Overall average oven-dry moisture content was calculated to be 14.1%. Because of some knots, pitch pockets, heartwood and variable density within measuring sections, their oven-dry moisture content will inevitably vary from that found in the absolutely clear areas, where meter readings were made. Some differences noted in the table of overall averages below include the results of such variation as well as other experimental errors:
| Delmhorst meter readings | Wagner meter readings | ||||||||
| OD MC (%) | Sou. pine (%) | Lnglf. pine (%) | Shrtlf. pine (%) | SYP (%) | Lob. pine (%) | Lnglf. pine (%) | Shrtlf. pine (%) | Slsh. pine (%) | Approx. sp. gv. |
| 14.10 | 15.3 | 15.0 | 15.1 | 11.8 | 12.7 | 10.8 | 12. | 10.8 | 0.53 |
As can be seen, no calibration of either meter produced overall average readings equal to the overall average oven-dry moisture content. Tests for determining differences between paired values, indicated all meter readings were statistically significantly different from oven-dry moisture content values. However, the Delmhorst averages were all closer to oven-dry values than the Wagner averages.
With only one moisture level examined, the range in moisture content values was relatively small. Therefore, regression equations relating initial meter readings to oven-dry moisture content are somewhat irrelevant. However, the range in wood density (specific gravity) for the sections examined was large, and regressions indicating the relationship between those values and moisture meter readings are found in the table below:
| R Squared | 0.0925 | ||||
| Delmhorst meter regression equation | |||||
| Specific gravity vs. Southern pine calibration readings | Specific gravity vs. Longleaf pine calibration readings | Specific gravity vs. Shortleaf pine calibration readings | |||
| R Squared | 0.38 | R Squared | 0.39 | R Squared | 0.26 |
| Wagner meter regression equation | |||||
| Specific gravity vs. Southern pine calibration readings | Specific gravity vs. Loblolly pine calibration readings | Specific gravity vs. Longleaf pine calibration readings | |||
| R Squared | 0.78 | R Squared | 0.78 | R Squared | 0.76 |
| Specific gravity vs. Shortleaf pine calibration readings | Specific gravity vs. Slash pine calibration readings | ||||
| R Squared | 0.76 R | Squared | 0.77 | ||
Although specific gravity accounts for almost none of the variation in oven-dry moisture content, it appears to explain about 1/4 to 1/3 of the variation in Delmhorst meter readings, which is not much. However, as has been shown before, specific gravity accounts for a large amount of variation in Wagner meter readings (more than 75%).
Examining differences between individual meter readings and oven-dry moisture content produced results seen in the following table:
| Delmhorst meter readings--deviation from OD MC | Wagner meter readings--deviation from OD MC | |||||||
| Sou. pine | Lnglf. pine | Shrtlf. pine | SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | |
Ave. of absolute diffs.  |
1.2 | 1.1 | 1.2 | 2.6 | 1.9 | 3.4 | 1.9 | 3.4 |
| Standard deviation |
1.0 | 1.0 | 1.1 | 1.2 | 0.9 | 1.3 | 1.0 | 1.3 |
| Straight averages |
1.2 | 0.9 | 1.0 | -2.3 | -1.4 | -3.3 | -1.4 | -3.3 |
| Max. values |
5.0 | 4.8 | 6.1 | 5.2 | 4.2 | 5.9 | 4.2 | 5.9 |
| Min. values |
0.0 | 0.0 | 0.0 | 0.2 | 0.1 | 0.2 | 0.1 | 0.2 |
As was obvious in the first table, average differences between meter readings and oven-dry moisture-content values, higher or lower, were considerably greater for the Wagner meter (1.9-3.4) than for the Delmhorst meter (1.1-1.2). However, the variation in differences was quite similar for both meters (1.0 to 1.1 for Delmhorst, 0.9 to 1.3 for Wagner). Looking at simple averages, it is clear that Delmhorst readings are usually greater than oven-dry values determined for the specimens, and that Wagner readings are usually less. Maximum and minimum differences (higher or lower) between meter readings and oven-dry values are similar for both meters.
The Delmhorst meter readings, restricted by the pin spacing to small clear areas of wood, may be closer to clearwood moisture content than the oven-dry calculations made on the entire measuring section. If so, a different analysis of results may be in order. Using the average of Delmhorst longleaf and shortleaf calibration readings (which were statistically no different from one another) as the ``standard'' moisture content, the following table summarizes the results:
| Assumed standard OD MC | Delmhorst meter readings | Wagner meter readings | ||||||
| (%) | Sou. pine (%) | Lnglf. pine (%) | Shrtlf. pine (%) | SYP (%) | Lob. pine (%) | Lnglf. pine (%) | Shrtlf. pine (%) | Slsh. pine |
| 15.1 | 15.3 | 15.0 | 15.1 | 11.8 | 12.7 | 10.8 | 12.7 | 10.8 |
| Delmhorst meter readings--deviation from OD MC | Wagner meter readings--deviation from OD MC | |||||||
| Sou. pine | Lnglf. pine | Shrtlf. pine | SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | |
| Ave. of absolute differences |
0.3 | 0.2 | 0.2 | 3.3 | 2.4 | 4.3 | 2.4 | 4.3 |
| Standard deviation |
0.5 | 0.4 | 0.4 | 1.4 | 1.2 | 1.3 | 1.2 | 1.3 |
| Straight averages |
0.2 | -0.0 | 0.0 | -3.3 | -2.3 | -4.3 | -2.3 | -4.3 |
| Max. values |
4.3 | 4.0 | 4.0 | 6.0 | 5.1 | 6.9 | 5.0 | 6.8 |
| Min. values |
0.0 | 0.0 | 0.0 | 0.1 | 0.0 | 0.6 | 0.0 | 0.7 |
This analysis obviously makes the Delmhorst readings look better in all respects than before, and the Wagner readings look worse. Maximum deviations from the assumed standard oven-dry moisture content are still greater than desirable, for both the Delmhorst as well as the Wagner meters. However, most of these problem readings can be explained by the presence of knots, resin pockets, etc.
