ľ¹Ï¸£ÀûÓ°ÊÓ

What if the same data could lead to opposite conclusions? Science is supposed to be objective, but researchers’ choices in analyzing data can lead to very different results. Suzanne Hoogeveen illustrates this with projects on topics such as health, religion, and racial bias. The seminar dives deeper into analytic heterogeneity and how research findings can be made more trustworthy.

In a traditional research project, a researcher formulates a hypothesis, collects data, and conducts a single statistical analysis to test their predictions. However, a growing body of work shows that research outcomes are not always as straightforward: subjective and seemingly arbitrary decisions in the analysis process can strongly shape the conclusions we draw. Recent many-analysts and multiverse projects have highlighted this issue by exposing analytic heterogeneity: when different researchers analyze the same data set, they often make different choices and reach different conclusions. 

In this talk, Suzanne will share what these large-scale projects can teach us about the robustness of research findings. Using examples from studies on topics such as religiosity and well-being, the mortality salience effect in social psychology, and implicit racial bias,

Suzanne will discuss three key questions:

1. what do these projects reveal about how research works?
2. how should we understand and respond to analytic heterogeneity?
3. what practical steps can researchers take to make their findings more reliable?

Presenter 

Suzanne Hoogeveen, assistant professor Methodology & Statistics, Utrecht ľ¹Ï¸£ÀûÓ°ÊÓ. 

Registration

The seminar can be attended on location in Utrecht or online via Microsoft teams. Please register via the .