Writing Analytically, 6th Edition - Rosenwasser, David & Stephen, Jill [95]
The Role of Context in Interpreting Numerical Data
In the previous example, the writer chose an interpretive context that seemed to best explain a pattern of detail in his evidence. In order to make a case for his interpretation, the writer needed to demonstrate the appropriateness and relevance of his chosen context, including his reasons for choosing one possible interpretive context over another.
The process is similar when a writer seeks to interpret numerical data. The writer must decide the extent to which his or her numerical data confirm or fail to confirm an expectation defined in the study’s hypothesis. Here is a brief example of statistical analysis from a political science course on public opinion research. The study uses a data set generated to test the hypothesis that “Republican defectors who have been members of the party for over 11 years are less likely to change party affiliation to Democrat because of the Republican Party’s policies than Republican defectors registered with the party for under 10 years.” Note how the writer integrates quantitative data into her discussion of the findings, a move characteristic of interpretation in the social sciences, and how she establishes the context in which this data might be best understood. Notice too how her findings complicate her original research question.
The data suggest that the longer a Republican defector was a member of the Republican Party, the more likely that person was to switch party affiliation to Democrat because of the Republican Party’s policies as opposed to changes in his or her own belief system. For example, 35% of Republican defectors who had been members of the party for 1-5 years agreed with the statement ‘the Republican Party’s policies led me to leave the party,’ while 35% said it was due to changes in their personal beliefs. Thus, it appears as though both reasons have equal influence on an individual’s decision to switch parties.
However, when you look at the defectors who were members of the party for over 6 years, roughly 20% more of them left because of the party’s policies than because of a change in their personal beliefs. This suggests that people don’t change their views—it is the party’s change of views that prompts defection by even long-time members.
In the case of statistical data, an interpretive problem arises when writers attempt to determine whether a statistical correlation between two things—blood cholesterol level and the likelihood of dying of a heart attack, for example—can be interpreted as causal. Does a statistical correlation between high cholesterol levels and heart attack suggest that higher levels of cholesterol cause heart attacks, or might it only suggest that some other factor associated with cholesterol is responsible? Similarly, if a significantly higher percentage of poor people treated in hospital emergency rooms die than their more affluent counterparts, do we conclude that emergency room treatment of the poor is at fault? What factors, such as inability of poor people to afford regular preventive health care, might need to be considered in interpretation of the data? (For more on interpreting numerical data, see “Interpreting the Numbers: A Psychology Professor Speaks” in Chapter 8, Reasoning from Evidence to Claims.)
INTENTION AS AN INTERPRETIVE CONTEXT
An interpretive context that frequently creates problems in analysis is intention. People relying on authorial intention as their interpretive context typically assert that the author—not the work itself—is the ultimate and correct source of interpretations.
FIGURE 6.1 The Dancers by Sarah Kersh. Pen-and-Ink Drawing, 6″ × 13.75″
©The Dancers, by Sarah Kersh. Pen and ink drawing, 6″ × 13.75″.
Used by Permission of Sarah Kersh.
Look at the drawing titled The Dancers in Figure 6.1. What follows is the artist’s statement about how the drawing came about and what it came to mean to