Qualitative Forecasting Methods

• Forecasting is a crucial aspect of operations management, enabling informed decisions about production, inventory, and resource allocation

Qualitative Forecasting Methods

• Rely on expert opinion and subjective judgment
• Useful when historical data is limited or unavailable
• Often used for new product introductions or significant market changes

Jury of Executive Opinion

• Combines the opinions of a small group of high-level managers
• Relatively quick and easy to implement
• May be dominated by one individual's opinion

Delphi Method

• Uses a panel of experts to answer questionnaires iteratively
• Maintains anonymity of individual responses to avoid groupthink
• Aims to reach a consensus forecast through multiple rounds of feedback

Sales Force Composite

• Gathers forecasts from individual salespeople in the field
• Salespeople are likely to be aware of customer demand
• Can be overly optimistic or pessimistic due to individual biases

Consumer Market Survey

• Directly solicits input from customers regarding their future purchasing plans
• Can provide valuable insights into customer preferences and intentions
• Can be time-consuming and expensive

Quantitative Forecasting Methods

• Utilize historical data and mathematical models to predict future outcomes
• Appropriate when historical data is available and stable
• Can be more objective and consistent than qualitative methods

Time Series Analysis

• Analyzes historical data patterns to predict future values
• Assumes that past patterns will continue into the future
• Common time series components include trend, seasonality, cycles, and random variation

Moving Average Method

• Averages data from a specified number of past periods to generate a forecast
• Simple to calculate and understand
• Equally weights all data points in the average
• Sensitive to the number of periods included in the average (larger number of periods = less sensitive)

Weighted Moving Average Method

• Similar to the moving average method, but assigns different weights to each data point
• Allows more recent data to have a greater impact on the forecast
• More flexible than the simple moving average method but more complex to utilize

Exponential Smoothing Method

• Averages past data, but weights the most recent data more heavily
• Requires a smoothing constant (alpha) to determine the weight assigned to recent data
• Simple and widely used forecasting technique
• Alpha values close to 1 place more emphasis on recent data
• Alpha values close to 0 place more emphasis on older data

Trend Projection

• Fits a trend line to historical data to project future values
• Can be used to forecast linear or non-linear trends
• Linear trend equation: y = a + bx, where y is the forecast, x is time, a is the y-intercept, and b is the slope
• The trend line can be determined with regression analysis

Seasonal Variations

• Regular, predictable patterns that occur within a year
• Can be incorporated into forecasts using seasonal indexes

Multiplicative Seasonal Method

• Seasonal factors are multiplied by an estimate of average demand to produce a seasonal forecast.

Additive Seasonal Method

• Seasonal amounts are added to an estimate of average demand to produce a seasonal forecast.

Cyclical Variations

• Patterns in the data that occur every several years
• More difficult to predict than seasonal variations
• Often influenced by economic or political factors

Associative Forecasting Methods

• Also known as causal forecasting
• Incorporates factors that might influence the quantity being forecast
• Used when changes in one or more independent variables can be used to predict changes in the dependent variable

Regression Analysis

• Uses statistical techniques to determine the relationship between variables
• Simple regression involves one independent variable
• Multiple regression involves multiple independent variables
• Objective is finding the best fit.

Correlation

• Indicates the strength of the relationship between variables
• Ranges from -1 to +1, where 0 indicates no correlation
• Positive correlation indicates that variables move in the same direction
• Negative correlation indicates that variables move in opposite directions

Forecast Error Measurement

• Essential for evaluating the accuracy of forecasting methods
• Helps to identify areas for improvement

Mean Absolute Deviation (MAD)

• Average absolute difference between actual and forecasted values
• Easy to understand and interpret
• Does not indicate the direction of the error

Mean Squared Error (MSE)

• Average of the squared differences between actual and forecasted values
• Penalizes larger errors more heavily than smaller errors
• More sensitive to outliers than MAD

Mean Absolute Percentage Error (MAPE)

• Average absolute percentage difference between actual and forecasted values
• Expresses error as a percentage of actual values
• Easy to compare across different datasets

Tracking Signal

• Measures how well the forecast is keeping pace with actual values
• Calculated by dividing the cumulative forecast error by the MAD
• A large tracking signal indicates a potential bias in the forecast

Choosing a Forecasting Method

• Accuracy: how well the method predicts future values
• Data availability: the amount and quality of historical data
• Time horizon: the length of the forecast period
• Cost: the expense of developing, implementing, and using the method
• Ease of use: the simplicity and understandability of the method
• Important to consider the specific context and requirements of the forecasting situation

Explore qualitative forecasting methods: jury of executive opinion, Delphi method, sales force composite. These techniques rely on expert opinions and are useful when historical data is limited. They help in making informed decisions in operations management.