MEASURES OF CENTRAL TENDENCY IN STATISTICS
Measures of central tendency in statistics are summary measures that represent the
center or typical value of a data set. They describe where the data tends to cluster and
provide a single value that summarizes the entire distribution. The three main measures
of central tendency are the mean, median, and mode.
• The mean (arithmetic mean) is the sum of all data values divided by the number of
values. It is the most commonly used average and represents a balance point in the
data set. However, the mean is sensitive to outliers and skewed data, which can
distort the value.
• The median is the middle value that separates the higher half from the lower half of
the data when the values are arranged in order. For an odd number of data points,
it is the center value, and for an even number, it is the average of the two middle
values. The median is better than the mean for skewed distributions and ordinal
data.
• The mode is the most frequently occurring value in the data set. It is the only measure
of central tendency that can be used with nominal (categorical) data. Some data
sets may have more than one mode or none at all.
Each measure provides a different perspective on what is considered the "center" of the
data, and the best choice depends on the type and distribution of the data. For normally
distributed continuous data, the mean, median, and mode often coincide. For skewed
distributions, the median is more reliable. For categorical data, the mode is the
appropriate choice.
Additional measures exist, such as the geometric mean and harmonic mean, which are
used in specific contexts, but mean, median, and mode are the most fundamental and
widely used measures of central tendency
ARITHMETIC, WEIGHTED AND GEOMETRIC MEAN
The arithmetic mean, weighted mean, and geometric mean are three different types of
averages used in statistics and data analysis. They differ in how they are calculated and
the types of data or scenarios in which they are most appropriate.
Arithmetic Mean
• The arithmetic mean is the simple average of a set of values.
• It is calculated by summing all the values and then dividing by the number of values.
𝑥1 +𝑥2 +⋯+𝑥𝑛
• Formula: Arithmetic Mean = where 𝑥𝑖 are the data values and 𝑛 is the
𝑛
number of values.
• It is useful when all values contribute equally and is sensitive to extreme values
(outliers).
Weighted Mean
• The weighted mean accounts for the varying importance (weights) of different
values.
• Each value is multiplied by a weight representing its importance or frequency, and
then the sum of these weighted values is divided by the sum of the weights.
𝑤1 𝑥1 +𝑤2 𝑥2 +⋯+𝑤𝑛 𝑥𝑛
• Formula: Weighted Mean = 𝑤1 +𝑤2 +⋯+𝑤𝑛
where 𝑤𝑖 are the weights and 𝑥𝑖 are the
data values.
• It is used when data points contribute unequally, such as grades with different credit
units or survey results with varying respondent weights.
Geometric Mean
• The geometric mean is the nth root of the product of 𝑛 values.
• It is calculated by multiplying all data values together and then taking the root
corresponding to the number of values.
1
• Formula: Geometric Mean = 𝑛√𝑥1 × 𝑥2 × … × 𝑥𝑛 = (𝑥1 × 𝑥2 × … × 𝑥𝑛 )𝑛
• It is appropriate for data that are multiplicative or growth rates, such as investment
returns or population growth.
• The geometric mean is always less than or equal to the arithmetic mean and
handles skewed data better.
• For weighted geometric mean, weights are applied as
1
𝑤 𝑛
exponents: Weighted Geometric Mean = (∏𝑛𝑖=1 𝑥𝑖 𝑖 )∑𝑖=1 𝑤𝑖
Key Differences
• Arithmetic mean uses addition and division; geometric mean uses multiplication
and roots.
• Weighted mean adjusts for different importance of values, unlike arithmetic mean.
• Geometric mean is better for rates and multiplicative data and reduces the impact
of very high or low values.
Summary Table
Mean Type Calculation Method Use Case Sensitivity to Outliers
Arithmetic Sum of values ÷ Equal weight data High sensitivity
Mean number of values points
Weighted Sum of (value × Unequal weight Moderate sensitivity
Mean weight) ÷ sum of data points (depends on weights)
weights
Geometric nth root of product of Growth rates, Less sensitive than
Mean values ( (∏ 𝑥𝑖 )1/𝑛 ) multiplicative data arithmetic mean
MEASURES OF DISPERSION OR VARIATION
Measures of dispersion, also known as measures of variation, quantify the extent to which
data values in a dataset are spread out or scattered around a central point. These
measures help to understand the variability, consistency, and diversity within the data,
providing a clearer picture of its distribution beyond just the central tendency.
Definition and Purpose
• Measures of dispersion describe how much the values in a dataset deviate from the
central measure (like the mean).
• A dispersion measure equals zero if all data points are identical, and increases as
the data becomes more spread out.
• Dispersion is important for comparing the variability between different datasets and
identifying outliers or inconsistency.
