2. Load "LJ.trans.csv" data posted in [Resources] -> [R Exercise] @ ReggieNet.

LJ.trans<-read.csv("LJ.trans.csv")

Variable description

cust.id: unique customer ID

age: customer age

email: "yes" if LJ has the customer's email address; "no" otherwise

city: living in Bloomington or Normal

sex: female or male

distance.to.store: distance between home to the closest offline store (mile)

online.visits: total number of visits to online store for a month

online.trans: total number of transactions via online store for a month

online.spending: total spending ($) in online store for a month

online.return: total product return amount ($) via online store for a month

offline.trans: total number of transactions via offline store for a month

offline.spending: total spending ($) in offline store for a month

offline.return: total product return amount ($) via offline store for a month

review: "yes" if the customer posted her/his reviews; "no" otherwise

income: income ($)

loyalty: customer segments (high, med, low) based on customer's level of loyalty

2.1. A researcher wants to predict whether customers will leave reviews based on the independent variables, email, sex, online. visits, online.return, offline. return. Please run a proper regression model and interpret all of the estimated coefficients for the independent variables. What are your suggestions that help LJ increase customers' reviews?

2.2. A researcher wants to predict the number of visits to an online store for a month based on the independent variables, age, email, sex, distance.to.store, online.spending, online.return, offline.spending, offline.return, income, loyalty. Please run a proper regression model.

2.3. Is your model for question 2.2 appropriate? More specifically, is there an overdispersion problem? If so, please resolve it and compare your results in question 2.2 with your new results in this question. Also, please interpret all of the statistically significant estimated coefficients.