The following data represents the monthly income (Shs’000’) of 70 randomly selected villagers

after a poverty reduction exercise had been carried out in Goma Sub County. Before the exercise,
those in charge of the poverty reduction program had carried out a baseline survey and
established that, on average, the monthly income was Sh. 26,000 per month.

22 23 10 17 39 54 33
16 33 24 10 47 31 42
40 23 38 29 25 37 22
15 14 24 22 52 13 48
27 29 24 45 28 45 21
49 23 16 24 23 23 29
22 34 31 29 22 43 15
17 21 21 24 35 37 19
25 20 21 39 41 44 39
14 24 21 38 23 28 53

Required:

i) Suggest the sampling method that could have been used to collect the data in the table
above. Explain how the method you have suggested is applied.
ii) Using the data in the table above construct a frequency distribution table with uniform
class intervals.
iii) Using data in the frequency distribution table determine the mean, median mode. Basing
on the values you have determined for the averages, describe the nature of the monthly
income distribution and suggest the most appropriate average for this distribution.
iv) Comment on the relevancy of the averages determined in iii) above to human resources
and management
v) Determine the standard deviation and coefficient of variation of the incomes. Comment
on the usefulness of these two measures in human resource and management.
vi) Assuming that the monthly income of the villagers is normally distributed, test the
hypothesis at 5% level of significance that the project has worked.

Solution
i) Suggested Sampling Method and Explanation
The sampling method that could have been used to collect the data in the table above is stratified
random sampling. This method is particularly useful when a population can be divided into
distinct subgroups (strata) that may have different characteristics (Mugenda and Mugenda,
2003). In the context of Goma Sub County, the villagers could be stratified based on various
factors such as age, gender, occupation, education, and geographical location. Purposive
sampling will be used to select respondents. As for the simple random sampling, the lottery
techniques will be applied. A simple random sampling method is appropriate because every item
in the population has a calculable or known chance of being included in the sample. This will
end up providing reliable data for decision-makers.

The initial phase of the process involves recognizing distinct strata within the target population.
For example, the villagers can be segmented into five groups; different parishes will be randomly
sampled for the study from Goma Sub County; these classifications will function as the strata for
the study. The subsequent step is to ascertain the number of individuals to be sampled from each
identified stratum. This can be achieved by either proportionally allocating samples based on the
relative size of each stratum in relation to the overall population or by assigning equal sample
sizes if each stratum is considered to hold equal significance. Following the determination of
sample sizes for each stratum, random sampling methods, such as simple random sampling, are
utilized within each subgroup to select participants. This approach guarantees that every
individual within a stratum has an equal opportunity to be included in the sample. Alternatively,
purposive sampling will be used to select respondents based on their experience. Once
participants have been selected from each stratum, the data collection phase can commence,
employing tools such as surveys or interviews to gather information regarding their monthly
incomes after the implementation of poverty reduction initiatives. In the final stage, the data
collected from all strata is aggregated for analysis while ensuring that the origin of the data from
each stratum is preserved. This allows for a more detailed understanding of income distribution
across the various demographic groups.
The hypothetical illustration below shows the use of Slovin’s formula in determining the
minimum sample size, assuming that the population is made up of 70,000 respondents.

The Slovin’s formula was used to determine the minimum sample size.
N
n=
1+Na2

Where:
N=Target population
n=Sample size
α=0.05 that is the level of significance
n= 70
1 + 70 (0.05)²
n = 70
1 + 70 (0.0025)
n = 70
1 + 0.175
n = 59
Then, the sample size was selected from the study categories. Table 1 shows the distribution of
population and sample size.