Types of Measures of Dispersion
Absolute Measures of Dispersion
• These have the same units as the original data.
• Common absolute measures:
o Range: The difference between the maximum and minimum values. It gives a
quick sense of the spread but is sensitive to extreme values.
o Variance: The average of the squared deviations from the mean. It measures
how data points vary from the mean.
o Standard Deviation: The square root of variance, expressing dispersion in the
original units. It is widely used for normal distributions.
o Mean Deviation (Mean Absolute Deviation): The average of the absolute
deviations of data points from the mean or median.
o Quartile Deviation (Interquartile Range): Half the difference between the third
quartile (Q3) and first quartile (Q1), representing the spread of the middle 50%
of data.
Relative Measures of Dispersion
• These are unitless ratios or percentages that allow comparison between datasets
with different units or scales.
• Examples include the coefficient of variation, which is the ratio of standard
deviation to the mean, useful for comparing relative variability.
Measures of dispersion are essential for understanding the spread or variability in
data. While measures like the range provide crude estimates, more robust measures such
as variance, standard deviation, and quartile deviation offer detailed insights into data
variability and distribution characteristics. Measures of dispersion complement measures
of central tendency to provide a comprehensive statistical summary of data
• Use absolute measures (e.g., range, standard deviation) when focusing on the
spread within a single dataset measured in meaningful units.
• Use relative measures (e.g., coefficient of variation) when comparing variability
between multiple datasets with different units or scales, enabling normalized
comparisons.
Here are more detailed real-world scenarios illustrating the use of absolute and relative
measures of dispersion:
Absolute Measure of Dispersion Scenario
• In a company's salary analysis, the HR department wants to understand how much
individual salaries vary around the average salary.
• Suppose the average salary is $50,000 and the standard deviation is $7,000.
• This standard deviation is an absolute measure of dispersion, indicating that on
average, salaries deviate from the mean by $7,000.
• It helps HR see the extent of salary spread in actual dollar terms, useful for budgeting,
salary adjustments, or compensation fairness reviews.
• Here, using the absolute measure (standard deviation in dollars) makes sense
because the unit (dollars) is meaningful and consistent across the data.
Relative Measure of Dispersion Scenario
• A financial analyst compares the risk of two mutual funds: Fund A with an average
return of 8% and a standard deviation of 5%, and Fund B with an average return of
15% and a standard deviation of 8%.
• Using the coefficient of variation (CV), which is the ratio of standard deviation to
mean expressed as a percentage, allows comparing the relative risk.
• Fund A's CV = (5 / 8) × 100 = 62.5%, Fund B's CV = (8 / 15) × 100 = 53.3%.
• Although Fund B has a higher absolute standard deviation, relative to its higher
average return, it has lower relative risk as shown by the lower CV.
This relative measure is crucial for comparing data sets with different scales or units,
allowing normalized comparison of variability.
Summary Table of Examples
Measure Type Scenario Example Why Use It Interpretation
Absolute Salary variability in Units matter for Salaries typically vary by
Dispersion dollars practical $7,000 around the
interpretation mean salary
Relative Risk comparison Compare variability Fund B is relatively less
Dispersion between two relative to mean risky despite larger
investment funds return absolute SD
Chebyshev's Theorem
Chebyshev's Theorem (also known as Chebyshev's Inequality) is a fundamental result in
statistics that applies to any data distribution, regardless of its shape. It provides a
minimum proportion of values that fall within a specified number of standard deviations
(k) from the mean.
Implications
• The theorem places a lower bound on how much data falls near the mean,
making no assumptions about the distribution being normal or otherwise.
• For example:
• At least 75% of values lie within 2 standard deviations.
• At least 88.89% lie within 3 standard deviations.
• It is often used in situations where the distribution is unknown or not normal,
providing a conservative estimate of data spread.
• Example usage of Chebyshev's Theorem:
Empirical Rule:
The Empirical Rule, also known as the 68-95-99.7 Rule or the Three-Sigma Rule, is a
statistical guideline that applies specifically to data following a normal (bell-shaped)
distribution.
Definition
• The Empirical Rule states that for a normal distribution:
• About 68% of the data falls within 1 standard deviation of the mean.
• About 95% of the data falls within 2 standard deviations of the mean.
• About 99.7% of the data falls within 3 standard deviations of the mean.
Use and Importance
• The rule helps quickly estimate the spread and probable range of data when the
distribution is normal.
• It allows statisticians and analysts to understand how much data clusters around
the mean and identify potential outliers beyond 3 standard deviations.
• The empirical rule is valuable in fields like quality control, finance, healthcare, and
any area where data conformity to normality is expected.
Example
• For pulse rates of 100 students with a mean of 75 beats per minute and standard
deviation of 4 beats:
• Approximately 68 students have pulse rates between 71 and 79 (mean ± 1
SD).
