SAMPLE SIZE CALCULATION

Prepared by:

Kalonga Mwiinga (PhD).

• Why does sample size matter?
• Is a critical step in research design.
• Ensures that studies are:
• Scientifically valid
• Resource-efficient
• Choosing an appropriate sample size affects:
• Precision & Reliability of results
• Generalizability of study results.

2
• Definition
• Sample size refers to the number of observations
or participants included in a study.
• Simply put, the sample size (often denoted as 'n')
represents the number of individual units (e.g.,
people, animals, cells, products) that are selected
from a larger population to be part of the
research investigation.
• It is a key factor in determining whether the study
can produce:
• meaningful results
• statistically significant results.
3
• Importance of Sample Size
• The size of the sample has a significant impact
on the study's ability to provide meaningful
answers to the research questions.
• The consequences of having a sample size that is
either too small or too large are substantial:

4
• Importance of Sample Size
• Too small: May not detect true effects
(underpowered).
• A study with an insufficient sample size is said
to be underpowered.
• This means that even if a real effect or
difference exists in the population, the study
might lack the statistical "muscle" to detect it.
• Imagine trying to see a faint star with a weak
telescope – it might be there, but you won't be
able to see it clearly.
5
• Importance of Sample Size
• Too small: May not detect true effects
(underpowered).
• Underpowered studies are more likely to
produce false negatives (Type II errors).
• This is where we fail to reject a null
hypothesis that is actually false.
• Here, a real effect is missed.
• In practical terms, this could mean missing a
potentially significant market trend.
6
• Importance of Sample Size
• Too small: May not detect true effects
(underpowered).
• This can lead to:
• wasted time
• wasted resources
• potentially incorrect conclusions
• Hinders:
• scientific progress
• informed decision-making. 7
• Importance of Sample Size
• Too large: Wastes resources and may be
unethical.
• Conversely, a sample size that is unnecessarily
large can lead to a waste of resources.
• This includes:
• financial costs associated with recruiting and
testing participants.
• The. time and effort of researchers.

8
• Importance of Sample Size
• Too large: Wastes resources and may be unethical.
• More importantly, an excessively large sample size
can raise ethical concerns.
• If a smaller sample could adequately answer the
research question, exposing more participants to
the potential risks or burdens of the study
without additional benefit is ethically
questionable.
• For instance, in clinical trials, unnecessarily large
samples might expose more individuals to a new
treatment with unknown side effects when a
smaller group could have provided sufficient
evidence of its efficacy. 9
• Importance of Sample Size
• Too large: Wastes resources and may be
unethical.
• Furthermore, with very large samples, even
trivial or practically insignificant differences
can become statistically significant.
• This leads to misleading interpretations of the
findings.

10
• Importance of Sample Size
• "Just right": Balances statistical power with
resource constraints.
• The goal of sample size calculation is to find the
optimal balance between having enough
statistical power to detect a real effect (if it
exists) and the practical limitations of
resources, time, and ethical considerations.

11
• Importance of Sample Size
• "Just right": Balances statistical power with
resource constraints.
• A well-calculated sample size ensures that the
study has a high probability of detecting a
meaningful effect while minimizing waste and
ethical concerns.
• This leads to more reliable and impactful
research findings.

12
• Core Concepts in Sample Size Calculation
• Population
• The entire group of individuals, objects, or
events that are of interest in the study
• About which the researchers want to draw
conclusions.
• For example, if a researcher is studying the
prevalence of diabetes in Zambia, the
population would be all individuals living in
Zambia.
13
• Core Concepts in Sample Size Calculation
• Sample
• A subset of the population that is selected and
actually studied.
• Researchers collect data from the sample and
then use statistical methods to make inferences
or generalizations about the entire population.
• Ideally, the sample should be representative of
the population to ensure that the findings can
be reliably extrapolated.
14
• Core Concepts in Sample Size Calculation
• Statistical power
• The probability of detecting an effect when it truly
exists in the population.
• It is often denoted as (1 - β).
• where β is the probability of making a Type II
error (failing to reject a false null hypothesis).
• Higher statistical power (typically set at 0.8 or
80%) means a lower chance of missing a real
effect.
• Power is directly influenced by sample size; larger
samples generally lead to higher power. 15
• Core Concepts in Sample Size Calculation
• Effect size
• The magnitude of the difference or relationship that
the researcher is trying to detect in the population.
• It quantifies the practical significance of the findings.
• A larger effect size is easier to detect with a smaller
sample size.
• A smaller effect sizes require larger samples to
achieve adequate power.
• Effect size can be expressed in various ways depending
on the type of data and statistical analysis (e.g.,
Cohen's d for the difference between two means,
correlation coefficient r for the strength of a
relationship). 16
• Core Concepts in Sample Size Calculation
• Sampling Error:
• Is the natural variation or discrepancy that
inevitably arises between a sample statistic (a
value calculated from a sample, such as the
sample mean or sample proportion) and the
corresponding population parameter (the true
value in the entire population, such as the
population mean or population proportion).

