Frequency distributions provide an indication of the relative 

 magnitude of occurrence of all the measurements in an inspection survey. 

 A frequency distribution shows the number of observations from the data 

 set that fall into each measurement interval. The measurement interval 

 must be selected to be at least as large, as the accuracy of the measure- 

 ment tool and should provide meaningful intervals covering the entire 

 range of possible measurement values. Frequency distributions are 

 usually depicted by a histogram as shown in Figure 3. The frequency 

 distribution can be shown on either an absolute or relative scale. An 

 absolute distribution plots the actual number of occurrences for each 

 interval, while a relative distribution shows the percentage of occur- 

 rences compared to the total number of observations in each interval. 

 For ease of comparison of frequency distributions with other statistical 

 distributions, the relative distribution is generally used. 



Probability distribution functions (PDF) are theoretical frequency 

 distributions developed to deal with statistical expectations. These 

 distributions are commonly used to make inferences under conditions of 

 uncertainty. The normal (or Gaussian) distribution is the most common 

 of the probability distributions dealing with a continuous random variable, 

 Its properties make it useful in many situations in which it is necessary 

 to make inferences from sample data. This type of distribution has been 



fc o.i 



I 1 I I 



RTHn I I J— J 



10 12 14 16 



Diameter (in) 



Figure 3. Frequency distribution for timber pile diameters. 



