Empirical Tests on Regressive Estimations: Implications on Surveys and Diaries