Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
13 Cards in this Set
- Front
- Back
List types of forecasting techniques
|
1. Qualitative
2. Causal modeling 3. Time series methods |
|
Describe qualitative methods
|
based on judgement rather than records of past data
inferior to quantitative methods |
|
Describe Causal Modeling
|
means the variable to be forecast is related statistically to one or more which are thought to cause changes in it
|
|
Time series methods
|
predict future values of a variable solely from historical values of itself (with regression analysis)
|
|
When are time series methods of forecasting best?
|
1. Conditions are stable and will continue
2. Short-term forecasts are required 3. Base forecast is needed on which can be built changes in future conditions |
|
Which forecasting method is cheapest and easiest to maintain?
|
Time series
|
|
What is the time horizon for forecasts?
|
Short term - up to a year
Medium term - 1 to 5 years Long term - more than 5 years |
|
Which methods are used for which time frames?
|
Short-term - time series
Medium-term - causal modeling Long-term - qualitative techniques |
|
Describe qualitative forecasting techniques
|
combination of experience, judgement, intuition. Use Delphi technique.
|
|
What are the two motives behind qualitative forecasting?
|
1. a lack of adequate data
2. factors affecting the forecast may be better handled qualitatively |
|
Common qualitative methods
|
Visionary forecasting - 1 person
Panel consensus - generally inaccurate Brainstorming - can be useful - to panel type Market Research - accurate but expensive Delphi - no communication between Scenario writing - Cross-impact matrices |
|
Common Qualitative methods (cont)
|
1. Analogies - when no data - a second variable who history is completely known is used
2. Catastrophe theory - like contingency planning 3. Relevance trees - starts in future and what must happen today to make future happen (normative) |
|
Procedure for building relevance tree
|
1. Draw the relevance tree
2. Establish criteria 3. Weight the criteria 4. Weight sub-criteria (elements) - equals to 1 for each criteria 5. Calculate PRN - partial relevance numbers PRN = Criterion wt. x element wt. 6. Calculate LRN - local relevance number for each element (sum of PRNs for that element) 7. Calculate CRNs - Cum. relevance numbers for each element (CRN = sum of LRNs for each associated element 5. |