30/5/2016

iScience Maps

 

A free tool for researching Twitter content.

iScience Maps is a set of web applications designed to help researchers mine and conduct social media analyses of Twitter content.

The tool is free and browser-based, leveraging a simple interface that allows users to examine content from both a global and local level.

Using Global Search, researchers can analyze worldwide trends over a time period of their choosing. Results are obtained from a random sample of geolocated tweets fetched via the Twitter Streaming API, and then assigned to the appropriate global region based on the iScience Maps taxonomy. Once a search is complete, users can build upon the temporal nature of this data to create animations that show change over time.

The Local Search allows the user to take Global Search results and study occurrences for more specified geographic areas. To provide these advanced location definition features, iScience Maps utilizes the Google Maps API v3.

Results from both the Global and Local Search features can be downloaded in Excel or CSV format, as well as being copied to the clipboard.

Research applications

iScience Maps was created by Ulf-Dietrich Reips and Pablo Garaizar in order to make comparisons between cities, regions, or countries regarding psychological states and their evolution following a trigger event. Applying the tool to a research question, they sought to replicate a 1993 study on personality characteristics inferred from first names.

Taking three male names associated with success (Alexander, Charles, Kenneth) and three with weak associations (Otis, Tyrone, Wilbur) from the 1993 study, the researchers conducted a social analysis via the iScience Maps’ Local Search feature. Searches looked for tweets containing each name in isolation as well as with an attribute related to being successful; namely, ambitious, intelligent, and creative. By applying a localized search, they were able to whittle results down to the Western US and the UK with Ireland.

Fig-5-Comparison-interface-directly-available-from-The-example-shows-a-comparison-for.png

Image: Shows a comparison for the Boolean search “Charles” AND “intelligent” in the western U.S. and the U.K./Ireland.

From their iScience Maps analysis, the researchers were able to make the following conclusions:

  • In both the US and UK, there were no tweets that included a low connotation name with any of the success attributes, and all of the high connotation names appeared with some of the success attributes.
  • Yet,  high connotation names featured at a much higher base rate overall compared to low connotation names, suggesting that less frequent names may cognitively be less associated with any personality characteristics.

Visit the iScience Maps website here.

Comments