These methods construct a forecasting logic through a process of identifying the factors that cause some effect on the forecast and building a functional form of the relationship between the identified factors. In other words, a set of independent variables are identified and associated with the dependent variable through a functional relationship. For example, let us consider the demand in the country for a new product such as Direct to Home receivers (DTH). Since this is a new product, we may not have adequate past data on the demand and may need other means of establishing the potential demand. Even in the case of existing product, the number of factors that influence demand may be several requiring us to understand interaction among these, Several factors – including exchange rate fluctuation, installed capacity in the country, new product launches customers tariffs and price of raw material at the international markets—influence the demand. Forecasting in these situations uses casual methods.

In general, let us consider the forecast for a dependent variable Y using n independent variables X1, X2, X3, … Xn. Then developing a forecasting logic requires establishing a establishing as follows: Y= f(X1, X2, X3, … Xn)

Use of casual method to extract the trend component in times series is a frequent application of casual method. Other casual methods include econometric models, multiple regression models and technological forecasting techniques.

Casual methods of forecasting require greater degree of mathematical treatment of data. There are several computer packages such as SPSS available today to help the forecast designer in this process

**Example:** A manufacturer of tricycles in the age group of two to four years commissioned a market research firm to understand the factors that influence the demand for its product. After some detailed studies, the market research firm concluded that the demand is a simple linear function of the number of newly married couples in the city. Based on this assumption, build a causal model for forecasting the demand for the product using the data given below for a residential area in the city Also estimate the demand for tricycles if the number of new marriages is 150 and 250

X | Y |

New marriages | New marriages |

200 | 165 |

225 | 184 |

210 | 180 |

197 | 145 |

225 | 190 |

240 | 169 |

217 | 180 |

225 | 170 |

**Solution:** Since the causal relationship is a simple linear regression the method of least squares is used to determine the coefficient of linear regression Y= a + b

New marriages | Demands for tricycles | ||

X | Y | X*Y | X*X |

200 | 165 | 33,00 | 40,000 |

235 | 184 | 43,200 | 55,225 |

210 | 180 | 37800 | 44100 |

145 | 197 | 28,565 | 38,809 |

225 | 190 | 42,750 | 50,625 |

240 | 169 | 40560 | 57600 |

217 | 180 | 39060 | 47089 |

225 | 170 | 38250 | 50,625 |

Sum: 1749 | 1383 | 303,225 | 384,073 |

Average 216-625 | 172.875 | – | – |

From the equation b= $X1Y1-nXY

$X1*X1-nX*X

a =Y-bX

We have b=303,225-(8*218,625*172.875) =0.5104

384,073-8*218,625*218,625

a= 172.875-0, 5104 ore the demand for tricycles is given by relationship

Number of tricycles demanded= 61.29+0.5104 *no. of new marriages

If the no. of new marriages is 159 then the demand-138 tricycles

If the no. of new marriages is 250 then the demand = 189 tricycles