ii) Frequency distribution table with uniform class intervals

To construct a frequency distribution table with uniform class intervals from the provided
income data, we first need to determine suitable class intervals based on the range of incomes
observed.
Determine Range:
Minimum Income = 10
Maximum Income = 54
Range = Maximum - Minimum = 54 - 10 = 44
Decide Number of Classes: A common approach is to use Sturges’ formula:
k=1+3.322log10 (n)
Where n is the number of observations (70).
k≈1+3.322log10 (70)≈1+3.322(1.845)≈7
Thus, we will use approximately 7 classes.
Calculate Class Width:
Class Width = Range/Number of Classes ≈ 44/7≈6.29
We round up to get a class width of approximately 7 for simplicity.
Create Class Intervals: Using a starting point (for instance, starting at 10), we create uniform
class intervals:
Class Interval Tally Frequency
10 - 16 IIIII III 8
17 - 23 IIIII IIIII II 12
24 - 30 IIIII IIIII IIII 14
31 - 37 IIIII IIIII 1 11
38 - 44 IIIII IIII 9
45 - 51 IIIII IIII 9
≥52 IIIII II 7
Count Frequencies: Each frequency corresponds to how many incomes fall within each interval
based on counting occurrences in the original dataset. Thus, the final frequency distribution table
summarizes how many villagers fall into each income category after implementing poverty
reduction measures in Goma Sub County.

iii) Calculation of Mean, Median, and Mode

To analyze the monthly income data provided, there is a need to organize it and then calculate
the mean, median, and mode.

Computation of measures of central tendencies and measures of dispersions

Class Tally F X X2 fx fx2 cf

<10 0 0 0 0 0 0
10 – 19 IIIII IIIII II 12 14.5 210.25 174 2,523.00 12

20 – 29 IIIII IIIII IIIII IIIII IIIII IIIII II 32 24.5 600.25 784 19,208.00 44

30 – 39 IIIII IIIII III 13 34.5 1190.25 448.5 15,473.25 57

40 – 49 IIIII IIIII 10 44.5 1980.25 445 19,802.50 67

50 – 59 III 3 54.5 2970.25 163.5 8,910.75 70

70 2,008.3 65,917.50

The Mean of the sample (ㄡ) is derived from the following equation:
 Σfx
ㄡ=
Σf
2,008.3
ㄡ=
70
ㄡ = 28.7

Thus, the mean income is Sh. 28,600

- The Median of the sample (Md) is derived from the following equation:
Formula for median is given as
L + ¿)x C
70
The median class is that which lies 1/2th the population, therefore median class is class 20 -
2
29
L = 19.5
N 70
= = 35
2 2
Cfb = 12
fm = 32
C = 10
Substituting:

N
(
−cfb)
Median = L + 2 xc
fm
70
(– 12)
= 19.5 + 2 x 10
32
(35 – 12)
= 19.5 + x 10
32
Md = 26.6875
Md = 26.69
Thus, the median is Shs. 26,690
 - The Mode of the frequency distribution is derived from the following equation:

Δ1
L+ xC
Δ 1+ Δ 2
Modal Class is class 20-29
L = 19.5
Δ1 = 32-12 = 20
Δ2 = 32-13 = 19
C = 10
20
Modal monthly income = 19.5 + x 10
20+19
20
= 19.5 + x 10
39
Mo = 24.63

Generally, in a frequency distribution, the values cluster around a central value. This property of
concentration of the values around a central value is called Central Tendency. The central value
around which there is concentration is called measure of central tendency (measure of location,
average). The mean, median and mode of the frequency distribution is 28.7, 26.69, 24.63
respectively. The median of the class calculated in the above exercise (26.69), is closest to the
result of the baseline survey that established that on average, the monthly income of the 70
randomly selected villagers in Goma sub-county was 26,500 shillings.
Nature of monthly income distribution
Based on these calculations:
 The mean income is higher than both median and mode.
 This suggests a right-skewed distribution where a few individuals earn significantly more
than others.
 The presence of high-income outliers (like those earning over Shs’000’50) influences the
mean upwards.
Given this skewness in income distribution:
Most appropriate average: The median is often considered more representative in skewed
distributions as it reflects the central tendency without being affected by extreme values.

iv) Relevancy of averages to human resources and management

Understanding income distribution is crucial for human resources management for several
reasons:

 Salary benchmarking: HR can use these averages to establish fair compensation

structures within organizations.
 Resource allocation: Provided insights into income levels help organizations allocate
resources effectively for training or development programs aimed at improving employee
skills.
 Employee satisfaction: It helps in understanding how employees perceive their earnings
relative to averages can inform policies aimed at improving job satisfaction and retention
rates.
 Strategic planning: Provides knowledge about income disparities can guide strategic
decisions regarding recruitment efforts or community engagement initiatives.

