The project deals with the estimation of day-to-day variation in aggregate expectations using information about the day on which individual (survey) expectations are formed. This kind of data is sometimes gathered in surveys. Methods are developed both for quantitative and qualitative survey expectations data. Moreover, it will be shown how additional variables likely to be related with day-to-day changes in aggregate expectations can be used to improve the estimate of latent aggregate expectations, especially on days when no survey responses are available.