Question #1 You want to evaluate the effects of a placement policy for immigrant
ID: 3295343 • Letter: Q
Question
Question #1 You want to evaluate the effects of a placement policy for immigrants. You are interested in investigating if it is better or worse to place newly arrived immigrants in areas where many earlier immigrants already resides. You are interested in the effects of the placement on the immigrants' future earnings. Your have been informed that when an immigrant arrives, a government officer decides where the immigrant should be placed. The officer makes the decision based on information on the immigrants' country of origin, age, marital status, education and the results from a host-country language test. The officer does not meet the immigrant before making the decision.Explanation / Answer
Solution:
In this case since there is random selection we can do sampling regression analysis to judge the percentage of error in the wrong evaluation score . Since the data has been randomly changed among the sample units . So the average statistics of the data remains more or less same . Random samples drawn and checking the percentage error from the parameter values would still give us a good evaluation .
checking the percentage error means: Percentage error is the deviation of the values we estimate from the actual values ..... Suppose first you have the actual values even if they are flawed , and you calculate the mean of the values . That mean is parameter . Then you take up a sample from the actual values randomly and calculate the mean for the sample , that is statistic . The %diff. between these two means is the error .