Survey Techniques

Surveys and opinion polls are often used to make short term forecasts when quantitative data are not available. These qualitative methods can also be very useful for supplementing quantitative forecasts that anticipate changes in consumer tastes or business expectations about future economic conditions. That can also be useful in forecasting the demand for new product that the firm intends to introduce.
Consumer Survey Methods
Consumer are asked what quantity of the product, they would be willing to buy under various conditions such as price and income levels. This method is known as direct interview method. This method may cover almost all the potential consumer’s or only selected groups of consumers from different cities or parts of the area of consumer concentration. When all the consumers are interviewed, the method is known as complete enumeration and when only a few selected representative consumer’s are interviewed, it is known as sample survey method. In case of industrial inputs, interviews or postal inquiry of only end users of a product may be required.
Regression Analysis
Regression analysis is most popular statistical techniques of demand estimation. Economic theory is employed to specify the determinants of demand and to examine the nature of relationship between the demand for a product and its determinants. In regression method of demand forecasting the firm estimates the demand function for a product. In the demand function, quantity to be forecasted is “dependent” variable and all other variables that affect the demand are called “independent” or “explanatory” variables. For example, demand for meat in Kathmandu may be said to depend largely on ‘per capita income’ of the city and its population. Here demand for ‘meat’ is a ‘dependent variable’ and ‘per capita income’ and population are the ‘explanatory’ variables.
While determining the demand function for particular commodity, the analyst may come across many commodities whose demand depends on a single independent variable. For example, suppose, in a city, demand for items like salt, gas, sugar, is found to depend on the population of the city then demand functions for such commodities are single variable demand functions. On the other hand if analyst find that demand for vegetables, fruits, beer, etc. depends on a number of variables like commodity’s own price, price of its substitutes, household income, population, customs, habits etc. such demand function are called Multi-variable demand function. Simple regression equation is used for single variable demand functions and multi-variable regression equation is used for multiple variable regression are explained below.
Simple Regression (Single Variable)
In simple regression method, a single independent variable is used to estimate a statistical value of the dependent variable which is to be forecasted. This technique is similar to trend fitting (least square) method. A key difference between trend fitting and regression method is that, in the case of trend fitting, independent variable is time (t) where as in regression equation the chosen independent variable is the single most important determinant of demand.
Multi- Variable Regression
Liner Function: The multi variety regression equation is employed for cases in which number of explanatory variables is greater than one.
The first step here is to identify the determinants of the variable to be forecasted. For example, for estimating demand for capital goods the relevant variables are additional corporate investment, rate of depreciation, cost of capital goods, cost of labor and raw materials, market rate of interest, etc. Similarly in forecasting the demand for breakfast cereals, the relevant variables are price of the breakfast cereals (p), consumer s disposable income (Y), the size of the population (pop), price of substitutes (Ps), the price of milk (Pc – a compliment), level of advertising by the firm (A). Once independent variables are specified and necessary data are collected, the next step is to0 specify the form of equation.
Power Function: In linear equation the marginal effect of independent variables on demand is assumed to be constant and independent of change in other variables. But there may cases in which the marginal effects of each independent variables as well as on the value of all other variables in the demand function.
Simultaneous Equations or Multiple Equation Method
Although single equation models (liner equation) are often used by firms to forecast demand or sales, economic relationship may be so complex that a multiple-equation model may be required. This is particularly used in forecasting macro variables such as GNP or the demand and sales of major sections or industries. Multiple-equation models may consist only a few equations or hundreds of equations.