Another way to evaluate the performance of the Wagner meter is to ``recalibrate'' readings from the various factory-supplied calibrations, and look at resulting average and maximum deviations from oven-dry moisture content values. The table below illustrates results of this procedure:
| Wagner meter readings--deviation from ovendry moisture content | ||||||||||
| SYP | SYP adj. calib. | Lob. pine | LOB adj. calib. | Lnglf. pine | Lnglf. adj. calib. | Shrtlf. pine | Shrtlf. adj. calib. | Slsh. pine | Slsh. adj. calib. | |
| Ave. | -2.35 | 0.00 | -1.39 | -0.00 | -3.33 | -0.00 | -1.36 | -0.00 | -3.32 | -0.00 |
| Ave. ABS | 2.57 | 1.19 | 1.85 | 1.19 | 3.38 | 1.06 | 1.85 | 1.23 | 3.39 | 1.11 |
| Max. ABS | 5.22 | 4.26 | 4.22 | 4.30 | 5.92 | 4.34 | 4.22 | 4.37 | 5.92 | 4.43 |
| Min. ABS | 0.18 | 0.01 | 0.09 | 0.01 | 0.17 | 0.02 | 0.14 | 0.00 | 0.17 | 0.01 |
(ABS--the absolute values, or the total deviation above or below the standard)
For each set of meter readings, individual readings were ``adjusted'' by the amount the average deviation was above or below the standard oven-dry value. For instance, for the SYP calibration, each meter reading had 2.35 added to it before being compared to the oven-dry standard moisture content value. For the loblolly pine calibration, each meter reading had 1.39 added to it before the comparison was made, and so on. The ``adj. calib.'' summaries show that in all cases, the average absolute deviation from the standard oven-dry value dropped greatly, and the maximum deviations were little changed. In fact, adjusted average absolute deviations become very similar to the deviations for the Delmhorst meter readings.
Although testing the Wagner ``new'' calibrations on data from which they were developed doesn't prove anything, using them separately on data from the North, middle and South Georgia mills may be informative. The tables below illustrate those evaluations:
| orig. deviations of meter readings OD MC | deviations or recalibrated meter readings from OD MC | ||||||||||
| SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | ||
| North Georgia Mills | Ave. | -2.82 | -1.89 | -3.85 | -1.86 | -3.79 | -0.47 | -0.50 | -0.52 | -0.50 | -0.47 |
| Ave. ABS | 2.91 | 2.09 | 3.85 | 2.08 | 3.83 | 1.04 | 1.10 | 0.86 | 1.09 | 0.95 | |
| Max. | 4.91 | 4.01 | 5.81 | 4.11 | 5.71 | 3.71 | 3.55 | 2.48 | 3.92 | 3.78 | |
| orig. deviations of meter readings OD MC | deviations or recalibrated meter readings from OD MC | ||||||||||
| SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | ||
| Middle Georgia Mills | Ave. | -2.64 | -1.66 | -3.52 | -1.58 | -3.54 | -0.29 | -0.27 | -0.19 | -0.22 | -0.22 |
| Ave. ABS | 2.77 | 1.88 | 3.59 | 1.87 | 3.62 | 1.09 | 1.06 | 1.00 | 1.17 | 1.03 | |
| Max. | 5.22 | 4.22 | 5.92 | 4.22 | 5.92 | 4.26 | 4.30 | 4.34 | 4.37 | 4.43 | |
| orig. deviations of meter readings OD MC | deviations or recalibrated meter readings from OD MC | ||||||||||
| SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | SYP | Lob. pine | Lnglf. pine | Shrtlf. pine | Slsh. pine | ||
| South Georgia Mills | Ave. | -1.59 | -0.63 | -2.62 | -0.64 | -2.63 | 0.76 | 0.76 | 0.71 | 0.72 | 0.69 |
| Ave. ABS | 2.02 | 1.59 | 2.69 | 1.61 | 2.72 | 1.44 | 1.42 | 1.32 | 1.43 | 1.33 | |
| Max. | 4.32 | 3.31 | 5.22 | 3.52 | 5.22 | 3.94 | 3.98 | 3.85 | 4.05 | 3.94 | |
It's obvious that even using the overall average ``recalibration'' figures, applying them separately to the North, middle and South Georgia lumber improves accuracy of the meter readings.
Finally, when data from the seven reexamined test pieces were included in regression equations demonstrating relationships between meter readings and ovendry moisture content the following r-square values resulted:
| Regression Output: | Regression Output: | ||
| Delmhorst Sou. pine calibration readings vs. oven-dry MC | Wagner SYP calibration readings vs. oven-dry MC | ||
| R Squared | 0.754588 | R Squared | 0.469703 |
Even the minor expansion of the range of moisture content examined (from the original average of 14.1% for 90 pieces down to 6.6% for seven pieces), immediately increased the size of the r-square values. Wagner meter readings obviously do not account for nearly as much variation in moisture content as Delmhorst readings do, again suggesting the need for special handling of Wagner readings for individual-piece moisture content.