Once you create an index score, the calculation of counts can get really confusing. Let's start with an example:
We have 3 respondents:
| respondent id | scale1 | scale2 | index | weight |
| 1 | 1 | 1 | 0,75 | |
| 2 | 2 | 2 | 2 | 1,5 |
| 3 | 3 | 2 | 2,5 | 0,75 |
UNWEIGHTED:
Once you calculate the mean of the index score, you will get a final result of 1,8 based on 3 respondents (base size).
However, once you calculate a count it will just give you the result of all 1, 2, and 3's given:
| 1 | 2 | 3 | |
| Count | 1 | 3 | 1 |
This calculation is based on a total count of 5 (1+3+1 = 5, which is the base size)
And, once you calculate a percentage share it will just give you the result of all 1, 2, and 3's given:
| 1 | 2 | 3 | |
| percentage share of index | 20% (count =1) | 60% (count =3) | 20% (count =1) |
This calculation is based on a total count of 5 (base size is 5) too.
WEIGHTED:
Once you calculate the mean of the index score, you will get a final result of 1,88 based on 3 respondents (base size).
However, once you calculate a count it will just give you the result of all 1, 2, and 3's given:
| 1 | 2 | 3 | |
| Count | 0,75 | 3,75 | 0,75 |
This calculation is based on a total count of 5 (0,75+3,75+0,75 = 5, which is the base size)
And, once you calculate a percentage share it will just give you the result of all 1, 2, and 3's given:
| 1 | 2 | 3 | |
| percentage share of index | 14,3 % (count =1) | 71,4 % (count =3) | 14,3 % (count =1) |
This calculation is based on a total count of 5 (base size is 5) too.