One of the goals of statistical analysis is to generalize a sample-based outcome onto the population of interest. In order to do this, the sample you collect should be as random as possible. This is done to mitigate the effects of bias in your sample. This is one of the rst steps in the statistical process and I cannot stress enough how important good information is. If you've ever heard the phrase \"garbage in, garbage out" it most denitely applies here. There is no level of statistical elegance that can x bad data. Let's look at an example to try and make this more clear. Suppose we have a politician we'll call Candidate Zero. Zero has no genuine political opinion on any topic; their only concern is to get elected to a federal position. To do this they desperately need to know how the American voting population feels about certain hot button issues. Forthis example we'll use marriage equality. Now suppose Zero hires 3 polling rms, Firm A, Firm B, and Firm C, to inform Zero of what stance he should take on the issue of marriage equality. Firm A sends their pollsters to churches in rural Mississippi to collect their sample data. Firm B sends their pollsters into the Castro neighborhood in San Francisco to collect their sample data. Firm C collects their sample data through random selection of registered American voters. Remember, Candidate Zero has no position on this issue, they only want to pander to the majority, i.e. they only want to get elected. 1} Identify the population of interest for this scenario. 2} Identify any issues that might be problematic with the methodology employed by each rm. Does anything feel off here? 3} Draw a connection between the data collected by Firm A and Firm B. What is it about these two data sets that makes them similar? 4} Explain how polling might help a candidate "evolve" on an issue