The interval determination of indicators is a method of research for precisely indeterminate indicators under the state of uncertainty natural for the economy. The interval representation of the indicator enables a description to be made of its most likely values. The interval method of economic indicators will open a logical path from the predicament of explaining its quantitatively precise indefinability and deploying a complete picture of causal links in the economy.

The relations of economic indicators can be represented using mathematical models. However, the most detailed record of economic links will not change the fundamental impossibility of determining the exact values of economic indicators. Considering the principle of uncertainty, the causal links of key indicators can be presented by defining their range of possible values. The interval uncertainty of key economic indicators represents the infeasibility to pinpoint an economic indicator within the minimal uncertainty interval.

The difference between the lower and upper bounds of the key indicator is its minimal uncertainty interval for a given probability. This range is the T-interval.

When forecasting key economic indicators, the use of the minimal uncertainty interval is an underlying idea. Based on this idea, a system of intervals of key economic indicators can be deduced, which is a necessary condition for economic efficiency. Thus, economic predictability can be maximized.

Scenario forecasts can reduce the uncertainty interval of an economic indicator in each scenario. However, they cannot eliminate uncertainty by turning the interval into a point estimate. An overlap can be purely accidental, like a shot in the dark. Decision making which is subject to uncertainty amounts to the determination of the most possible minimal uncertainty interval for key indicators and designing scenarios for economic development with their implementation methods.

The diagnosis of the minimal interval of an economic indicator, including its extreme values, and possible consequences when obtaining such values, is no less important than the forecast itself. When carrying out an interval forecast of an indicator, its diagnosis is also necessary. The utility of the forecast depends on the interval bounds—the less, the better. It also depends on how accurate the estimates of the possible consequences in case of obtaining results outside the interval are.