For a web survey asking opinions about drug policy, think of the whole sample as all those who visited the website in question. The sample then splits into $n=1000$ people who chose to complete the questionnaire, and an unknown number of people who visited but chose not to complete the questionnaire.

Amongst the respondents, indicated support for safe-injection sites, while the other indicated opposition. Thus the proportion of respondents supporting safe injection sites is .

Now, consider the odds that someone sampled responds to the questionnaire given he/she supports safe injection sites, and the odds of responding given opposition to safe injection sites. A drug-policy expert tells you she isn't sure which of these two odds is larger. But she is sure that the larger odds will be no more than times the smaller one.

[A] Based on the respondent data and the expert opinion, give an upper-bound on the proportion of the sample supporting safe-injection sites (report your proportion to three decimal places).


[B] Similarly, give a lower-bound on the proportion of the sample supporting safe-injection sites (again reported to three decimal places).