H-indices in PER

The h-index is a measure of scholarly productivity and impact for individual researchers.  Larger numbers are better, and typical numbers vary by field.  In PER, typical numbers are in the single- or low-double digits, but there's no official list of h-indices.   This is important because there's a document floating around stating that an appropriate h-index for physicists going up for tenure is somewhere in the teens, and that's wildly inappropriate for PER. 

H-index is pretty easy to calculate: a scholar's h-index the number of papers h which have h or more citations.  Google Scholar (and other systems like Scopus or Web of Science) will each calculate it.  The numbers differ each time because of the ways that the systems index papers and citations, so it's important to draw all your numbers from the same source, and to pick the right source for the right field.  

I spent an idle afternoon in late spring 2015 looking up h-indices for people in PER using Google Scholar, which has opt-in participation.  I also coded people's positions to find trends by seniority. 

Data Selection

• The top 100 researchers in PER measured by number of unique coauthors in PERC, PhysRev, and AJP, 1980-2013.  These are the top three journals in the field.

• Everyone on the Researchers in Physics Education listserv, which is for active researchers in PER, mostly in the US.

• Except anyone in any of the following categories:

  • Not ever in PER (e.g. chem ed folks, people who have never published in PER)

  • Not in a US-like position, but still in academia (non-academics and quasi-academics can stay; groups from countries with wildly different academic systems cannot).

  • People I haven't heard of and can't find easily by googling for their name or their name + physics.  There are not very many people in this list.

• This gives me a list of about 150 active researchers in PER, biased towards researchers who are more senior and have lots of co-authors.  That means that I am probably overestimating average h-indices.

Data reduction

• Looked up everyone on google scholar.  Only 53 researchers survived, so this might not be representative.

• Recorded h-index.  Google scholar lists "all time" and "last 5 years"; I picked the "all time" number.

• Because some people publish in multiple fields (Mazur, Pritchard, etc), I recounted everyone's h-index to only include their PER/education papers.  I did not make judgements as to the quality of the papers, only as to their field (HEP no; teaching people about HEP yes).  

• Remembered (about 2/3) or looked up (about 1/3) everyone's current position, and coded it as follows

  • f: Full professor or emeritus

  • d: associate professor, has been for a while

  • c: associate professor, recently tenured

  • b: assistant professor, nearing tenure

  • a: assistant professor, recently hired

  • z: postdoc, VAP, or other recently PhD'd

  • x: RAP

  • o: other, usually non-academics or non-TT/non-research people