In: Statistics and Probability
The editor of the student newspaper was in the process of making some major changes in the newspaper’s layout. He was also contemplat- ing changing the typeface of the print used. To help himself make a decision, he set up an experiment in which 20 individuals were asked to read four newspaper pages, with each page printed in a dif- ferent typeface. If the reading speed differed, then the typeface that was read fastest would be used. However, if there was not enough evidence to allow the editor to conclude that such differences existed, the current typeface would be continued. The times (in seconds) to completely read one page were recorded. What should the editor do?
Typeface 1 | Typeface 2 | Typeface 3 | Typeface 4 |
110 | 123 | 115 | 115 |
118 | 119 | 110 | 134 |
148 | 184 | 139 | 143 |
147 | 145 | 141 | 185 |
159 | 191 | 152 | 171 |
200 | 209 | 194 | 222 |
114 | 116 | 102 | 123 |
148 | 147 | 143 | 151 |
132 | 138 | 127 | 150 |
158 | 175 | 134 | 152 |
159 | 150 | 161 | 177 |
127 | 142 | 131 | 130 |
189 | 202 | 182 | 183 |
167 | 195 | 153 | 174 |
146 | 167 | 136 | 151 |
135 | 126 | 115 | 165 |
124 | 135 | 133 | 122 |
146 | 150 | 143 | 151 |
122 | 136 | 113 | 138 |
129 | 147 | 110 | 137 |
The following are the continuous observatins. So, one can use is one way ANOVA . I have used R where R code is above and output is at lower side.
The null and alternative hypotheses will be:-
In words with each typerface, reading speed is same and alternative hypotheses will be that reading speed differ significant.
Taking significance level at 0.05 here p-value is 0.0899 > 0.05 , therefore we fail to reject null hypothesis, So typerface does not effect reading speed. So, there is not enough evidence to allow the editor to conclude that such differences existed, the current typeface would be continued.
NOTE:- Here, tukey test is applied which is not necessary, if null hypothesis is rejected then for see the significant difference we apply tukey test.