ABSTRACT

This project deals with time series analysis on the volume of Eagle Stout in Nigeria between January 1994 and December 1999. This data was obtained from the Marketing Division of the Golden Guinea Breweries PLC, Umuahia, Abia State. The basic assumptions of the time series were tested. The analysis of the volume of sales was done with autocorrelation and partial autocorrelation functions. A statistical package in computer, Minitab was employed during the analysis. The analysis showed no trend but the seasonal variation in the volume of sales each year. This revealed that the series was non-stationary. The data was differenced. The 95% confidence limit was 0.237, which is also the standard error of the autocorrelation. It could be seen that four out of the 48 values of PK lie just outside the significant limits. These occurred at lags 3,6,9,12.This showed the presence of seasonal variation. The fitted model was shown as Seasonal autoregressive integrated Moving Average of order (101 )(01 1)12. With the models I forecast the volume of sales of Eagle stout. Finally, recommendations on the possible ways of regulating production were made at the end of the work.