Góc học tập
17/10/2017 09:04:16 AM

QUALITATIVE FORECASTING METHODS

Panel Concensus.

The method, is by bringing the internal members or experts from all level in the company together, and have a open discussion about a product, any people is allowed to give their own opinions. The behoof of this method is that it involves large amount of people in forecasting process, the group forecast should be better than an individual one, and the outcome would be much more balanced, as the members are from variant aspect. Company executives’s involvement might make the forecast fit better with company’s situation, since these people make the policy.  In addition, this method is relatively quick. The drawback is that this method highly relies on experience and opinion, and with just a little resource to the fact. During the process, the top-level may affect the speech by the people from lower-level. In addition, due to the group dynamic,  a strong personality person might have more affect on forecasting, these factors will introduce bias to the final forecasts.

Historical Anology

The expert might estimate the probable demand for a product the company will introduce in one regional market based on his review of historical data regarding the introduction of similar products in other regional markets. Assuming the expert has experience with analogous situations, he will be able to gauge the similarities between the proposed and prior product introductions. Based on these similarities or differences, the expert will consider to what degree the probable outcome of the new product introduction will mirror the outcomes of previous product introductions.

Delphi method

  1. Design the questionnaire
  2. Invite participants to take part
  3. Send out first round of questionnaire- a typical question may be-

Example:

Which of the following CLINICAL areas do you think are HIGH PRIORITY for development of an improved evidence base relating to minority ethnic groups and their health needs?

Clinical Area

PRIORITY for development of an improved evidence base

Comments –including any particularly important topics for action.

Mental Health

Low 1 2 3 4 5 High 0 (don’t know)

 

Cancer

Low 1 2 3 4 5 High 0 (don’t know)

 

Immunisation/Vaccination

Low 1 2 3 4 5 High 0 (don’t know)

 

The questionnaire provides space for respondents to raise any other issues relating to the topic. The first round of the questionnaire aims to categorise opinions under common headings.

4.Analyse responses from round 1 questionnaire

5.Prepare the second round questionnaire

6.Send out second round questionnaire - the follow up to the first round question above might be-

Example:

Cancer has been identified as a high priority for developing an evidence base relating to minority ethnic groups. Within this clinical area, what aspects should research focus on?

Research area

PRIORITY for development of an improved evidence base

Comments –including any particularly important topics for action.

Identifying risk factors of disease

Low 1 2 3 4 5 High 0 (don’t know)

 

Identifying barriers to access of services

Low 1 2 3 4 5 High 0 (don’t know)

 

Improving the patient experience for minority ethnic groups

Low 1 2 3 4 5 High 0 (don’t know)

 

In this example, the research areas chosen for the second round questionnaire were identified from the free text column in the round 1 questionnaire. Participants have the chance to suggest further areas of focus in the second round questionnaire. The aim of second round is to score agreement or disagreement with statements from first round.

7.Analyse responses from round 2 questionnaire

8.Design the third round questionnaire. For the third round, the second questionnaire is repeated but incorporates scores from the second questionnaire results. This gives participants a chance to see how the rest of the group prioritised the areas and if the participant then wants to change their opinion on the basis of the group consensus, has the opportunity to do so.

9.Analyse the results of the third round questionnaire for agreement and degree of consensus

10.Report findings