CHAPTER 2 



The Summarization of Sets 

 of Data Involving One Type 

 of Measurement 



Whenever a statistical investigation is to be made, two initial steps 

 must be taken: (a) A group of objects (persons, plants, bolts, or any- 

 thing capable of being measured) is specified as the subject to be 

 studied. (6) A decision is made regarding the feature of these 

 objects that is to be measured numerically or by some qualitative 

 designation. Such a set of measurements is called a population if it 

 includes every member of the group to be defined in a. For example, 

 suppose that an economist proposes to study the net cash incomes 

 of beef-cattle ranchers in Kansas during the decade from January 

 1, 1944, to January 1, 1954. It would be necessary first to define 

 the group of ranchers to be included in this study. How many beef 

 animals must he raise? Must the raising of beef cattle be his major 

 source of income according to some standard? Are absentee owners 

 included? There are many other matters which would have to be 

 considered. When a specific group of Kansas ranchers has been 

 defined, part a above has been completed. 



Next it is necessary to decide upon the specific meaning of the 

 term, net cash income. Is the measurement to be on a per-animal 

 basis, or the total for the ranch regardless of its size? Is any ad- 

 justment to be made for inflation, cost-of-living indexes, and the 

 like? When net cash income has been defined specifically, part b 

 listed above has been completed, and the population is defined. 



In some situations it is feasible to obtain every possible one of 

 the measurements in a population, as would be the case if every 

 beef-cattle rancher in the group discussed above were to be inter- 

 viewed and his net cash income determined according to the defini- 

 tion adopted by the investigator. Under these circumstances, the 

 purpose of statistical analysis is to summarize the information in the 



9 



