. J.
Zhang,Morris Jinhui

15. If you learned that a company uses data you shared for a different purpose, e.g., to trick you to listen to their message, would you be less likely to buy from them? Can you give a specific example?

Based on the survey result, clearly misuse of data has an impact on customers (92.45% are affected). More than half of the customers will be less likely to buy from a company that uses data for a "bad" purpose, the most referred to reason as the loss of trust. 39.62% of the people will be affected but will decide according to the specific situation (i.e., how malicious the use of data is directed, how loyal the customer is to the products). For those who are not really affected, they consider this kind of company behavior an understandable way of marketing, and somewhat irrelevant to the products themselves.

Cox, Tayler
Of those respondents who provided explanations of the circumstances under which they would continue to buy products from the company, I assigned them into one of two categories. The first categories I called "principle." Respondents in this category indicated that the loss of trust, lack of honesty, loss of reputation, or simply feeling "tricked" by the company was enough reason not to buy from them again. Respondents fell into the second category, "utility," if they indicated that they would still buy as long as the information was used to provide relevant or useful information to them, and on balance did not cost them additional time or money. The breakdown for the respondents was 31 in "principle" and 34 in "utility."

On balance, it seems that as a long as a company provides useful or relevant information to consumers and does not waste valuable resources, 50% consumers are willing to overlook the breach of privacy or unapproved use of information and will continue to buy products from the company. However, no matter how useful that ad, the other 50% of consumers will be less likely to buy based on principle, seemingly regardless of how well the information is used. It might be worth it for companies to simply ask the customers for an explicit "opt in" permitting them to use the information to target ads in the future. While not all customers will opt-in, at least that way you avoid offending and losing the trust of a large chunk of customers.

Of course, it would be interesting to see in practice how many of the principalists actually stopped buying less if the company succeeded in providing relevant and useful ads to them all the time. I suspect that in practice, utility usually wins out over principle.

Prashanth, Sai
Lin,Christopher Guo

16. Do you receive marketing messages from businesses via your mobile phone? How do you generally respond to these messages? What are your hopes and fears for marketing via mobile phones?

Benson,Catrina D - People tend to view marketing via mobile phones in a very different way that marketing via e-mail spam. Almost everyone views phones as much more private that e-mail. Phone messages and texts from markets are viewed almost as an invasion of privacy. Also, people are concerned about the possibility of advertisers collecting private information such as position data and loosing privacy in the area. Most people simply ignore marketing messages via cell phone anyway, so it seems the practice is not very effective.
Griffith,E. Bradley Donaldson

18. If you gave a marketer access, what would you consider fair in exchange? Pick a couple of areas and explain: how much cash, what kind of personalized services, ...?

Todd H Sullivan: Two main groups: those that want money in return and those that want personalized services. There was some crossover, but not much. The money group ranged wildly from a one-time payment of $100 to receiving $100,000 annually. The personalized services group was mostly interested in receiving better targeted ads and deals on things that they are interested in. Most people mentioned a discount from 20% to 40% off. The primary concern is that the targeted ads must be relevant. Most responders were not optimistic about the current state of targeted advertising (for example on Gmail), and would only share information if they received far less spam and much better targeted opportunities. A large concern for responders of both groups was about automated vs. human processing. Automated (by computer) processing of the information and in-aggregate processing was mostly seen as okay as long as the responder received some kind of compensation. No one was okay with a human viewing and processing his/her individual information.

Bryant,Sylvie Elaine: Some people were quite specific about the type of information they would or wouldn't provide (eg. cell phone texts vs. facebook wall content vs. e-mail headers, etc). This graph depicts a generalization of the results. If a person was willing to accept some sort of compensation for some access (generally less private information), but unwilling to accept compensation for more private information, I still categorized them as willing to accept compensation in the categories listed below in the graph. There are also people who wanted a level of control over the information and type of information is provided to the marketer. In addition, there were people who want to know what the marketer does with the information provided. Here is the graph:

Bhatia, Nipun
Malek,Alan John

20. If you had one wish that Facebook could implement, what would it be?

Walker,Tristan Jawon - I would really like to see something similar to what Twitter does with @replies. If my friend mentions something about me in the status i would love to be notified of this. Will be especially important with the real time stream. That is 'social media'. i Should be notified every time my name is mentioned or referenced. At the very least i should have the opportunity to opt-in or opt-out of something like this. Would be particularly effective in Facebook's closed network.
Bradford, Ray
Mapham,Thomas Russel

21. If you had one wish that Twitter could implement, what would it be?

I wish Twitter could define the importance of updates. Lots of the updates were of no importance to me. I would like to show my vote to people who i follow. Then gradually, important updates will show up more often.

