30.06.2013 Views

Introduction to Health Physics: Fourth Edition - Ruang Baca FMIPA UB

Introduction to Health Physics: Fourth Edition - Ruang Baca FMIPA UB

Introduction to Health Physics: Fourth Edition - Ruang Baca FMIPA UB

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

HEALTH PHYSICS INSTRUMENTATION 493<br />

The statistical error associated with a count ratemeter is a function of the meter’s<br />

time constant (response time) and is given by<br />

<br />

count rate<br />

σCRM =<br />

. (9.48)<br />

2 × time constant<br />

Note that the count rate and the time constant in Eq. (9.48) must be in the same<br />

time units.<br />

W Example 9.12<br />

A survey meter whose time constant is 15 seconds reads 400 cpm. What is the standard<br />

deviation of this measurement?<br />

Solution<br />

In this case, the time constant, 15 seconds, corresponds <strong>to</strong> 0.25 minutes. Therefore,<br />

from Eq. (9.48), we find<br />

<br />

400 cpm<br />

σCRM =<br />

=±28 cpm.<br />

2 × 0.25 min.<br />

The standard deviation is a measure of the dispersion of randomly occurring<br />

events around a mean. It was pointed out above that if a large number of replicate<br />

measurements were made, 68% of the measurements would fall between ±1σ of the<br />

mean. For this reason, we say that 1σ is the 68% confidence interval. Similarly, the<br />

96% confidence interval is ±2σ . If we report data within the limits of 2σ , it means that<br />

we are 96% certain that the true value lies within the given interval. Several levels<br />

of confidence, <strong>to</strong>gether with their corresponding number of standard deviations,<br />

are given in Table 9-9. The numerical value of a confidence interval represents<br />

the percentage of the area under the normal curve that is included between the<br />

corresponding number of standard deviations.<br />

W Example 9.13<br />

A preliminary measurement made during a short counting time suggested a gross<br />

counting rate of 55 cpm. The background counting rate, determined by a 1-hour<br />

measurement, is 25 cpm. How long should the sample be counted in order <strong>to</strong> be<br />

96% certain that the measured net counting rate will be within 10% of the true<br />

counting rate?

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

Saved successfully!

Ooh no, something went wrong!