Practise two important special questions: finding missing frequencies from a given median and using percentiles to make age-limit decisions.
In missing-frequency problems, first use the total frequency to form one equation. Then use the given median as \(P_{50}\) to form the second equation.
In percentile decision problems, convert the percentage into position first. Then locate the correct class using cumulative frequency.
The main skill is identifying the correct class and using the correct class boundary.
\[ \text{Median}=P_{50} \]
\[ i=\frac{p}{100}N \]
\[ P_p=L+\left(\frac{i-c.f.}{f}\right)h \]
The expenditure of 1000 families is given below. If the median of the frequency distribution is 87, find the missing frequencies.
| Expenditure | 40 - 59 | 60 - 79 | 80 - 99 | 100 - 119 | 120 - 139 |
|---|---|---|---|---|---|
| No. of families | 50 | \(x\) | 500 | \(y\) | 50 |
Convert the classes to continuous boundaries. Then use total frequency and \(P_{50}\).
The given class intervals are not continuous. So first make them continuous:
| Original Class | Continuous Class Boundary |
|---|---|
| 40 - 59 | 39.5 - 59.5 |
| 60 - 79 | 59.5 - 79.5 |
| 80 - 99 | 79.5 - 99.5 |
| 100 - 119 | 99.5 - 119.5 |
| 120 - 139 | 119.5 - 139.5 |
Let the missing frequencies be \(x\) and \(y\).
| Class Boundary | Frequency | Cumulative Frequency |
|---|---|---|
| 39.5 - 59.5 | 50 | 50 |
| 59.5 - 79.5 | \(x\) | \(50+x\) |
| 79.5 - 99.5 | 500 | \(550+x\) |
| 99.5 - 119.5 | \(y\) | \(550+x+y\) |
| 119.5 - 139.5 | 50 | \(600+x+y\) |
Since \(N=1000\),
\[ 600+x+y=1000 \]
\[ x+y=400 \]
Since median is \(87\), we treat it as \(P_{50}\).
\[ \text{Median}=P_{50} \]
\[ i=\frac{50}{100}N \]
\[ i=\frac{50}{100}\times1000=500 \]
The median value \(87\) lies in the class \(79.5-99.5\). So this is the \(P_{50}\) class.
\[ L=79.5,\qquad h=20,\qquad f=500,\qquad c.f.=50+x \]
Now apply the percentile formula:
\[ P_{50}=L+\left(\frac{i-c.f.}{f}\right)h \]
\[ 87=79.5+\left(\frac{500-(50+x)}{500}\right)20 \]
\[ 87=79.5+\left(\frac{450-x}{500}\right)20 \]
\[ 7.5=\left(\frac{450-x}{500}\right)20 \]
\[ 7.5=\frac{450-x}{25} \]
\[ 187.5=450-x \]
\[ x=262.5 \]
Since frequency must be a whole number,
\[ x=263 \]
Now use \(x+y=400\):
\[ 263+y=400 \]
\[ y=137 \]
Following is the age distribution of 1000 persons working in a factory. Due to heavy loss, the management decides to bring down the strength to 50 percent of the present number according to the scheme given below.
| Age Group | 15-20 | 20-25 | 25-30 | 30-35 | 35-40 | 40-45 | 45-50 | 50-55 | 55-60 |
|---|---|---|---|---|---|---|---|---|---|
| Number of People | 60 | 122 | 135 | 242 | 148 | 107 | 85 | 63 | 38 |
First prepare cumulative frequency. Then find \(P_8\), \(P_{40}\), and \(P_{90}\).
Prepare cumulative frequency:
| Age Group | Frequency | Cumulative Frequency |
|---|---|---|
| 15-20 | 60 | 60 |
| 20-25 | 122 | 182 |
| 25-30 | 135 | 317 |
| 30-35 | 242 | 559 |
| 35-40 | 148 | 707 |
| 40-45 | 107 | 814 |
| 45-50 | 85 | 899 |
| 50-55 | 63 | 962 |
| 55-60 | 38 | 1000 |
\[ N=1000 \]
First 8% from the lower group:
\[ i=\frac{8}{100}\times1000=80 \]
The 80th value lies in the class \(20-25\). Here \(L=20,\;c.f.=60,\;f=122,\;h=5\).
\[ P_8=20+\left(\frac{80-60}{122}\right)5 \]
\[ P_8=20.82 \]
Next 32% means the limit up to \(8\%+32\%=40\%\). So find \(P_{40}\):
\[ i=\frac{40}{100}\times1000=400 \]
The 400th value lies in the class \(30-35\). Here \(L=30,\;c.f.=317,\;f=242,\;h=5\).
\[ P_{40}=30+\left(\frac{400-317}{242}\right)5 \]
\[ P_{40}=31.71 \]
Highest 10% retire means the lower limit for that group is \(P_{90}\):
\[ i=\frac{90}{100}\times1000=900 \]
The 900th value lies in the class \(50-55\). Here \(L=50,\;c.f.=899,\;f=63,\;h=5\).
\[ P_{90}=50+\left(\frac{900-899}{63}\right)5 \]
\[ P_{90}=50.08 \]
Decision:
First 8% below \(20.82\) years are retrenched.
Next 32% from \(20.82\) years to \(31.71\) years are transferred to other branches.
Highest 10% above \(50.08\) years retire prematurely.
Therefore, the persons retained in the mill are from \(31.71\) years to \(50.08\) years.