COMS 4995: Internet Technology, Economics and Policy (Fall 2017) - HW1

The assignment is due on Thursday, October 5, 5 pm EDT, via CourseWorks.

  1. Payments: As an alternative to paying for content via advertising, consider what it would take to allow for a no-ad, fee-based version of web sites and apps in the United States. Estimate how much it would cost a consumer each month to replace advertising revenues for Facebook, Google, and the New York Times.
  2. Anonymous browsing: Document the type of ads shown (e.g., travel, men's clothing, Internet services) on a news page for three of the top 50 news sites with and without Firefox anonymous browsing, with tracking protection either enabled or disabled. You can use screen shots, but a simple description ("video ad for Volvo") will do.
  3. Tracking cookies: We discussed tracking cookies that follow you after you visit certain web sites. By experiment, determine whether car manufacturers, financial services companies or ecommerce sites employ tracking on Facebook and on a general-interest newspaper site. You may not find any, so it is sufficient to document your experiment.
  4. How many ads: Consulting the revised July 2017 IAB guide for ads, tally the ad types displayed and whether they use native ads (called "paid posts", "sponsored content" or similar), Google Text Ads or other advertisement types listed in the IAB document. Summarize your tally for ten web different ad-supported web sites, chosen from one category and sub-category of popular sites. Thus, pick an Alexa category and sub-category and then consult the front page for each of the ten most popular sites. Some categories, such as shopping, may not carry advertising (except for their own products), so they may not be suitable for this comparison. We will then "crowd source" statistics across categories from your data.
  5. Internet usage - money or education?: Internet usage in the US strongly depends on income, education and rural vs. urban. However, the three factors are closely correlated - households with higher educational attainment generally have a higher income. Also, households in urban areas have, on average, a higher income than those in rural areas and may also differ in education. Using data from the Census and the Pew Research Center, compare Internet usage for the three variables (rural/urban, post-highschool education, income, e.g., in three bands). If you have taken statistics, you can create a prediction function and show how well your function explains the data (r-squared). If not, choose two variables to keep constant and see how the third changes the outcome.