10.01.2013 Views

Sampling Vegetation Attributes - Natural Resources Conservation ...

Sampling Vegetation Attributes - Natural Resources Conservation ...

Sampling Vegetation Attributes - Natural Resources Conservation ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

STUDY DESIGN AND ANALYSIS<br />

Wiegert (1962), summarized in Krebs (1989:67-72), gives a quantitative<br />

method for determining optimal quadrat size and/or shape. The method considers<br />

the costs involved in locating and measuring quadrats and the standard<br />

deviation (or its square, the variance) that results from samples of that size and<br />

shape. Refer to Krebs’ book for details (and an example).<br />

c Sequential <strong>Sampling</strong> The estimate of the standard deviation derived through<br />

pilot sampling is one of the values used to calculate sample size, whether one<br />

uses the formulas given in Technical Reference, Measuring & Monitoring Plant<br />

Populations, or uses a computer program.<br />

When conducting the pilot sampling, employ sequential sampling. Sequential<br />

sampling helps determine whether the examiner has taken a large enough pilot<br />

sample to properly evaluate different sampling designs and/or to use the standard<br />

deviation from the pilot sample to calculate sample size. The process is<br />

accomplished as follows:<br />

Gather pilot sampling data using some arbitrarily selected sample size. Calculate<br />

the average and standard deviation for the first two quadrats, calculate it again<br />

after putting in the next quadrat value, and continue these iterative calculations<br />

after the addition of each quadrat value to the sample. This will generate a<br />

running average and standard deviation. Look at the four columns of numbers on<br />

the right of Figure 5 for an example of how to carry out this procedure.<br />

Plot on graph paper (or use a computer program) the sample size versus the<br />

average and standard deviation. Look for curves smoothing out. In the example<br />

shown in Figure 5, the curves smooth out after n = 30-35. The decision to stop<br />

sampling is a subjective one. There are no hard and fast rules.<br />

A computer is valuable for creating sequential sampling graphs. Spreadsheet<br />

programs such as Lotus 1-2-3 allow for entering the data in a form that can later<br />

be analyzed while at the same time creating a sequential sampling graph of the<br />

running average and standard deviation. This further allows the examiner to<br />

look at several random sequences of the data before deciding on the number of<br />

sampling units to measure.<br />

Use the sequential sampling method to determine what sample size not to use<br />

(don’t use a sample size below the point where the running average and<br />

standard deviation have not stabilized). Plug the final average and standard<br />

deviation information into the appropriate sample size equation to actually<br />

determine the necessary sample size.<br />

19

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!