This morning, just for fun, I ran a few statistical tests on the composition of the NOSB. First, I linked the year the person was appointed to the president in office at the time. Note that some error might have been introduced at this stage, because a board member’s term usually begins in January, and there could be a change in presidency at the exact time. Next I calculated the percent of the members whose board position matched their jobs for each administration (similar to what I posted yesterday). The table (sorry for the crudeness of its appearance) below shows the results.
Note this is a first pass, and thus is a rough approximation. Many factors influence the outcome of who is selected for a specific position on the board, and while Christina Bronsing and I are exploring this, we haven’t yet fully specified our models. In other words, we are still at the exploratory stage.
| Positions filled | Percent “good fit” | |
|---|---|---|
| Bush 41 | 13 | 38 |
| Clinton | 21 | 71 |
| Bush 43 | 24 | 46 |
| Obama | 16 | 56 |
I conducted pairwise t-tests because I was curious to see whether bush 41 differed from clinton, and so on. The tests indicate that under the Clinton administration, a statistically significant greater percent of the board members passed my definition of a “good fit” for their roles on the board, when compared to Bush 41 and Bush 43. There was no statistical difference to the goodness of fit of the appointees during the years of Clinton and Obama.
I also ran a few logistic regressions, with the dependent variable set to “1″ if the role on the board matched the person’s work, and “0″ otherwise. Preliminary results indicate that (see table below, with log odds output from logistic regression) board members with roots in the organic industry (vs conventional or sustainable worlds) were more likely to have a role commensurate with their work life. In addition, the statistical link between “big food” and the goodness of fit between job/board role was not statistically significant.
jobmatch Odds Ratio Std. Err. z P>z
bigfood 0.64 .35 -0.81 0.42
rootscode 4.07 2.07 2.76 0.01
Our next steps are to assess the effect of the NOP leadership on the board composition. So stay tuned!