40 Directory UMM :wiley:Public:college:statistics:Johnson:

Continuous variable: 2.1

2.2 40

CHAPTER 2 DESCRIBING PATTERNS IN DATA On the other hand, there are quantitative variables whose values can, in principle, take any value in an interval. A quantitative variable whose value can be any value in an interval To a reasonable approximation, net income, time to equipment failure, total sales, and weight are examples of continuous variables. In these cases, the measurement scale does not have any gaps. Ideally, any value along a continuum is possible. A truly continuous scale of measurement is an idealization. In practice, continuous measurements are always rounded either for the sake of simplicity or because the measuring device has a limited accuracy. However, even though weight may be recorded to the nearest pound or time to failure to the nearest minute, their actual values occur on a continuous scale so the data are referred to as continuous. Variables that are inherently discrete are treated as such, provided they take relatively few distinct values. When the values for a discrete variable span a wide range, however, they may be treated as continuous. The number of shares volume of stock traded per day is a discrete variable, but daily volume data may, for practical purposes, be viewed as continuous. The point to keep in mind is that, regardless of whether the variables char- acteristics of interest are naturally numerical, ultimately we will be dealing with numbers. Classify the following as enumerative or analytic studies. Justify your choice. a. A telephone company wishes to estimate the proportion of all telephones in a city that are working at a given time. b. An airport executive wants to know the number of on-time arrivals at a municipal airport yesterday. c. A construction company executive wants to estimate the amount of supervi- sory time for each worker-day of time allocated to its jobs. d. During the summer, a university administration wants to estimate the number of admitted freshmen who will attend school in the fall. Classify the following as enumerative or analytic studies. Justify your choice. a. A rating service wishes to estimate the number of households in the United States watching a particular Monday night television program. b. A company wants to determine the number of defective golf balls in a recently produced batch. c. A consumer products company wants to know whether increasing advertising expenditures will lead to increased sales of an item. d. A mail order firm wants to estimate the time it takes to ship the goods once an order is received. EXERCISES 4 4 4 4

2.4 GRAPHICAL DISPLAYS OF DATA DISTRIBUTIONS