Description

At the beginning of our research, we focused on tracing who was participating to the debate online. We found the most popular website talking about marijuana legalisation and we examined if they were always the same actors talking across all the country.

We started by taking a map of the United States and positioning each website - as a dot, coloured with his target category- within the corresponding state, once we had all the dots in place we linked all the websites - therefore dots - appearing in more than one state.

The buttons allows to see just one category at a time and the list of websites on the right side to look at each websites behaviour in the map. The websites are ordered according to their popularity. This way, we can see which are the main actors of the debate, where do they speak and how they spread across the United States. The above output is a new geography of U.S.A. when talking about marijuana legalisation.

Below the first map there are to maps that shows the most and least interconnected states. The former five states, North and South Dakota, Michigan, Delaware and Maryland have seven domain in common with the others while the latter, Tennesee and Ohio share just one domain with other states. We were looking for a pattern or some corrispondence whereas all these states share a similar legal situtation, some of them allow the medical use only of marijuana while others consider it a misdemeanour. We didn't find the pattern we expected through this protocol.

Another finding is about the most popular websites throughout all the United States which are norml.org and thecannabist.co, both with the mission of bringing awarness and news about marijuana and its legalization in the U.S.A.

Protocol

We started our protocol searching the query “marijuana legalization” associated to all the States within the U.S. in order to retrive our actors. We gathered 10 websites for each state, a total of 500 websites, then proceeded to cluster them in the categories found in WebPulse Site Review. In excel we then created a pivot table to see the most popular website categories and selected the top five: “news”, “political”, “marijuana”, “government” and “reference”; we assigned one color for each of them.

Data

Timestamp: 01/12/2016 - 05/12/2016

Data source: Google, Webscraper, WebPulse site review

The datas are one Excel file with one sheet for each State containing the 10 wesites retrieved with the scraping. The informations contained for each website are its title, href link, domain, category and keywords. There are also other 10 websites for each state downloaded just to be sure to have at least 10 usable links.

The other Excel file contains all the 10 websites for each state in one sheet, then there's a pivot table that counts how many websites for each category and another sheet that is the result of the websites associated to each state just of the selected categories.