Delphi Method: Forecasting Approaches

The Delphi method is part of the large family of forecasting approaches used in Statistics. Forecasting is the art and science of predicting future events in an increasingly complex world economy making forecasting challenging. Forecasting involves taking past events, such as online sales, and predicting them using mathematical or intuitive techniques. We might use intuition because of the consideration regarding the product cycle, where sales are in Introduction, growth, maturity, or decline. This is why there is rarely a perfect method of forecasting. Demand can also influence forecasting, for example, the demand of aluminum in the aeronautical market. There are seven steps in the forecasting system. First, by determining the utilization of the prediction. After we decided over the use of the forecasting, we should select the population about which statistical inference should be drawn. The duration of the forecast should determine the third step in forecasting, whether the prediction is made on a short, medium, or long-term basis. After these 3 steps, we should choose carefully the forecasting models and gather the data we need for the prediction. In the end, making the forecast and validating and implementing the results stand as the last steps. There are 2 types of forecasting approaches: quantitative and qualitative. There are 5 quantitative methods which include the nave approach, moving averages, and exponential smoothing part of the time series models group. Associative models are linear trend regressions that are part of the extended family of quantitative methods of forecasting. Moving to qualitative methods, there are four different forecasting techniques: Market survey, Sales force composite, Jury of executive opinion, and Delphi method.

Delphi's method of forecasting is a qualitative method. 'Qualitative forecasts incorporate such as the decision maker's intuition, emotions, personal experiences, and value systems'. A qualitative method of research has its weaknesses that can lead to inaccuracies. The vision in the way we approach the research can have its advantages and disadvantages. On one hand, A smaller sample size can lead to a more in-depth investigation of the subject under study. It can also facilitate the discovery of new ideas. On the other hand, a smaller size cannot be generalized to the wider population. If we vary the questions, one of the disadvantages would be a long time to interpret the data, and also it can lead to follow-up questions during the data collection time. Moreover, varied questioning can facilitate different directions to the researchers and get more elaborated deeper exploration during the data collection process can constitute the disadvantages of a diversified set of questions. 

A natural setting can help with more valid results. Disadvantages might be that the data can not be reliable if the respondent becomes impartial. Meanwhile, the research can produce no information if the study is unfocussed if it is not properly planned. Delphi technique was introduced in the 1950s by RAND Corporation and its name derives from the Oracle of Delphi. Delphi method is also known under the name of ETE- Estimate talk Estimate. The traditional version was a multi-round survey, widely used in business, and is used as a means of communication in a panel of experts. A big advantage of the Delphi technique over other forecasting methods, especially for business, as an anonymous inquiry method, is that maintains the advantages of an interacting group without the opposite effect of a dominant member of the panel who might not be the best expert. Another advantage of using a multi-round method is that the respondent can modify or add comments to the primary answer. Compared to the quantitative research method, the Delphi technique which is a qualitative method is a combination of number based and text-based which allows the respondent to use his spontaneity. Answers are generally numeric estimates, yesno, or ratings on the scale. 

First used in the 1950s, developed in the way we communicate and we trade information, the Internet will have a large impact because it will possibly expand the number of participants in the forecast. A good example is, the state of Alaska, for example, which has used the Delphi method to develop its long-range economic forecast. A large part of the state's budget is derived from the million-plus barrels of oil pumped daily through a pipeline at Prudhoe Bay. The large Delphi panel of experts had to represent all groups and opinions in the state and all geographic areas.