Thread: statistics

I have two surveys asking the same question, but the question is worded different. I did a confidence interval for one survey, and the other survey does not fit in the interval, so this means that the wording in each does make a difference on the response.

Is a pooled value a reasonable estimate?
Even though there is different results?
If not, what would be reasonable?

Originally Posted by Morgan82

I have two surveys asking the same question, but the question is worded different. I did a confidence interval for one survey, and the other survey does not fit in the interval, so this means that the wording in each does make a difference on the response.

Is a pooled value a reasonable estimate?
Even though there is different results?
If not, what would be reasonable?

Yes wording does make a difference.

No you cannot pool them.

Look at the questions and try to determine what the difference in the wording might do, then choose the survey whos wording gives the answer to the question really wanted.

RonL