In: Statistics and Probability
You are a public health nurse that is interested in comparing the total blood cholesterol levels in clients from different census tracts in the city that are enrolled in a community-based lifestyle program offer by the Public Health Department. Previous research has shown that cholesterol levels are generally normally distributed, and you have no reason to believe that the variances in cholesterol levels would differ across census tracts. The table below provides the data that you collected on cholesterol levels in clients enrolled in your program from four census tracts in the city. Assuming a 0.05 level of significance, can you conclude that cholesterol levels differ across the census tracts? If there is a difference, which census tracts differ from the others?
Area A |
Area B |
Area C |
Area D |
96 |
120 |
138 |
101 |
126 |
122 |
139 |
104 |
166 |
128 |
140 |
104 |
168 |
131 |
140 |
107 |
173 |
132 |
141 |
108 |
178 |
147 |
159 |
122 |
190 |
148 |
166 |
125 |
194 |
149 |
172 |
136 |
195 |
151 |
176 |
139 |
198 |
160 |
180 |
149 |
212 |
160 |
182 |
160 |
212 |
162 |
185 |
162 |
213 |
164 |
188 |
162 |
215 |
165 |
188 |
168 |
216 |
174 |
192 |
171 |
227 |
184 |
194 |
173 |
185 |
194 |
174 |
|
188 |
200 |
175 |
|
192 |
202 |
180 |
|
194 |
203 |
182 |
|
197 |
206 |
182 |
|
198 |
210 |
184 |
|
201 |
215 |
184 |
|
221 |
220 |
190 |
|
224 |
222 |
191 |
|
230 |
224 |
192 |
|
242 |
230 |
193 |
|
250 |
231 |
194 |
|
251 |
232 |
195 |
|
233 |
200 |
||
237 |
206 |
||
243 |
207 |
||
245 |
207 |
||
263 |
209 |
||
274 |
210 |
||
294 |
212 |
||
307 |
215 |
||
327 |
221 |
||
224 |
|||
227 |
|||
229 |
|||
231 |
|||
239 |
|||
241 |
|||
248 |
|||
262 |
|||
317 |
Here we have to test,
H0 : All census tracts are insignificant.
Vs H1 : At least one census tract differ significantly.
I had done this using MINITAB as follows -
stat - one way ANOVA as follows -
It opens the following window
Then click on option & click on appropriate as follows
To check which census track differ from other I had done Tuckey's test for that go in comparison as follows-
It gives output as follows-
From above output of ANOVA table we observe that P-value is 0.295 > l.o.s(0.05) . Thus, we accept H0 indicating there is no significant difference in census tracts.
Also from Tuckey's confidence interval
We observe that all confidence interval includes 0 . Thus , we accept H0 indicating that All census tract are insignificant. i.e. no census tracts differ significantly from other.