17
• Core Concepts in Sample Size Calculation
• Sampling Error:
• Imagine trying to estimate the average height of
all adults in Zambia by measuring the heights of
100 randomly selected individuals in Lusaka.
• The average height you calculate from your
sample of 100 might be slightly higher or lower
than the true average height of all adults in
Zambia.
• This difference is due to sampling error – your
sample, while hopefully representative, is not a
perfect mirror of the entire population.
18
• Core Concepts in Sample Size Calculation
• Sampling Error:
• Key Point: Sampling error is inherent in
sampling and cannot be completely eliminated
unless we study the entire population (which is
often impractical or impossible).
• However, we can reduce sampling error by
increasing the sample size.
• Larger samples tend to provide more accurate
estimates of population parameters because
they are more likely to be representative of the
whole group.
19
• Core Concepts in Sample Size Calculation
• Confidence Level
• Confidence level is the degree of certainty you
want to have that your sample results
accurately reflect the true population
parameter.
• It is usually expressed as a percentage (e.g.,
90%, 95%, 99%).

20
• Core Concepts in Sample Size Calculation
• Confidence Level
• A 95% confidence level means that if you were to
take many random samples from the same
population & calculate a confidence interval for
each sample, approximately 95% of those
intervals would contain the true population
parameter.

21
• Core Concepts in Sample Size Calculation
• Confidence Level
• If you construct a 95% confidence interval for
the average height of adults in Zambia based
on your sample, you can be 95% confident that
the true average height of all Zambian adults
falls within that calculated interval.

22
• Core Concepts in Sample Size Calculation
• Confidence Level
• A higher desired confidence level generally
requires a larger sample size.
• To be more certain that your sample
accurately represents the population, you need
to gather more information.

23
• Core Concepts in Sample Size Calculation
• Z-Score
• A Z-score (also known as a standard score) is a
measure of how many standard deviations a
particular data point is away from the mean of
its distribution.
• In the context of sample size calculation for
proportions (when the population distribution
can be approximated by a normal distribution),
the Z-score is linked to the chosen confidence
level.
24
• Core Concepts in Sample Size Calculation
• Z-Score
• For a given confidence level, there is a
corresponding Z-score that defines the
boundaries of that level under the standard
normal distribution.
• Common Z-Scores for Typical Confidence Levels:
• 90% Confidence Level → Z ≈ 1.645
• 95% Confidence Level → Z ≈ 1.96
• 99% Confidence Level → Z ≈ 2.576
25
• Core Concepts in Sample Size Calculation
• Role in Sample Size Calculation:
• The Z-score is used in formulas for calculating
the margin of error and subsequently the
required sample size.
• especially when dealing with proportions.
• It essentially scales the standard error (the
standard deviation of the sample statistic) to
match the desired level of confidence.

26
• Core Concepts in Sample Size Calculation
• Significance Level (α or alpha):
• The significance level (α) is the probability of
rejecting a true null hypothesis.
• The null hypothesis is a statement of no effect
or no difference in the population.
• It is the threshold we set for deciding whether
our observed results are statistically
significant.

27
• Core Concepts in Sample Size Calculation
• Significance Level (α or alpha):
• The significance level is commonly set at:
• α = 0.05 (5%) (most frequent in research)
• α = 0.01 (1%) for stricter thresholds (high stake
studies like clinical trials.
• Its role is to determine the threshold for
statistical significance.
• If the p-value ≤ α, we reject H₀.

28
• Core Concepts in Sample Size Calculation
• Significance Level (α or alpha):
• By setting a significance level (commonly 0.05
or 5%), we are saying that we are willing to
accept a 5% chance of incorrectly concluding
that there is an effect (rejecting a true null
hypothesis).
• This incorrect rejection is known as a Type I
error (also called a false positive).

29
• Core Concepts in Sample Size Calculation
• Significance Level (α or alpha)
• Inverse relationship with confidence level:
• The significance level and confidence level are
related.
• A confidence level of (1 - α) corresponds to a
significance level of α.
• For example, a 95% confidence level
corresponds to a significance level of 0.05 (1 -
0.95 = 0.05).
30
• Core Concepts in Sample Size Calculation
• Margin of Error (E)
• The margin of error (E), also known as the
confidence interval width or sampling error
tolerance, is the maximum difference you are
willing to accept between your sample statistic
and the true population parameter.
• It is usually expressed as a plus or minus value
(e.g., ±3%, ±5 units).

31
• Core Concepts in Sample Size Calculation
• Margin of Error (E)
• The margin of error creates a range around your
sample statistic within which you are
reasonably confident the true population
parameter lies (based on your chosen
confidence level).
• A smaller margin of error indicates a more
precise estimate.

32
• Core Concepts in Sample Size Calculation
• Margin of Error (E)
• If a survey reports that 55% of Zambians support
a particular policy with a margin of error of
±3%, it means that the true percentage of
support in the entire Zambian population is
likely to be between 52% and 58%.

33
• Core Concepts in Sample Size Calculation
• Margin of Error (E)
• The margin of error is a key component in
sample size formulas.
• Researchers often decide on an acceptable
margin of error based on the practical
significance of their findings.
• A smaller desired margin of error will
necessitate a larger sample size.

