Stated intentions and purchase behavior: A unified model

Intentions data often contain systematic biases; intentions change over time and may not accurately predict actual purchases. Ignoring the discrepancies between intentions and purchasing can produce biased estimates of variable coefficients and biased forecasts of future demand. This study proposes a unified model that takes into account various sources of discrepancies between intentions and purchasing and forecasts purchasing probability at the individual-level by linking explanatory variables (e.g., socio-demographics, product attributes, and promotion variables) and intentions to actual purchasing. The proposed model provides an empirically better explanation of the relationship between stated intentions and purchasing and offers more accurate individual-level purchase predictions than do other existing intention models.