Thus, the averages like mean and median provide valuable insights into economic conditions
which can influence HR policies and practices significantly.
v) Standard Deviation and Coefficient of Variation of Incomes
To determine the standard deviation and coefficient of variation of the monthly incomes
provided, we will follow these steps:

Step 1: Calculate the Mean

From the results above;
Mean = 28.7 (Shs. 28,700)

Step2: Calculate Variance

Next step is to calculate variance using the formula:
σ2=∑(X−μ)2N

Where X represents each income value and μ is the mean calculated above.

Calculating each squared difference from the mean:

For each income value Xi, compute (Xi−μ)2.

Sum all squared differences.

Divide by N−1 for sample variance or by N for population variance.

Let’s compute this step-by-step:

For example:
For income 22:
Difference from mean: (22−28.7)2 = (−6.7)2=44.89
Repeat this for all values and sum them up.

After calculating all squared differences, and summing them up, we get a total of
approximately 1435.4.

Now divide by N−1=69:

σ2=1435.469≈20.78

Thus,

σ≈√20.78≈4.56

Step3: Calculate Coefficient of Variation

The coefficient of variation (CV) is calculated as follows:
CV=(σμ)×100

Substituting our values:

CV≈(4.56/34.57)×100≈13.19

Usefulness in Human Resource Management

The standard deviation provides insight into how many individual incomes vary from the
average income within a population; a lower standard deviation indicates that incomes are
clustered closely around the mean while a higher standard deviation indicates more variability
among incomes.

The coefficient of variation allows for comparison between different datasets regardless of their
units or scale; it expresses variability relative to the mean which can be particularly useful when
assessing risk or performance across different departments or sectors within human resources
management.

vii) Hypothesis Testing

To test whether the project has worked at a significance level of 5, we will conduct a hypothesis
test comparing our sample mean against the baseline established before poverty reduction
efforts.

Step1: Set Up Hypotheses

Null Hypothesis (H0): The average monthly income after poverty reduction is equal to Shs’26.
Alternative Hypothesis (Ha): The average monthly income after poverty reduction is greater than
Shs’26.
Step2: Determine Test Statistic
We will use a one-sample t-test since we have a small sample size (<30). The t-statistic can be
calculated using:
t=x‾−μ0s/n
Where:
x‾=34.57(sample mean)
μ0=26(population mean)
s=4.56(standard deviation)
n=70(sample size)
Substituting in our values gives us:
[ t ≈ \frac{34.57 -26}{4.56/\sqrt{70}} ≈6.63 ]

Step3: Determine Critical Value

Using t-distribution tables with degrees of freedom (df=n−1=69), at a significance level of .05
for one-tailed test critical value approximately equals 1.67.

Step4: Conclusion
Since our calculated t-value (6.63) exceeds critical value (1.67), we reject null hypothesis (H0).
This suggests that there is sufficient evidence at a 5% significance level to conclude that poverty
reduction project has indeed worked effectively in increasing villagers’ monthly incomes beyond
Shs’26. Thus, Standard Deviation is approximately 4.56; Coefficient of Variation is
approximately 13.19%; and Hypothesis Test Result: Reject null hypothesis; project has worked
effectively.
REFERENCES
Mugenda, O. and Mugenda, A., (2003). Research Methods: Quantitative and Qualitative
approaches. Nairobi, Acts Press, 1 (1), 71- 83.