34
• Core Concepts in Sample Size Calculation
• Population Proportion (p):
• Population proportion (p) is the estimated
proportion of individuals in the population who
possess a specific characteristic of interest.
• This is relevant when studying categorical
variables (e.g., the proportion of people who
support a certain political party, the proportion
of defective products).
• When calculating the required sample size for
estimating a population proportion, we need an
estimate of this proportion.
35
• Core Concepts in Sample Size Calculation
• Population Proportion (p)
• Dealing with Uncertainty:
• If you have a prior estimate: You can use the
proportion observed in previous studies or
pilot studies.

36
• Core Concepts in Sample Size Calculation
• Population Proportion (p)
• Dealing with Uncertainty:
• If you have no prior estimate:
• The most conservative approach is to use p = 0.5
(or 50%).
• This value maximizes the required sample size
because the variance of a proportion is highest
when p = 0.5.
• Using this "worst-case scenario" ensures that
your sample size will be large enough to achieve
the desired margin of error regardless of the
actual population proportion. 37
• Core Concepts in Sample Size Calculation
• Population Proportion (p)
• The estimated population proportion (p) and its
complement (q = 1 - p) are directly used in the
sample size formula for proportions.

38
• Core Concepts in Sample Size Calculation
• Variability (Standard Deviation, σ or s):
• Variability refers to the extent to which
individual data points in the population (σ) or
sample (s) differ from the average (mean).
• It measures the spread or dispersion of the
data.

39
• Core Concepts in Sample Size Calculation
• Variability (Standard Deviation, σ or s):
• High Variability: If the data points are widely
spread out, it means there is a lot of individual
difference.
• In this case, a larger sample size is needed to
obtain a stable and representative estimate of
the population mean.

40
• Core Concepts in Sample Size Calculation
• Variability (Standard Deviation, σ or s):
• Low Variability: If the data points are clustered
closely around the mean, the population is
more homogeneous.
• A smaller sample size might be sufficient to
achieve a precise estimate.

41
• Core Concepts in Sample Size Calculation
• Variability (Standard Deviation, σ or s):
• Using σ vs. s:
• If the population standard deviation (σ) is
known, it should be used.
• However, this is often not the case.
• More commonly, we use the sample standard
deviation (s) as an estimate of the population
standard deviation.

42
• Core Concepts in Sample Size Calculation
• Variability (Standard Deviation, σ or s):
• The standard deviation (σ or s) is a crucial
component in sample size formulas for
estimating means.
• Higher variability requires a larger sample size
to reduce the impact of this spread on the
precision of the estimated mean.

43
• Core Concepts in Sample Size Calculation
• Final thoughts on understanding importance of
Sample Size Core Concepts
• Understanding these concepts is fundamental to
grasping the logic and mechanics of sample
size calculation.
• By carefully considering each of these factors,
researchers can determine an appropriate
sample size that balances statistical rigor with
practical constraints.
• This ultimately leads to more reliable and
meaningful research findings.
44
• Sample Size Calculation Formulars
• (1) Descriptive Study
• Example: Surveying a population/group
• Imagine a Lusaka-based entrepreneur wants to
estimate the average monthly sales of small
grocery shops in Kalingalinga.

45
• Sample Size Calculation Formulars
• (1) Descriptive Study
• Key Formula:

46
• Sample Size Calculation Formulars
• (1) Descriptive Study
• Assume:
• Confidence = 95%
• Margin of error = 5%

• Estimated proportion = 0.5

• Therefore, you would need a sample of 384
grocery shops. 47
• Sample Size Calculation Formulars
• (1) Descriptive Study
• Adjusting for finite population:
• If the population size (N) is small, the
calculated sample size might be a significant
portion of the population.
• This calls for adjustment.
• For example, if we were surveying employees in
a small Zambian company (N=500):

48
• Sample Size Calculation Formulars
• (1) Descriptive Study
• Adjusting for finite population:

49
• Sample Size Calculation Formulars
• (2) Comparative Study (2 independent groups)
• Example:
• A Zambian retail chain wants to compare
average monthly sales (in Zambian Kwacha)
between two supermarket branches in Lusaka
(Shoprite vs. Pick n Pay).
Formula for comparing Means:

50
• Sample Size Calculation Formulars
• (2) Comparative Study (2 independent groups)
• Example:
Where:

51
• Sample Size Calculation Formulars
• (2) Comparative Study (2 independent groups)
• Example 2:
• You want to compare the proportion of farmers
using organic vs. chemical fertilizer in Central
Province.
• Assumption:

52
• Sample Size Calculation Formulars
• (2) Comparative Study (2 independent groups)
• Example 2:

53
• Sample Size Calculation Formulars
• (2) Comparative Study (2 independent groups)
• Example 2:

54
• Sample Size Calculation Formulars
• (3) Correlational Study
• Example 2:
• You want to examine if there’s a relationship
between advertising budget and sales in
Lusaka’s Chain stores.
• Formula:

55
THANK YOU
Kalonga Mwiinga (Ph. D)

+260977702254/0976619869

[email protected]

[email protected]

www.unza.zm

56