Why Use Weighted Averages in Survey Results?

[DaveC426913]

We got some results back from a company survey today. The speaker showed us the numbers and told us they had weighted the averages.

Say, there were two questions:

                                    All the time | often | sometimes | rarely
How well does our company communicate?   60%     |  30%  |    10%    |   0
How much are your skills utilized?	 40%     |  50%  |    10%    |   0

I asked why.

They said that they had given more weight to the aspects that are more important, for example, we are interested in the positives of our company, so they were weighted heavier. They might have set them as All the time = 1, often = .75, sometimes = .5 and rarely = .25.

It seems to me, all this does is skew the results in favour of what they want to see. What is the purpose of weighted averages in this example?

 

[Number Nine]

In that specific example? Nothing, as far as I can tell. It's as you say: they wanted to skew the results in their favour. The results don't even mean the same thing anymore (I'm not sure what "60% of respondents said 'always'" is supposed to mean when they took a weighted average).

 

[gsal]

You are correct, they seem to be skewing the results, period...there is not proper weighting here.

If they had gotten answers from offices in two different cities, one with 2000 employees and one with 3000 employees and they had only received 100 surveys from each of the offices, I can see weighting them so that the 100 answers from the 3000-employee building represent 60% of the population and the answers from the 2000-employee building represent only 40%...other than that...giving more weight to one answer over other one is just cheating.

my 2 cents