Andrews,Georgia Marie
A large amount of the replies asked for features that would allow the user to filter or sort the tweets they receive. Some suggestions were to sort by user-created groups (such as family, friends, news, etc), hashtags, keywords, date, etc. The idea of groups came up several timesSeveral responses asked for a "noise filter" to eliminate or hide tweets they regard as spam. One entertaining response along this line: "let me rate my posts for relevance. i have one user that just has brain diarrhea. she blabs about nothing!! but i'd like to stop following her "politely" or give her polite feedback that her stuff is obnoxious. just turning her off isn't what i want. maybe she can rate her own and my filter will only let in what's important." Another common response asked for a recommender system to help them find other users with similar interests.

Several responses asked for longer tweets. One of those complained about URL shorteners, because it can be useful to see where it's pointing, and suggested a way to attach URLs, possibly similar to Facebook. There were also several requests for a Chinese or Asian version of Twitter. Other suggestions included: integration with Facebook or IM platforms, insights about a user page, better (more filters) search, an API for developers, and better user verifaction, i.e. that public figures have the username everyone thinks they do.

Some particularly entertaining or interesting responses:
"I wish Twitter would die a slow and painful death. I'm really not familiar enough with the service to purpose changes."
"A better way to filter the people you follow on the site itself; we should not have to use outside programs to make the information we are getting more relevant and organized. - Twitter should have a filtering system so that you can group your tweets into different categories and structures for easier monitoring. You should not have to rely on outside platforms like TweetDeck to have functionality that allows you to manage your Twitter-based information in a coherent, categorical way that makes sense for your individual use case. For different people, there will be different measures of what information is relevant. These could include: geography, follower group based on a topic (e.g. the people I follow for political information), follower group based on a category (e.g., coworkers, friends), time of tweet, etc. "
"Create a real-time daily newspaper/newsfeed based on tweets all over the world"


22. What would be your one suggestion to improve email?

Polcari,Michael Peter
Students offered a variety of techniques to overcome the overwhelming volume of email. Approaches can be broken down into a few broad categories:
  • email client changes: Improved categorization, sorting, and filtering.
  • structural changes to the protocol: Authorization, payment systems, or centralization.
  • education or behavioral changes to nudge correspondents into behaving more responsibly.

Duong,Constance Gwendolyn
Chang, Eric

23. Current email systems either let a message through or mark it as spam. How often do you look in your spam folder?

Jonas Jacobson:

When it comes to checking your spam folder, the determining factor is trust. People who check everyday “don't trust” the spam filter, feel it's “not reliable at all,” or feel that even if the algorithm is pretty good, the “cost of false positives (identifying important emails as spam) is high.” This is a similar problem to a recommendation engine giving you a ridiculous recommendation - it kills trust in one blow.
Those who check yearly or never feel the opposite. They trust the spam filter and “almost never” find emails in it that are not spam.

Ramamurthi, Jayendraraj

How often do you look in your spam folder?
Frequency of check
# people
every message
twice daily
twice a week
every week
every other week
every few weeks
every month

Note The question asks how often, and the response every message is not a unit of time, unlike the others. However, I have classified responses into that bucket since the real intension of the question is to judge the confidence that people have in spam classification systems. One example of what I counted into that bucket is Every time I have spam.


