Hi Santhosh ,
I think there is some confusion over here. The terms moving average and weighted average refer to 2 totally different concepts.
Moving average is always over a period of time ; the term moving implies that the window of time which is used to average the values keeps moving ; thus you can have a 12 month average over the period Jan , Feb , Mar ,..., Nov , Dec. This is not a moving average , since it is an average over 12 months. With the same 12 months of data , you can have a moving average with a window of 3 months ; in this case , you will have the following averages :
1. Jan , Feb , Mar
2. Feb , Mar , Apr
3. Mar , Apr , May
.
.
.
10. Oct , Nov , Dec
Thus , with the same 12 months of data , you can have 10 averages.
If you now change the window to 4 months , you will have 9 averages.
Weighted average is in no way connected with time ; it is connected with the relationship between 2 or more values , where some values are more important or weighty than others. A straightforward example is when the winner of a competition is decided using votes from judges as well as votes from an audience ; suppose the number of judges is 5 while the audience numbers 100.
If each vote of a judge carries the same weightage as the vote of a member of the audience , then effectively the judges will be outnumbered by the audience , since they can at the most pitch in with 5 votes.
Now , suppose a weightage of 10 is given to each judge's vote ; effectively , the 5 judges can now contribute 50 votes.
Thus suppose one candidate receives 3 votes from the judges , and 40 votes from the audience ; this means that the other candidate received 2 votes from the judges and 60 from the audience ; the totals received by the two candidates are 30 + 40 vs. 20 + 60 i.e. 70 vs. 80 which means the second candidate won.
Now to your measure in cell F4 ; this is not the way percentages should be averaged. Taking the average of a series of percentages is never done , since a percentage is always on a base of 100. The individual percentage values may or may not be on a base of 100.
Suppose a student appeared in 2 tests ; in the first he got 90 out of 100 , while in the second he got 10 out of 10. Thus taking a straightforward average of the percentages will result in an overall percentage of 95 % ( 90 % + 100 % will equal 190 , divided by 2 will equal 95 ). However , this will not be a correct representation of the overall result , since the second test had a maximum marks of just 10 , compared to the 100 which the first test had. The correct overall result would be to add the marks obtained in the 2 tests , and divide this by the addition of the maximum marks in the 2 tests , thus giving ( 90 + 10 ) divided by ( 100 + 10 ) , which will give 100/110 or approximately 91 %. Whether this can be called the weighted average or not is a different matter altogether , but the fact is that as a composite measure of performance , it gives a better representation than 95 %.
Narayan
