REV10162 - Laserfiche WebLink

Home Browse Search

Both years of data collected needs to be analyzed using the same methodology. The Division ran the adequacy statistics for 200] data using each production quadrat data as an individual value. Analyzing the 2001 production data in the same method as Greystone analyzed the 2002 data found the 2001 reclaimed pasture data to have been sampled to statistical adequacy, however, the 2001 reference area pasture did not meet sample adequacy using the same method of analysis. Rule 4. /5.8(4)(a) Requires that, "For herbaceous production the revegetated area shall be considered acceptable if it is: Not less than 90% of the production as determined from the reference area or approved standard with 90% statistical confidence. Fifteen transects lines with three production quadrats per line were collected in July 2001 irrigated pasture. This data achieved statistical sample adequacy. 2001 irrigated pasture reference area achieved sample adequacy with twelve transect lines with three production quadrats collected per transect. 2001 Dry production weight on the irrigated pasture reference area was 50.3 g/0.5 Mz. Productivity success required on the irrigated pasture for the 2001 data is ninety percent of this value, 45.3 gf0.5 MZ. Herbaceous productivity measured on the irrigated pasture in July 2001 was 137 g/0.5 M2. In 2001 productivity on the reclaimed irrigated pasture far exceeded productivity on the reference azea pasture. Since statistical sample adequacy was not achieved on the 2002 productivity data, the Division used the reverse null hypothesis testing to determine reclamation success. The "reverse null" hypothesis presumes that the sample (reclaimed area) mean is less than or equal to 90% of the standard and thus unsuccessful. Calculation of the reverse null hypothesis: "The reclaimed area mean is less than or equal to 90% of the reference area mean." was proven to be a false statement indicating that the reverse is true, "the reclaimed area mean is greater than 90% of the reference azea mean." Thus, using the reverse null hypothesis, the Division can state that the data supports that the reclamation is successful. The "reverse null" hypothesis test is a more stringent test because the assumption is that the reclamation has not been successful. In this situation, it is in the operator's interest to design a powerful test so as to detect a true difference between the reclaimed area parameter and the standazd, if one exists, in order to reject the null hypothesis. For this reason, no statistical sample size adequacy demonstration is required, as long as the minimum sample size as specified in the bond release guidelines is met. In situations where the sample mean exceeds the standard by a substantial margin, it may be more efficient for an operator to demonstrate success by employing a "reverse null" hypothesis test, than by obtaining the number of sample observations necessary to meet statistical sample adequacy. For normally distributed data, a minimum sample size of 15 is recommended. For sampling data that is not normally distributed, a minimum sample size of thirty should be collected. For 2002 irrigated pasture reclamation azea Greystone collected 50 samples. Although the data did not achieve sample adequacy application of the "Reverse Null" hypothesis demonstrated that reclamation success was achieved and this more stringent test eliminates the requirement to achieve statistical sample adequacy since the minimum sample size was achieved.