QF Short Course: Sergey Frolov

Date and Time

This short course is offered to registered attendees only.

 

Title: "How to Tell if a Paper I am Reading is Legit?"

Instructor: Professor Sergey Frolov, University of Pittsburgh

Abstract: Despite belonging firmly in natural sciences, known for their mathematical rigor, a lot of the quantum experimental papers in their essence are just essays illustrated by figures. The interpretation of figures is often qualitative. Figures are chosen to best represent the hypothesis that the authors prefer. In particular, these works often lack a statistical dimension that is central to many other fields of research. There are valid reasons behind this --- for example, when interesting phenomena are found in unicorn samples that are hard to make and reproduce. The consequence of this is that the only means to assess the validity of claims comes from whether they were published in a trusted peer-review journal. But in reality, neither the journals nor their referees have reliable means to evaluate conclusions, due to limited access to data. They typically just resort to trust in the authors' diligence.

While I don't have the ultimate foolproof method of verifying any paper, I will take you through several basic considerations - using real-world examples. This two-part minicourse will focus on claims assessment in the field of quantum transport experiments. However, the core principles and the ways of thinking should be applicable to a variety of fields. As one example, we will study the case of disappearing and reappearing resonances known as Shapiro steps, and what deep conclusions can be gleaned, and overturned, based on this evidence. (I have been making Shapiro steps show up and vanish for a good 20 years!). We will talk about Majorana research and how striking patterns considered by many a smoking gun in the signal can nevertheless have other explanations. Beyond critically reading the literature, we will discuss ways of writing papers in a way that would make their assessment by peers more straightforward.