The purpose of this data essay is to study what determines whether a violent act gets covered in international media. Your task is to evaluate two di erent hypotheses that might explain why some events get news coverage while others go unnoticed.
Focus on a professional presentation of your results (for example, meaningful and profes- sionally looking tables and publication-quality graphs). Include a write up of no more than 2000 words ( 10%), font size 12pt, double-spaced with at least 2.5cm margins on all sides. Tables and gures do not count towards the word count (please only include one table/ gure per page, appropriately named and referenced in the text). Essays not adhering to these formatting guidelines will be marked down, as will sloppy presentations of tables and gures.
Apply in this essay what you have learned in this class about the statistical techniques, the interpretation of your estimates, and the presentation of statistical results. Do not include a theoretical discussion of violent events and news coverage. Instead, focus on the interpretation of the results and the description of the models. Keep in mind that there are di erent ways how to present results; try to choose the one that conveys the most informa- tion to the reader in an understandable way. Think about this data essay as the methods, data, and results" section of a research paper that you would like to submit to a journal for publication. Please keep in mind that absolutely no collaboration with others is allowed.
The authors of a recent study found that the probability of an act of rebel violence to get covered in international media depends on the coverage of the mobile phone network in the area where the event took place.
To test their argument the authors need data on violent acts which does not rely on news reports. (If researchers relied on data from news reports this would lead to measurement error in the dependent variable that is correlated with the main independent variable.) For their dependent variable they need data on all the events that eventually got reported
Quantitative Methods in Political Science
in international media. For their main independent variable they also need information on the mobile phone network in the area of the violent events. Furthermore, they iden- ti ed some control variables that might determine whether an event of rebel violence gets reported. First, the severity of an attack; the number of total casualties probably also in uences the likelihood of international news coverage. Similarly, the proportion of western casualties of the total casualties might as well in uence the international news coverage. Second, the level of remoteness of the areas where the events happened could be driving international news coverage. The authors o er three di erent operationalizations of remoteness.
With such a dataset they want to test two hypotheses.
First, they hypothesize that the coverage of the mobile phone network makes the reporting of rebel violence more likely, controlling for the possible other determinants.
Second, they suspect that the e ect of mobile phone coverage depends on the level of remoteness. They speculate that the e ect of mobile phone networks on the probability of reporting a violent act in the international media should decrease for higher levels of remoteness.
Your task is to test the two hypotheses using appropriate models.
The dataset contains information on more than 1,200 events of rebel violence in one country. Each observation is one distinct act of rebel violence with information on the variables of interest. The dependent variable, whether an event was reported in interna- tional media, is coded 1 if was reported, and 0 if was not. The codebook below contains information on all variables in the data set.
Quantitative Methods in Political Science
Use descriptive statistics and variable descriptions tofamiliarizethe reader with the data that you use. Note that you do not have to use all the variables that we provided you with. However, do think hard about which ones to use and why.
Use appropriate models to empirically test the eectof mobile phone networks on the probability of international news coverage of rebel violence. Discuss and comparethe eectsfor your main independent variables in terms of signs, size, and statisticalsignicance.