Describe the types of statistical unit in detail or write a short note on units of collection.
Describe the types of statistical unit in detail or write a short note on units of collection.
Ans.
Types of Statistical Units
The statistical units may be broadly classified as follows:
(i) Units of collection.
(ii) Units of analysis and interpretation.
(i) Units of Collection: The units of collection may further be sub divided into the following two classes :
(a) Unit of Enumeration : In any statistical enquiry, whether it is conducted by sample’ method or ‘census’ method, unit of enumeration is the basic unit on which the observations are to be made and this unit, is to be decided in advance before conducting the enquiry keeping in view the objectives of the enquiry. The unit of enumeration may be a person, a household, a family, a farm (in land experiments), a shop, a live stock, a firm etc. As has been pointed out earlier, this unit should be very clearly defined in terms of shape, size etc. For instance, for the construction of cost of living index number, the proper unit of enumeration is household. It should be explained in clear terms whether a household consists of a family comprising blood relations only or people taking food in a common kitchen or all the persons living in the house or the persons enlisted in the ration card only. The concept of the household (to be used in the enquiry) is to be decided in advance and explained clearly to the enumerators so that there are no essential omissions or irrelevant inclusions.
(b) Units of Recording: The units of recording are the units in terms of which the data are recorded or in other words they are the units of quantification. For instance, in the construction of cost of living index number (consumer price index) the data to be collected from each household, among other things, include the retail prices of various commodities together with the quantities consumed by the class of people for whom the index is meant. The units of recording for quantity may be weight (in case of food-grains), say, in l.ilograms, quintals, tons, etc., in case of clothing the unit of recording may be metres; the prices may by recorded in terms of rupees and so on.
Units of measurement (recording) may be simple or composite. The units which represent only one condition without any qualification (adjective) are called simple units such as metre, rupee, ton, kilogram, pound, bale of cloth, hour, week, year, etc. Such units are generally conventional and not at all difficult to define. However, sometimes care has to be taken in their actual usage. For example, the bale of cloth must be defined in terms of length, say 20 metres, 50 metres or 100 metres. Similarly, in case of weight it should be clearly specified whether it is net weight or gross weight.
A simple unit with some qualifying words is called a composite unit. A simple unit with only one qualifying word is called a compound unit. Examples of such units are skilled worker, employed person, ton-kilometre, kilowatt hour, man hours, retail prices, monthly wages, passenger kilometres. For instance ton-kilometre means the number of tons multiplied by the number of kilometres carried; man hours implies the total number of workers multiplied by the number of hours that each worker has put in and so on. If two or more qualifying words are added to a simple unit, it is called a complex unit such as production per machine hour output per man hour and so on. Thus as compared to simple units composite compound and complex) units are much more restrictive in scope and difficult to define. Such units should be defined properly and clearly as they need explanation about the unit used and also about the qualifying words.
(ii) Units of Analysis and Interpretation : As the name implies, the units of analysis and interpretation are those units in the form of which the statistical data are ultimately analysed and interpreted. It should be decided whether, the results would be expressed in absolute figures or relative figures. The units of analysis and interpretation facilitate comparisons between different sets of data with respect to time, place or environment (conditions). Generally, the units of analysis are rates, ratios and percentages, and coefficients.
Rates involve the comparison between two hetrogeneous quantities i. e. when the numerator and denominator are not of the same kind e.g., the mortality (death) rates, the fertility (birth) rates and so on. Rates are usually expressed per thousand. For instance, the Crude Birth Rate (C.B.R.) is the ratio of total number of live births in the given region or locality during a given period to the total population of that region or locality during the same period, multiplied by 1,000. Rate per unit is called coefficient. However, ratios and percentages are used for comparing homogeneous quantities e.g., when the numerator and denominator are of the same kind. For example, “the ratio of smokers to non-smokers in a particular locality is 1 : 3” implies that 25% of the population are smokers.
From practical point of view for comparing data relating to different series, usually the unit of analysis is one which gives relative figures which are pure numbers independent of units of measurement. For instance, if we want to compare two series for variability the appropriate unit of analysis is coefficient of variation & for comparing symmetry of two distributions we study the coefficient of skewness.
