Help me understand something statistic related.
 

I've been saddled with this survey thing at work and so far, I've been able to muddle through, but I've hit a bit of a roadblock.

Let's say the survey asks the respondent to rate a series of services using a Likert (I think that's what she called it) scale of 1-4 where 1=very dissatisfied and 4=very satisfied.

So, let's say that--and I'm going to be making these numbers up and they will be deliberately low as it will help me to understand--that 100 people were very satisfied with our book selection; 113 were satisfied; 57 were dissatisfied and 30 were very dissatisfied.

What I'm looking for is the average response, for example, "On a scale of 1-4 with blah blah blah, the average response was 3.28."

I should point out that Mathematics...well, we aren't on good terms. Although I 'get' the big concepts in the abstract, anything beyond college level algebra in a practical sense, I don't really understand. So...the simpler the better, eh?

As always, any and all help is greatly appreciated. :lol:

I might use a larger scale and include a neutral point (neither satisfied, nor dissatisfied), but if you're just looking for the average, each person's response is a score, add up all of the responses and divide by the number of responses.

((113 x 3)+(57 x 2)+(30 x 1))/200

sapiens, i think you forgot the 'very satisfied' people, wouldn't you have ((100*4) + (113*3) + (57*2) + (30*1))/200 ?

Bascially gothmund, you're just adding up all the responses, and then dividing by the total number of responses. The multiplication steps in the above just speed it up. If you're tallying these in Excel, then it has an 'average' function that will calculate it for you.

Minor threadjack: I've heard you get a "better" result if you don't include a neutral option, because it forces people to express an opinion if they are unable to cop out with a neutral answer.

Wow.

That seems remarkably simple... :hmm: I was making it so much more complicated.

Thanks all :thumbsup:

"Better" only in the sense of getting more people to express an opinion. Most research shows, however, that most people don't have well formed opinions on most questions (see "The Nature and Origins of Mass Opinion" by John Zaller). To keep people from expressing an opinion they barely hold, or from choosing answers randomly, you should usualy include a "don't know" option.

There's a really intersting study that found that upwards of 40% of the population will express an opinion on a totaly made-up issue if you don't include a "don't know" option.

Yikes! Thanks.

I agree with ickky. Omitting a neutral option is only better if better means "less representative of people's actual opinions".

More agreement with icky: I like "don't know" in addition to "Neither satisfied nor dissatisfied". Not knowing is different from being neutral, just ask Switzerland.