Unit 3: Processing Data

The same data can be used to tell different stories. As an example, consider the following two expressions. One: the majority of participants (51%) voted to support the resolution. Two: nearly half of the participants (49%) opposed the resolution. These two statements are indeed based on the same data, but they might communicate very different messages. In the research world, most of the time the “raw data” cannot be directly used to support or refute an argument: it has to be processed, converted, and analyzed using particular procedures, techniques, methods, and theories.

 

Is the glass half empty or half full?

Figure 1 “It Depends…”
(Half empty Half full By Ravikallianpur / CC BY-SA 4.0)

 

Overview

This unit explores some ethical questions in processing and representing data to the research community and to the public. In particular, we probe the strengths and limitations of statistical analysis, different types of falsification, as well as ways in which explicit and implicit biases impact the representation and interpretation of research data.

 

Learning objectives

After successfully finishing this unit, you will be able to:

  • Explain statisticians’ ethical responsibility in data analysis;
  • Recall the Office of Research Integrity’s definition of research misconduct;
  • Give examples in which conflict of interest impacts the representation of research findings;
  • Recognize the need to reflect on the researcher’s own interests and biases.