Naive forecasting encompasses a family of forecasting techniques based on averaging, or "smoothing". The techniques are so called because no attempt is made to use regression mathematics to establish causal relationships (i.e. statistically proved cause-and-effect relationships) among the data being subject to analysis (contrast Causal Forecasting). Instead, with naive forecasting, on each occasion a forecast is to be calculated, the data are analyzed (i.e. averaged) in precisely the way they were analyzed the previous time.