Observation periods starting in the first quarter Finally, in our main sample, observation periods start in different quarters of the year. Some differences in temporal consumption patterns between consumers may be due to seasonal effects. Although we control for seasonal variation (i.e., via quarterly dummies) in Subsection 22, seasonality may still affect the revealed preference tests. In addition, some sequences of observations may be more likely to coincide with a planning period than others.
