Daily and weekly smoking habits: a Box-Jenkins analysis.

The objective of this research was to ascertain if there are different temporal patterns of smoking. The method of data collection was to use voluntary subject smokers who recorded their daily cigarette consumption for 84 days. Subjects had smoked more than 5 cigarettes per day throughout the previous year; 29 subjects kept accurate autorecords. The daily smoking data of each subject were analyzed via the time-series procedure ARIMA(p,d,q)(P,D,Q)s of Box and Jenkins. 15 subjects (52%) showed simple autoregressive smoking models for which smoking on any given day was a function of the number of cigarettes smoked on the previous day or days, but 13 subjects (45%) showed autoregressive models of weekly seasonality, i.e., the number of cigarettes smoked on any given day is a function of the number smoked on the same day of the previous week, and only 1 subject's data (3%) had unpredictable smoking patterns.