  • It appears that most people (30%) never look at the spam folder.
  • The next significant behavior seems to be to check the folder daily (17.5%), followed by every week (12.6%).
  • I have classified anything more than a month (two months, six months, one year...) into occasionally.
  • ** It is a bit interesting to note that my metrics above are vastly different from another person's (Jonas Jacobson)! This, I believe, has primarily to do with the subjectivity in classifying the free-form user input. E.g., one response was I do not have a spam folder. This may be taken to mean that the person does not trust automated classification and thus checks every message, or to mean never for how can one look into a folder that does not exist!
  • ** The survey was great with very insightful questions, and valuable responses. But, as seen above, it was a bit difficult to classify user inputs! This was an important learning for me, that experiments need to be designed for data collection. If a defined set of choices had been given (of course, there were several questions in that format), the responses would have been very clear. However, a lot of other useful info, such as people explaining how they handle spam, would be missing! DJ briefly touched upon this in the LinkedIn Guest Lecture when discussing how to design experiments for the Mechanical Turk.

Chung, Ron

24. What would it take for you to use a system that ranks your incoming messages in order of predicted relevance for you?

Anderson, Chris
A summary of responses:
  • Most people would use the system, but a significant portion never would.
  • About half the people had a condition on their use, usually if it "works" as they expect.
  • Since people have extremely varying requirements, implementing a system that makes everyone happy would be
  • Several people cited business emails as being more applicable, which might make make sense for enterprise software companies.

About half of the respondents say that they will be very willing to accept a system like this. However, the other half is very reluctant to use it. The main reason that they are so reluctant is that they feel that developing such a system is very hard since the relevance is very different for each person, and for each situation. E.g. the relevance can be dependent on a very recent occurrence (an example that a student gave is if one just applied for a home mortgage, then any email regarding home mortgages will have a very high relevance). However, if there is such a system that does exactly as advertised and that it "works" (which can mean different things for different users), then the vast majority of the people will use such a system.

Cinelli, Chris
Naul,Brett A

25. One of the ingredients is for the sender to specify how relevant the message is for the recipient. Would you do the extra work and provide this information when sending a message?

The majority of respondents didn't give a straight "yes" or "no" answer. I read each answer and categorized whether it was more positive or more negative. Based on this, 56 said "yes" (in some form) and 41 said "no".
The positive "yes" answers had one of 5 general answers:
  1. Absolute "Yes" (15 people)
  2. Only if it was quick and easy (14 people)
    • a check box
    • something that requires only a few quicks
  3. Only if there was some guarantee that would make it worth their time (9 people)
    • a reminder is set for the recipient
    • some guarantee that the message is read
  4. Only if the system had limits (16 people)
    • receiver ranking should be coupled with sender's ranking
    • people shouldn't be able to mark ALL of their messages as high priority
    • limits to avoid spam abuse
  5. Yes--I already do this anyway (4 people)
    • using "IMPORTANT" in the subject line
    • setting priority

The negative "no" responses cited reasons:
  1. Absolute "No" (7 people)
  2. A systems based on the recipient's ranking is more useful (21 people)
  3. System abuse (14 people)
    • everyone is going to say their message is high priority
    • potential for spam abuse
  4. Alternatives already exist (4 people)
    • putting a recipient in the "To:" box rather than "CC:"ing
    • using bold face or CAPS in the subject line or message

Kim,Hoon Min
Many people seemed interested in the idea of "sender-initiated" marking of importance. But, these same people were highly skeptical of potential for real relevance or abuse. The possibility for false positives to occur seemed to be a large concern: A sender may think the message is important, but does the receiver really believe that's the case? If not, then the notion of importance becomes self-defeating. Some respondents (fewer than those who were skeptical) thought that if an intelligent algorithm were able to decide which messages should be flagged as important, then that method could be very valuable. It seemed that the objectivity presented by the algorithm seemed desirable.

Another interesting observation was that a good number of people believed that if the sender's message was important, then the sender should be able to communicate outside of email (e.g. telephone call, face-to-face) about the importance of the email. The complementary activity acted as a signal to the email recipient about the email's importance. Of course, this type of signaling mechanism would be more difficult to convey when talking to a stranger. But, some thought that was entirely appropriate. These people didn't like the idea of strangers noting what's important in their lives.

Overall, I got the impression that many email recipients were not comfortable with the senders' marking importance of emails. Thus, they felt rather reluctant to do this activity themselves.

Dahan, Ilan