In: Statistics and Probability
122. Suppose that a publisher conducted a survey asking adult consumers the number of fiction paperback books they had purchased in the previous month. The results are summarized in the Table 2.83.
# of books Freq. Relative Freq.
0 18
1 24
2 24
3 22
4 15
5 10
7 5
9 1
a. Are there any outliers in the data? Use an appropriate numerical test involving the IQR to identify outliers, if any, and clearly state your conclusion.
Answer:
Given that,
Suppose that a publisher conducted a survey asking adult consumers the number of fiction paperback books they had purchased in the previous month. The results are summarized in the table.
# of books | Frequency | Cumulative frequency |
0 | 18 | 18 |
1 | 24 | 42 |
2 | 24 | 66 |
3 | 22 | 88 |
4 | 15 | 103 |
5 | 10 | 113 |
7 | 5 | 118 |
9 | 1 | 119 |
Total | 119 |
From the cumulative frequency column, get the three quartiles
Q1 is the data value below which 25% of the data lies.
Position of Q1 is 119/4= 29.75th value
Position of Q2 is 2(119/4)=59.50th value
Position of Q3 is 3(119/4)=89.25th value
Thus,
29.75th value in cumulative frequency is between 0 and 1. Hence Q1=0.5
59.50th value in cumulative frequency is between 1 and 2. Hence Q2=1.5
89.25th value in cumulative frequency is between 3 and 4. Hence Q3=3.5
Interquartile range (IQR) is given by,
IQR=Q3-Q1
IQR=3.5-0.5=3
Interquartile range (IQR)=3
1.5 IQR=1.5 3 =4.5
Q1-1.5 IQR=0.5-45=-4
Q3+1.5 IQR=3.5+4.5=8
No value for number of books is less than -4.
However, there is one value which is 9, that is greater than 8.
Therefore, 9 books are an outlier value.