## Question 733:

1## Answer:

No answer provided yet.First we see that the absolute frequency of the responses in the 1st table have been adjusted to remove the unknown, blank or not ascertainable results. This leaves the values of interest being 156 did acknowledge getting suggestions or information that was helpful and 368 that did not, out of a total of 524 respondents.We're interested in the proportion who did find helpful suggestions, which is 156/524 = .29771 or just less than 30%. So from the percentages alone we can see that the overwhelming majority of respondents did not say they got helpful suggestions. We can compute a large sample confidence interval around the likely proportion of all taxpayers based on this sample using the the following steps.

- Multiply the proportion times 1-proportion = .29771*.70229 = .209079
- Divide the result from step 1 by the sample size = .209079/524 = .000399
- Take the square root of the result in step 2 = SQRT(.000399) = .019975, this is the standard error (SE).
- Multiply the SE times the critical value for a confidence level of 95%, which is 1.96 (found from a normal table or the percentile to z-score calculator) = 1.96*.019975 = .039151, this is the Margin of Error.
- Add and subtract the margin of error to the proportion to provides the interval between .258559 and .33686

Of the 156 respondents who did recall finding helpful information the second table shows us that about an equal percentage of those recall learning about another deduction (31.3%), receiving pamphlets/forms (27.2%) or some other piece of information (29.9%). Only a small percentage recalled learning about energy credits (5.4%) or which forms to use (6.1%). We see that these percentages represent a very small portion of the total respondents (e.g. only 8 out of 534 respondents recalled energy credits). As a final observation, there should probably be an additional category since so many reported the "other" , almost 30%.