• Approximately 95 students have pulse rates between 67 and 83 (mean ± 2
SD).
• Approximately 99 students have pulse rates between 63 and 87 (mean ± 3
SD).
Limitations
• This rule only applies for roughly normal distributions.
• It is not suitable for skewed or irregularly shaped data.
• Unlike Chebyshev's Theorem, which applies to all distributions, the empirical rule
assumes symmetry and a unimodal shape.
The Empirical Rule provides a quick, practical way to interpret standard deviations
and probabilities for normally distributed data, making it a foundational tool in statistics
for probabilistic estimates and outlier detection.
Measures of Position
Measures of position in statistics are tools used to determine where a specific data
value lies in relation to the entire data set or distribution. They help to understand whether
a value is about average, unusually high, or unusually low and allow comparison across
different datasets or scales.
Common Measures of Position
• Percentiles: Divide a ranked data set into 100 equal parts. For example, the
90th percentile is the value below which 90% of the data falls.
• Quartiles: Special percentiles dividing data into four equal parts. The first
quartile (Q1) is the 25th percentile, the median (Q2) is the 50th percentile,
and the third quartile (Q3) is the 75th percentile.
• Deciles: Divide data into ten equal parts, marking 10%, 20%, ..., up to 100%
of the distribution.
Purpose and Use
• These measures show where values fall within a distribution, indicating relative
standing.
• Percentiles and quartiles are especially useful in fields like education (test score
rankings), health (growth charts), and research where position matters more than
absolute value.
• Z-scores help standardize data for comparison across different scales.
Example
• If a student scores in the 85th percentile on a test, it means the student scored
better than 85% of the test takers.
• The median quartile (Q2) represents the middle value splitting the lowest 50% of
data from the highest 50%.
Summary
Measures of position provide insight into the relative location of data points within a set,
facilitating comparison, ranking, and interpretation of values in context of the overall
distribution.
Moments of Statistical Distribution:
Moments of a statistical distribution are quantitative measures that describe various
characteristics of the distribution's shape, including its central tendency, variability,
asymmetry, and peakedness. Moments help summarize important features of the data
or probability distribution.
Significance of Moments
• 1st Moment (Mean): Represents the center or average value of the distribution.
• 2nd Moment (Variance): Measures the dispersion or spread around the mean.
• 3rd Moment (Skewness): Indicates symmetry or whether the distribution tails lean
to one side.
• 4th Moment (Kurtosis): Describes the sharpness of the peak or the heaviness of tails
compared to a normal distribution.
Applications
• Describing distribution shapes in descriptive statistics.
• Used in statistical inference and hypothesis testing.
• Applied in machine learning for feature engineering and transformations.
• Important in finance to analyze asset returns' risk (volatility and skewness).
Significance of Moments:
The significance of the first four statistical moments—mean, variance, skewness, and
kurtosis—lies in their ability to comprehensively describe key aspects of a distribution’s
shape and behavior:
Mean (1st Moment)
• The mean represents the central location or average value of the distribution.
• It provides a measure of central tendency, indicating where data points tend to
cluster.
• The mean is sensitive to all data values and can be influenced by extreme values
or outliers.
• It serves as the reference point for calculating other moments such as variance
and skewness.
Variance (2nd Moment)
• Variance measures the dispersion or spread of data values around the mean.
• It quantifies how much the values deviate on average, considering the square of
deviations to ensure positive values.
• A low variance indicates data points are tightly grouped near the mean, while
high variance indicates greater spread.
• Variance forms the basis for the standard deviation, a widely used measure of
variability in the same units as the data.
Skewness (3rd Moment)
• Skewness assesses the asymmetry of the distribution around the mean.
• A skewness of zero indicates a symmetric distribution.
• Positive skewness means the right tail is longer or heavier, suggesting more extreme
higher values.
• Negative skewness means the left tail is longer, indicating more extreme lower
values.
• Skewness helps in identifying departures from normality and understanding tail
behavior.
Kurtosis (4th Moment)
• Kurtosis measures the tailedness or peakedness of the distribution.
• Higher kurtosis indicates heavier tails and a sharper peak, reflecting more frequent
extreme deviations (outliers).
• Lower kurtosis corresponds to lighter tails and a flatter peak.
• It helps in understanding the risk of extreme outcomes, especially in finance and
quality control.
Summary
Together, these moments provide a detailed statistical summary:
• Mean locates the center.
• Variance describes spread.
• Skewness reveals asymmetry.
• Kurtosis indicates tail risk and peak sharpness.
These interpretations enable analysts to capture not only where data lie but also how
they behave, assisting in model selection, hypothesis testing, and decision making across
fields such as finance, engineering, health sciences, and social sciences