1) Suppose you were hired to be on the campaign team of one of the presidential
ID: 3203946 • Letter: 1
Question
1) Suppose you were hired to be on the campaign team of one of the presidential candidates (including past party nominees). Your job is to advise the candidate on what courses of actions to take over the next number of months to maximize his or her chances of winning the election. What kind of data would you try to collect, what variables, and how would you construct a basic model used to advise the candidate?
Sketch a basic model using what we have learned. Also consider the problems your potential model could have (think Gauss-Markov assumptions) and how you could try to overcome them.
2) Despite Hillary Clinton having an almost "90%" predicted chance of winning the Presidential Election according to a number of models leading up to the election, she ultimately lost the election to now president Donald Trump. What does this does say, or not say, about the limits of these types of predictions?
Explanation / Answer
Q1)
A survey could be conducted where the people can be asked for several question and based on that question an index can be develped about the relevance of social media.Also their age can be noted.Based on age(independent) and social media likeness index(dependent) variable a regression model, can be developed so that it could be determined which age is mostly using social media and this could be used as an effective way to use this platform for advertising their candidature.
The basic model is the OLS regression model with the above metioned variables.The regression model may violate the Gauss-Markov assumptions of constant error variances and uncorrelated error.It may due to the biasness in the survey data when the data collected is not randomly selected.In this case a survey is collected from a particular palce or age group and not uniformly distributed which may lead to biasness.
Q2)Correct prediction is based on statisitcs and presence of similar condition of independent variables.The polling method used to collect the data from the samples to infer the population are unreliable .The other reason is that the result is biased as most of the sample used to get the information were Clinton supporter.In this case ,inference is not true due to selection bias.
Related Questions
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.