Originally Posted by Snipe
I have to enter my thoughts here..
We spend thousands upon thousands on electrofishing equipment and then measure and weigh in inches and pounds/ounces.
mm and grams are in order. It's been proven to me time after time that pounds/ounces/inches can create up to a 15% error in WR and bad assumptions are made because corners were cut on accuracy.
This is something I don't think is "good" in fisheries management, if it isn't as accurate as possible, how can you adjust or manage with faulty data????
Sorry guys, I'm OCD with this particular step.

Kenny, I agree that the degree of precision is very important. Like you I would prefer mm and grams because these units lend to very good degree of precision for fish that are > 5" in length. mm and grams are easy to read from scales and tapes.

That said. A similar degree of precision can be obtained if one has sufficiently sensitive measuring tools. For example, if a English unit tape is divided into 1/32 of an inch AND if the one taking the measurement doesn't take shortcuts of rounding then the degree of precision is actually just a tiny bit better than mm. Even so, a tape with mm is the hammer and the 1/32" scale is a rock. I would much rather drive a nail with hammer and than a rock because it is the most appropriate tool. So same goes for the tape. Your measurement will not need conversion into decimal notation and it is far easier and faster to take the measurement very accurately with a tape graded in mm.

A weight scale needs to have sufficient sensitivity to measure small increments of change. Some scales display a greater degree of precision than they are actually sensitive to. But if grams are the degree of precision sought, the scale must be sensitive to 1 gram increments. So this is easily checked (though most instruments declare sensitivity) by adding the increment to an existing weighed item and checking to see if the increment registers. If measuring in pounds on a decimal scale the scale must provide measurement in 1 thousandths of lbs and be sensitive to register .001 lb increments.

The problem with rounding and degree of precision is very much exacerbated with fish at the smaller end of the spectrum becoming more muted with large fish.

In looking over the LMB in the table, I think the surveyors may have attained a very acceptable degree of precision. The length measurements seem to be incremented to the nearest .05". So the tape was probably 1/10 graded with possible 1/2 ticks. Whether the tenths were half ticked or not, the data collected suggests that the measurements take were taken with care either to the finest degree of precision graded on the tape or possibly making additional effort to record precision finer than the tape graduation. The nearest .05 (1/20th) inch is a little larger than a millimeter by ~ 0.27 mm. So on the first fish, assuming a very accurate measurement to the nearest .05 inch the variation or uncertainty in relative terms is less than 9 parts in 10,000. Don't know if this alarmed you but the number of exactly 13" did catch my attention. With degree of precision at 0.05 this did seem a little odd but that doesn't of course mean that they were not 13" to the nearest 0.05"

On the weight end, increments of 0.01 lbs is about 4.5 times greater than grams. Assuming the sensitivity to 0.01 lbs the uncertainty of measurement relative to a gram sensitive instrument is about 4.5 parts in 1000 for first and lightest fish. The parts uncertainty for larger fish is less than that. All in all, there was sufficient precision here I think but one surely does not want to pull out a fish de-liar and round to the nearest quarter or half inch. This would yield results that are not very accurate.