In: Statistics and Probability
Specify completely two potential stratification variables for
grade goal; describe all levels of the variables(Levels refer to
the various categories. E.g. if you stratified the year according
to months, the levels would be Jan, Feb,..., Dec.). Discuss one
advantage and one disadvantage of stratifying in comparison to
simple random sampling.
index | goal |
1 | 100 |
2 | 100 |
3 | 100 |
4 | 100 |
5 | 100 |
6 | 100 |
7 | 100 |
8 | 100 |
9 | 100 |
10 | 100 |
11 | 100 |
12 | 100 |
13 | 100 |
14 | 100 |
15 | 100 |
16 | 100 |
17 | 100 |
18 | 100 |
19 | 100 |
20 | 100 |
21 | 99 |
22 | 98 |
23 | 95 |
24 | 95 |
25 | 95 |
26 | 95 |
27 | 95 |
28 | 95 |
29 | 95 |
30 | 95 |
31 | 95 |
32 | 90 |
33 | 90 |
34 | 90 |
35 | 90 |
36 | 90 |
37 | 90 |
38 | 90 |
39 | 90 |
40 | 90 |
41 | 90 |
42 | 90 |
43 | 90 |
44 | 90 |
45 | 90 |
46 | 90 |
47 | 90 |
48 | 90 |
49 | 90 |
50 | 90 |
51 | 90 |
52 | 90 |
53 | 90 |
54 | 90 |
55 | 90 |
56 | 90 |
57 | 90 |
58 | 90 |
59 | 90 |
60 | 90 |
61 | 90 |
62 | 90 |
63 | 90 |
64 | 90 |
65 | 90 |
66 | 90 |
67 | 89 |
68 | 89 |
69 | 88 |
70 | 88 |
71 | 88 |
72 | 87 |
73 | 87 |
74 | 86 |
75 | 86 |
76 | 86 |
77 | 86 |
78 | 86 |
79 | 85 |
80 | 85 |
81 | 85 |
82 | 85 |
83 | 85 |
84 | 85 |
85 | 85 |
86 | 85 |
87 | 85 |
88 | 85 |
89 | 85 |
90 | 85 |
91 | 85 |
92 | 85 |
93 | 85 |
94 | 85 |
95 | 85 |
96 | 85 |
97 | 85 |
98 | 85 |
99 | 85 |
100 | 85 |
101 | 85 |
102 | 85 |
103 | 85 |
104 | 85 |
105 | 85 |
106 | 85 |
107 | 85 |
108 | 85 |
109 | 85 |
110 | 85 |
111 | 85 |
112 | 85 |
113 | 85 |
114 | 85 |
115 | 85 |
116 | 85 |
117 | 85 |
118 | 85 |
119 | 85 |
120 | 85 |
121 | 85 |
122 | 85 |
123 | 85 |
124 | 85 |
125 | 85 |
126 | 85 |
127 | 85 |
128 | 85 |
129 | 85 |
130 | 85 |
131 | 85 |
132 | 85 |
133 | 85 |
134 | 85 |
135 | 85 |
136 | 85 |
137 | 85 |
138 | 85 |
139 | 85 |
140 | 85 |
141 | 85 |
142 | 85 |
143 | 85 |
144 | 85 |
145 | 85 |
146 | 83 |
147 | 83 |
148 | 83 |
149 | 82 |
150 | 80 |
151 | 80 |
152 | 80 |
153 | 80 |
154 | 80 |
155 | 80 |
156 | 80 |
157 | 80 |
158 | 80 |
159 | 80 |
160 | 80 |
161 | 80 |
162 | 80 |
163 | 80 |
164 | 80 |
165 | 80 |
166 | 80 |
167 | 80 |
168 | 80 |
169 | 80 |
170 | 80 |
171 | 80 |
172 | 80 |
173 | 80 |
174 | 80 |
175 | 80 |
176 | 80 |
177 | 80 |
178 | 80 |
179 | 80 |
180 | 80 |
181 | 80 |
182 | 80 |
183 | 80 |
184 | 80 |
185 | 80 |
186 | 80 |
187 | 80 |
188 | 80 |
189 | 80 |
190 | 80 |
191 | 80 |
192 | 80 |
193 | 80 |
194 | 80 |
195 | 80 |
196 | 80 |
197 | 80 |
198 | 80 |
199 | 80 |
200 | 80 |
201 | 80 |
202 | 80 |
203 | 80 |
204 | 80 |
205 | 80 |
206 | 80 |
207 | 80 |
208 | 80 |
209 | 80 |
210 | 80 |
211 | 80 |
212 | 80 |
213 | 80 |
214 | 80 |
215 | 80 |
216 | 78 |
217 | 75 |
218 | 75 |
219 | 75 |
220 | 75 |
221 | 75 |
222 | 75 |
223 | 75 |
224 | 75 |
225 | 75 |
226 | 75 |
227 | 75 |
228 | 75 |
229 | 75 |
230 | 73 |
231 | 73 |
232 | 70 |
233 | 70 |
234 | 70 |
235 | 70 |
236 | 70 |
237 | 70 |
238 | 70 |
239 | 70 |
240 | 70 |
241 | 70 |
242 | 70 |
243 | 70 |
244 | 70 |
245 | 70 |
246 | 70 |
247 | 70 |
248 | 68 |
249 | 62 |
250 | 60 |
251 | 50 |
We are having data on index and goal.
(1) stratification variable
(i) firstly consider an index variable which must have some order i.e. it is an ordinal scale. Then just divide the total number of observation(251) in some say 5 catagories each having 50 approximately observations hence this can be a stratification factor.
(ii) now sort the data on 'goal' find the quartiles say Q1 Q2 and Q3 which will divide the data into four parts then just treat them as stratas.
ADVANTAGE: stratification accurately reflects the population. In short it insures the all subgroups of population to be considered it does not ignore any part of population which is not exactly done in case of simple random sampling. In simple random sampling there is no guarantee that the all subgroups of population might be included in sample.
e.g in income study of a society simple random sampling draws a sample which may or may not include the highest and lowest income classes but in stratified sampling first make stratas according to there income then draw a sample from each stratum which insures that all subgroups have been considered.
DISADVANTAGE: Time consuming and tedious
Stratified sampling requires several conditions like non overlapping stratas, each unit must belong to one and only one strata, heterogeneity between strata and homogeneity within strata etc. And hence it is time consuming and tedious on the other hand simple random sampling is not that time consuming it exicuted with less efforts.