Methods used in Forecasting
User Rating: 0 / 5
- Details
- Hits: 19590
FORECASTING METHODS
Forecasting is a deliberate action to determine a possible future scenario as Graf, (2002) states. The business world has become a competitive pool and so each competitor always seeks an edge over the other. Forecasting helps managers to make the right decisions that would be beneficial to the business in the future.
Shah, (2009) illustrates two main types of forecasting methods; Qualitative and quantitative methods. The method used depends on the following factors; level of detail, units of analysis, time horizon required, cost, ease of use and accuracy required. The qualitative method is more of opinion oriented hence it is usually subjective in most cases.
The quantitative method relies more on mathematical data and can be proved scientifically.
According to Graf (2002) ‘‘the qualitative method is divided into; Executive opinion, market research and Delphi Method``. In the executive opinion, a panel of managers come together and discusses among themselves the possible future scenarios and from these discussions they come up with forecasts. In the market research method, interviews are conducted to get consumer opinions on particular products and surveys are also used to determine trends. In the Delphi method, a panel of experts in that particular field comes together and using their knowledge and experience, they come up with future possible scenarios.
The quantitative method is composed of two major models; Time series model and casual model. The time series model uses past patterns to predict future events. Patterns within a specific period are analyzed and then an assumption is made that in the future similar patterns will occur with a few variations. The time series model contains elements such as naive, simple mean, moving average, weighted moving average and exponential smoothing. The casual model employs the concepts of action and reaction, it taken that if a particular occurrence leads to a particular situation then in the future when such an occurrence occurs then the same situation is expected. The casual model, also known as associative model, also employs the concept of indicators to predict future scenarios. It employs the concepts of linear regression and multiple regressions, which is by large, an extension of linear regression.