Chapter 4: Forecasting Key Terms and Concepts

Vocabulary flashcards covering key definitions, techniques, time horizons, and concepts from Chapter 4 on Forecasting.

Forecasting

The process of making predictions about the future based on past and present data, most commonly by analysis of trends.

Prediction

A general term referring to estimates of future or unobserved values, using formal statistical methods or less formal judgmental methods.

Qualitative forecasting techniques

Subjective forecasting methods based on the opinion and judgment of consumers and experts, appropriate when past data are not available and usually applied to intermediate- or long-range decisions.

Quantitative forecasting models

Forecasting models used to project future data as a function of past numerical data, applied to short- or intermediate-range decisions when patterns are expected to continue.

Average approach

A forecasting approach in which predictions of all future values (or unobserved values) are equal to the mean of past observed data.

Naïve approach

A cost-effective time-series forecasting method where forecasts for a future period are set equal to the last observed value.

Drift method

A variation of the naïve method where forecasts are allowed to increase or decrease over time based on the average change seen in historical data.

Seasonal naïve approach

A time-series forecasting method that accounts for seasonality by setting each prediction equal to the last observed value of the same season.

Time series methods

Quantitative forecasting methods that use historical data patterns as the basis for estimating future outcomes under the assumption that past demand indicates future demand.

Autoregressive moving average (ARMA)

A time series forecasting method where forecasts depend on past values of the variable being forecast and on past prediction errors.

Autoregressive integrated moving average (ARIMA)

A time series forecasting method that applies an ARMA model to the period-to-period change in the forecast variable.

Causal / econometric forecasting methods

Forecasting methods that identify underlying external factors or relationships between variables that influence the variable being forecast.

Associative models

Quantitative models, also known as causal models, that assume the forecasted variable is associated with other variables and make predictions based on those associations.

Judgmental forecasting methods

Forecasting methods that incorporate intuitive judgment, opinions, and subjective probability estimates, particularly when historical data is lacking.

Forecast error (residual)

The difference between the actual value and the forecast value for the corresponding period.

Cross-validation

A sophisticated evaluation procedure where models are fitted on a training set of historical data and evaluated on a test set to determine forecast accuracy on new data.

Seasonality

A characteristic of a time series in which data experiences regular and predictable changes or patterns that recur over a fixed period, such as every calendar year or day of the week.

Cyclic behavior

Regular fluctuations in time series data that usually last for an interval of at least two years, where the length of the current cycle cannot be predetermined.

Self-destructing predictions

Predictions that undermine themselves by influencing social behavior in a way that alters the context and reduces or invalidates the original forecast.

Short-range forecast

A forecast typically covering less than 3 months up to 1 year, used for operational decisions such as weekly scheduling and inventory carrying.

Medium-range forecast

A forecast typically spanning 3 months to 1 year (up to 3 years), used for sales planning, production planning, and cash budgeting.

Long-range forecast

A forecast with a time span of 3 or more years, used for strategic decisions like product design, plant facility location, and long-range capital plans.