Female smoking prevalence forecast in Eastern Europe 2001-2029

The shown forecasts represent a blend of multiple input datasets from both internal (primary) and external (secondary) sources. Whereas primary data are generated via Statista's own surveys like the Global Consumer Survey, secondary input datasets are mostly sourced from international institutions (such as the IMF, the World Bank or the United Nations), national statistical offices, trade associations and from the trade press. These datasets are often incomplete as there are gaps between survey years or no or no reliable information might be available for a specific indicator in a specific country or region. Data for missing years are interpolated by various statistical means, such as linear or exponential interpolation, or cubic splines. Data for missing countries or regions are imputed by considering known information from other countries or regions that are found to be similar by cluster analyses like k-means or similar procedures.

Most indicators are composites of multiple input sources with slightly varying methodologies that have been processed by our analysts to be aligned and consistent with each other and with all other indicators in the KMI database. Due to ther addition of new data or changes in underlying methodologies it can be possible that data is not comparable any longer with previously published data or is changed retroactively to suit the new definitions.

Because of the high degree of processing, no specific external source can be named for each data point and all data for historical years (usually one to three years before the last finished calendar year before the current one) have to be considered Statista estimates. Future years are mostly Statista projections. These projections or forecasts are conducted by regression analyses, exponential trend smoothing (ETS) or similar techniques and extrapolate the found historical trend.