Hi,

I'm trying to create projections on a user survey.

Basically, in the question, there are 1000 computer users in a school:

Age range | No of children

10-12 200

12-14 200

14-16 400

16-18 200

The school is trying to work out how many Macs and PCs they should buy for next year.

A sample of 100 users reveals some like windows some like MACs etc, such as:

Age Range | No of children | Like MACs | Like PCs | Not Bothered

10-12 20 5 10 5

12-14 20 3 12 5

14-16 40 10 7 3

16-18 20 12 7 1

I calculated the % for each age range for each like mac or like pc and created a new table with the projections included. In the table, I totalled the number of people who would like macs and the number of people who would like windows using these projections.

However, if I use the sample above by taking the totals of like macs and like pcs and working out the percentage regardless of age range

e.g.

Total like Macs = 30

% of sample who like macs = 30/1 = 30%

Total like PCs = 36

% of sample who like pcs = 36/1 = 36%

The numbers come out differently.

I'm confused. I'm not sure whether to work out each individual age group projections then add up the total....OR....use the method above i.e. sample has 30% who like MACs therefore 300 macs are need

Please help !