The best way to learn about reproducibility is to actually do it. This exercise will teach you the basics of the Tier Protocol through the analysis of a dataset on racial attitudes in the United States. Our specific research question involves whether different racial and ethnic groups have different opinions about how much attention racial issues receive in the U.S.
The data we will use are provided by the Pew Research Center. Pew conducts regular surveys of American attitudes on a variety of topics. We will use their survey data from February 16th 2016. One of the questions they asked was, “In general, do you think there is too much, too little, or about the right amount of attention paid to race and racial issues in our country these days?”
The independent variable in this analysis is racial group. Social scientists know that race/ethnicity is a very difficult to operationalize because the concept is socially constructed. This means that how we as a society draw racial boundaries varies over time and place. Researchers must decide what question(s) they should ask to determine someone’s race/ethnicity and they also have to decide what categories to include in their analyses. In this exercise we will explore how Pew chose to operationalize race in their 2016 survey. We will learn how to recode the data to create what is (arguably) a better measure. Finally we will construct a cross tabulation to answer our research question. While we are doing this we will be recording everything we do: thus creating reproducible conclusions.
To start, do you have a hypothesis about what we will find? Do you think some groups are more likely than others to think there is too much or too little attention paid to racial issues? Why do you think that?