ABSTRACT

 We used time series analysis to analyze the number of letters posted in Abia State. In the analysis, the magnitude or the extent to which each of the time series components namely trend (Ii) seasonal (Si) Irregular (li) and cycle component (Ce) affected the series the number of letters posted every month for five years (2000-2004) were observed. The effects of the components were also estimated. The trend component variations were estimated using (a) Graphical Method (b) Principle of least squares method. (b) Moving average method. A cursory look on the graphical method shows a very slightly upward movement of the trend. This was due to the fact the posting of letters in the Abia Territory increase towards. 2002 and 2003. The trend line was estimated to be X+=20.58+0.49Y The seasonal components were estimated using the ratio to trend method. E.g. the seasonal index for the month of January is. 113.9, which means that the posted letters per day were about an average of 113.9 or -t vii thousand Approximately. Finally, the number of letters posted during the year 2005 was predicted using the multiplicative, model of time series.