Acensus around US was undertaken in 2000. Sixteen samples are randomly drawnfrom the cities over 2 million people around US. The sample data are shown asfollows:

v1=City
v2=Estimated Percent Undercounted
v3=Minority percent (Black or Huspanic)
v4=Crime rate per 1000 people (serious crimes)
v5=Poverty Percent
v6=Percent having trouble reading or speaking English
v7=Percent over 25 not completed  high school
v8=Percent of housholds in multiunit buildings

v9=Citygeological location categories (1.West,2.Midwest, 3.mideast, 4.east.)

v1	v2	v3	v4	v5	v6	v7	v8	V9
Baltimore	6.15	55.5	100	22.9	0.7	5.16	23.3	4
Boston	2.27	28.4	135	20.2	4.4	3.16	52.1	4
Chicago	5.42	53.7	66	20.3	6.7	4.38	51.4	3
Cleveland	5.01	46.7	101	22.1	1.6	4.91	36.4	4
Dallas	8.18	41.6	118	14.2	3.1	3.15	12.9	2
Detroit	4.33	65.4	106	21.9	1.6	4.58	18.6	3
Houston	5.79	45.1	80	12.7	5.1	3.16	8.9	2
Indianapolis	0.31	22.5	53	11.5	0.3	3.33	13.6	3
Los.Angeles	7.52	44.4	100	16.4	12.7	3.14	15	1
Milwaukee	3.17	27.2	65	13.8	1.6	4.64	27.2	3
New.York.City	7.39	44	101	20	8.9	3.98	32.2	4
Philadelphia	6.41	41.3	60	20.6	2.2	4.57	21.7	4
Saint.Louis	3.6	46.7	143	21.8	0.5	5.18	40.9	3
San.Diego	0.47	23.6	81	12.4	4.2	2.11	11.2	1
San.Francisco	5.18	24.8	107	13.7	9.2	2.6	20.3	1
Washington.DC	5.93	72.6	102	18.6	1.1	3.29	21	4

Thereare a lot of controversies on whether the census undercount varies bygeographical location. This is important to determine because census countsaffect representation in state and federal legislatures and the distribution oftax dollars back to localities. Are some cities being "victimized" bythe undercount? Are certain groups being "victimized" as well? We aregoing to use these data to try to address these questions.

我是统计菜鸟，拿到这个题目根本不明白从何下手，请高手指点一下