The Point of Reading Old Papers -- CAREFULLY
I suppose all of us, particularly husbands, have a handful of “this is how stupid I once was” stories. Here is one from my professional life.
A few years ago (about 1980), I came up with an idea that was eventually called “Big BAF” sampling. The use of a second prism was easy to imagine, since all of us have used more than one, and of the many ways to gather a subsample of Volume to Basal Area Ratio (VBAR) it was very effective and easy to check cruise.
We spent a lot of time looking for someone who had already done it. We asked reviewers, friends, experts in the field – no response. Nobody had ever heard of this simple technique that got twice as much benefit out of measured trees as previous methods had done. We checked out the literature, and nowhere could we find it.
Well, there came a day when John Bell cleaned out some of his many publications, and they came my way. Included was an original copy of Lew Grosenbaugh’s very early article “Plotless Timber Estimates -- New, Fast, Easy” published in the Journal of Forestry in 1952. I sat down by the fireplace in my library to read it yet again.
Now I have read this same article a number of times in the past. Lew was never a fan of the idea of using VBAR, preferring to measure all the trees on the plots - or at least use the very traditional equation for double sampling with “count plots” when it came to the statistics. Lew just hated to get snotty letters from university professors pointing out that the very excellent estimate of Sampling Error known as “Bruce’s Method” (and now used by virtually everyone) was not exactly right from a theoretical viewpoint. He knew that. The average did not change in the slightest – only the statistics, and those were virtually identical either way. Of course the assumptions for the usual formulas for those statistics was also wrong, but that never bothers the tedious folks who send letters to people doing practical work.
I was soon exchanging emails with one of my coauthors Dave Marshall, and the gist of them was ...
Dave, read the 1952 article by LRG.
Yes, I have read it several times Kim. Great article. Always like it.
Dave, read it. Read it again. Now, please.
OK, I did that … great article, like I said.
Dave, read it again, page 35, left column.
Oh drat. How could we have missed that ?
And there it was. “An obvious adaptation is to use a larger critical angle to obtain a smaller sample of the average ratio, while using a smaller critical angle to obtain a larger sample of average basal area per acre“. As Dave put it so well – “we should have known that if it was a good idea, Grosenbaugh or Bitterlich would have thought of it already”.
So, why did Lew Grosenbaugh not suggest that it be used ? He was clearly aware of the idea of subsampling for the volume/basal area ratio, and probably of the advantages. A little known memo he sent to his boss P.R. Wheeler in 1949 was almost certainly the first suggestion for the idea of subsampling for “VBAR” to apply against the basal area of the stand from the simple count of trees.
I would guess that Lew just thought VBAR was too much of a stretch for people in the field. The psychology was bad and the timing was wrong. Asking people to use a magic angle was already difficult to sell, and heaven knows there was great resistance to the idea of anything but fixed plots for many years. Using two prisms, and measuring trees for this strange conversion value of “BF/square foot” was probably a bridge too far. It was buried among the many ideas that one finds in any paper by Lew Grosenbaugh.
Oddly, 30 years later, we suggested two prisms for the same reasons – the psychology. It was easy to imagine and comfortable for people who were already using a prism in the woods to both count and sample trees. The second prism was easy to accept. There are, of course, many ways to choose a sub-sample of measured trees, and we have suggested some of these in the newsletter.
Perhaps we all should spend more time reviewing the ideas from the past. Many of them were not adopted because we did not have computers or databases at the time, or because the instruments of the day did not encourage it. The ideas were right – the timing was not. Many of these ideas seem to have composted in the piles of old research papers so they were independently invented later when the times and situation had changed. It’s hard to tell how many times this happens, but it is certainly useful to chat with the older members in the business. Many of their other good ideas never even ended up in print. The USFS has done a great job in scanning some of the old technical papers and making them available, but the Internet generally ignores most of that material.
So – I am now in the position of apologizing to a great biometrician who has already passed away. Although Lew was around when we first published this idea he never said a word. Perhaps he did not hear of it. Perhaps he was just too much of a gentleman to say anything. Perhaps in view of the hundreds of great ideas he had over the years it had slipped his mind (though it seems unlikely that anything would slip Lew’s mind).
The least we can do is give credit where credit is due. Since Lew did not suggest a name for the technique, I suppose the current “Big BAF” will stand (although the USFS always invents a name of their own for anything they adopt). It is now apparent that credit for the “Big BAF” method is entirely due to Lew Grosenbaugh, who was clearly the first to suggest it.
The Booby prize – well, now you know who that goes to.
Originally published October 2010
Regular Article Index