|
You collect data on crime and new studentenrollment from 97 randomly selected colleges across the United States for theyear 2010. You then run the followingmodels as shown in the table below:
Where: Enroll = total new student enrollment in 2010 for a college Crime = the total number of crimes reported on a campusin 2010 ln(Crime) = log of Crime Private = 1 if the school is a private school and 0 ifit is public ln(Enrollhat) = Predicted values of Enroll from Model 1 Enrollhat = Predicted values of Enroll from Model 3 Using the information in Table 2, would youprefer to use Crime or ln(Crime) as an independent variable when Enroll is thedependent variable? What evidence isthere to support your answer?
问题是根据表中的数据 用crime还是用对数形式的ln(crime)作为自变量更好?以及为什么.
1.顺便还有一题说是你为一家学校的快餐店做问卷调查 问卷内容包括性别,种族,年龄,是否为一年级新生,当前的GPA,以及6个月内吃过这家店的次数。 原句的问题说的是 Atthe end of the week, you begin analyzing the data and notice that age, race,and sex each have some missing entries. Under what conditions would this pose and not pose a problem for yourestimation? 我不太理解这个missing entries是代表了什么, 是指有几份问卷里这几项没有填写吗? [size=14.6667px]怎么理解 然后怎么答? [size=14.6667px]
就着第二题在问两个问题
2.说如果你用得到的数据来做回归
会有一个 Clou= β0 + β1*gpai + β2*firstyeari + β3*sexi + β4*blacki + μi 这样的方程 (CLou是快餐店的名字)
用CLou作为因变量能否得到可靠的估算数据? (原文:Suppose you plan to use the data to run a regression where how many times a student haseaten at Chicken’s Lou’s in the past 6 months is the dependent variable:)
3.然后如果打算用作一个LPM线性概率模型来根据他们是否愿意吃这个店来估算他们是不是一年级新生,firstyeari = β0 + β1*CLoui + β2*gpai + β3*sexi + β4*blacki + μi
那clou作为自变量能否得到可靠的数据? (原文:Now suppose you plan to run a linear probability model examining the probability that a student is a first-year student based on their tendency to eat at Chicken Lou’s (variable definitions are the same as part B). )
我实在是不太明白他在问什么 有数据的情况下不该都可以得到可靠的答案吗?
追加两小问 追加20分
|