Comparison and Contrast of Forecast Methods

Comparison and Contrast of Forecast Methods

MGT 554

Operations Management

University of Phoenix

Professor Leonard Enger

May 1, 2006

TABLE OF CONTENT

Cover Page................................................................1

Table of Contents.........................................................2

Seasonal Forecasting.....................................................3

Delphi Method............................................................4

Technological Method...................................................5

Time-series forecasting...................................................6

Company Forecasting Methods.........................................7

Conclusion.................................................................8

References.................................................................9

Comparison and Contrast of Forecast Methods

There are several different methods that can be used to create a forecast, this paper will compare and contrast the Seasonal, Delphi, Technological and Time Series method of forecasting. Factors to consider when selecting a forecast method is the experience of the forecaster, the amount of information available, the degree of accuracy or confidence needed from the forecast and the level of difficulty that the situation present.

Forecasting is the process of estimating or predicting future events or conditions. Forecasts may be long-term or short-term. The techniques used may be quantitative or qualitative. Quantitative forecasting models may be classified into (a) causal models in which independent variables are used to forecast dependent variables, and (b) time series models, which produce forecasts by extrapolating the historical values of the variables of interest by, eg, moving averages. www.indiainfoline.com/bisc/accf.html

Seasonality Forecasting

Seasonality is often displayed by time series forecasting. Seasonality can be simply described by periodic fluctuations. Seasonality is less common in engineering and scientific data but is commonly used in economic time series. Retail sales tend to peak for holiday seasons and then decline after the holidays. So time series of retail sales will typically show increasing sales from September through December and declining sales in January and February.

If seasonality is present, it must be incorporated into the time series model. Technique used in detecting seasonality are: run sequence plot to show seasonality; seasonal sub-series plot to specialize technique for showing seasonality; box plots used as an alternative to seasonal sub-series plot to detect seasonality; and autocorrelation plots can help identify seasonality. http://www.itl.nist.gov/div898/handbook/pmc/section4/pmc443.htm

Delphi Method

The Delphi method is a group decision process about the likelihood that certain events will occur. The objective of the Delphi method is to explore reliable creative ideas or produce suitable information for decision making.

The method selects experts based on areas of expertise. This method is based on the assumption that individuals with insight and experience on a particular subject will make better predictions of the future than theoretical approaches or extrapolation of trends. The Delphi method has been developed in order to make discussion between experts possible without permitting a certain social interactive behavior as happens during a normal group discussion and hampers opinion forming. The Delphi method has been widely used to generate forecasts in technology, education, and other fields. http://www.iit.edu/~it/delphi.html

The Delphi method uses questionnaires that are sent out to a pre-selected group of experts to elicit and develop individual responses to posed problems. This enables the experts to refine their views while the group continues to work on the assigned task. The goal of the Delphi method is to overcome disadvantages of conventional committee actions.

Never having to bring the experts together physically is an advantage of the Delphi Method. Since the majority opinion is represented by the median the process does not require that all experts agree. A disadvantage is that it is often difficult to keep experts available for the numerous rounds of questionnaires. Also, the prediction of future developments is not always accurate by iterative consensus nor by experts but by "off the wall" thinking or by "non-experts".

http://www.ryerson.ca/~mjoppe/ResearchProcess/841TheDelphiMethod.htm

Technological Method

The Technological Forecasting method is used to analyze the market for the life span of an existing technology to determine if its close to end of like and to see if a new product or technology is ready to enter an existing market. It is also used to identify competing new technology and to forecast sales.

Before a new innovative product enters into the market Technology Forecasting is one of several methods used to determine if customers will buy it. The Technology method should always be used in conjunction with other tools to identify prospective customers, prototypes, focus groups, interviews, market testing, internet polls and other tools to get a better understanding of the market.

The major techniques for technological forecasting is numeric data and judgmental. Numeric data-based forecasting extrapolates history by generating statistical fits to historical data. Judgmental forecasting can also be based on past projection but like the Delphi method it relies on the subjective judgment of experts. Keep in mind that technological forecasting is most appropriately applied to capabilities, not to the specific characteristics of specific devices.

Other Numeric data techniques are Trend Extrapolation, Qualitative Approaches, Growth Curves, Envelop Curves and Substitution models. Techniques used by Judgment-Base method are Monitoring,