To prepare for tomorrow’s round table discussion (Wed Sept 2, 11am ET) on FASRI’s Financial Statement Presentation research, take a look at this report on an experiment I conducted with Frank Hodge, Pat Hopkins and Kristi Rennekamp.  A few points I’d like to make off the bat:

  • I fully expect this research to result in a submission to a peer-reviewed outlet like The Accounting Review or Journal of Accounting Research.  However, this is not that paper.  Instead this is a report to the FSP Project Team, and ultimately the FASB and IASB.  As a result, its style is derived as much as possible from what this audience is used to reading:  internal staff memorandums.  This was a challenge, as I wanted to convey rigorous research results without requiring the staff and Board to take a course in applied statistics, and I wasn’t able to find a similar report to use as a template.
  • One example of how this report differs from a standard research paper: rather than reporting every statistical comparison indicating the p-value and the nature of the test (e.g., “p < 0.034 2-tailed Wilcoxon text”), I describe the nature and meaning of statistically-reliable differences and p-values in a couple of paragraphs before reporting any results, followed by this statement:  “To reduce the risk that readers of this paper will inappropriately rely on a difference that is not statistically reliable, the text of this report describes differences in behavior across versions of financial statements, or differences of measured variables from 0, only if those differences are statistically reliable evidence as indicated by p-values.” This approach allows me to simplify the write-up quite a bit, while avoiding misleading readers by mentioning differences that researchers wouldn’t view as meeting their standards.  Note that I also go very light on the theoretical contributions of the experiment, which aren’t of much interest to standard setters.
  • The manuscript the research team submits to a peer-reviewed journal will be co-authored by all of us.  This particular report is just written by me.  Naturally, it wouldn’t have been possible without a great deal of effort by Frank, Pat and Kristi.  I want to give them credit for everything they did, while shielding them from (1) work that isn’t directly resulting in a peer-reviewed publication, and (2) any implication that they are responsible for oversimplifications that might have crept into this report as I recast our study into standard-setter-friendly terms.

Join us tomorrow to discuss what its like to work with standard setters on research that serves two masters (standard setters and academic peer reviewers).   Details on participating in Round Tables are here.