CHAPTER IV
ABSOLUTE, RELATIVE, AND AVERAGE VALUES
4.1. The concept of statistical indicators, forms, and types of their expression
Regardless of the purpose and volume, form, and type of expression, any statistical research ends
with calculations and analysis of statistical indicators. The statistical indicator is a quantitative
characteristic of socio-economic phenomena and processes under conditions of qualitative certainty.
Qualitative certainty of indicators means that they are directly related to the content of the studied
phenomenon or process, its essence.
As a rule, the studied statistical processes and phenomena are quite complex and cannot be
directly characterized by one single indicator. Therefore, statistical indicator systems are mainly used
to describe them. The system of statistical indicators is a set of interconnected indicators, a structure
aimed at solving specific single- or multi-level statistical tasks. This may be a simple calculation of units
of the general population, the sum of the values of their attributes (characteristics), a comparison of
one or more values, or a more complex type of calculation.
Depending on the coverage of population units, statistical indicators can be individual and
aggregate, and according to the form of expression - absolute, relative, or average values.
Individual indicators characterize a separate object or a separate unit of the population
(enterprise, bank, household, etc.). Individual absolute indicators are, for example, the income of a
family, number of employees of an enterprise. Based on the ratio of two individual absolute indicators,
we get an individual relative indicator that characterizes the same object or unit. In statistics, average
individual indicators are also calculated for certain periods of time (for example, the average annual
number of employees of an enterprise).
Aggregate indicators, unlike individual indicators, characterize groups of units, describing part of
a statistical population or a statistical population as a whole. They, in turn, can be dimensional and
computational.
Dimensional indicators are calculated by summing the individual values of the population units.
Computational indicators (calculated using various formulas) are used to solve individual tasks of
statistical analysis (for example, measuring variation, characterizing structural shifts, assessing
interdependence, etc.). This group includes indices, correlation coefficients, sampling error, etc.
Computational indicators can also be absolute, relative, or average.
In addition to the form of inclusion and expression of the population unit, a number of other
criteria are used to classify statistical indicators. The time factor is also an important criterion for
classification. Socio-economic phenomena and processes are expressed as statistical indicators of a
certain moment in time or a certain period.
From the point of view of spatial certainty, statistical indicators can be general territorial,
characterizing the studied object (phenomenon) on the scale of the entire country, or regional and
local, which relate respectively to specific regions or individual objects.
27
4.2. Absolute values
Absolute statistical values express the sizes, volumes, and levels of phenomena and processes.
Aggregate statistical values are expressed in absolute quantities characterizing the scale of social
phenomena and processes in certain spatial and temporal conditions. Absolute statistical values can be
classified into individual and sum.
We get the individual absolute indicators directly during statistical observation as absolute
quantities expressing the quantitative characteristics of individual units of the population.
The sum of absolute quantities characterizes the sum of the values of a certain population included in
the statistical observation. We get this either by directly calculating the number of observation units or
by summing the units. Absolute values are nominal numbers that have direct measurement and units
of measurement.
Absolute statistical values are expressed in certain measurement units:
- physical (natural) units,
- conditionally physical (natural) units,
- labor units,
- cost units.
Physical (natural) units are used to characterize sizes and levels (e.g., centimetre (cm), meter
(m), and square meter (m2)); and volume (e.g., litre (l), (dm3), scope (units), kilogram (kg), and ton (t)).
Conditionally physical (natural) units are used if there is a need to combine types of data that
have one common property. Recalculation into nominal physical units of measurement is done with
the help of special coefficients. For example, an author’s sheet, a unit of a literary work volume, is used
to count print production in the publishing business and is equal to 40 thousand printed characters or
3 000 cm2 of printed material.
Labor units of measurement are used to determine time costs for a collection of people. For
example, labor hours are measured in man-hours or man-days.
Cost units of measurement enable the combination of data types that have different properties.
For example, sale volumes in a store, in cost terms, is a product of the unit price and the quantity of
physically counted sold commodities.
28
4.3. Types of relative values and methodology of their calculation
Taken separately, absolute values do not provide the necessary information about the studied
phenomena and processes. On their basis, relative values are calculated, characterized by the
quantitative ratio of the compared absolute values. Unlike absolute statistical values, relative values
are derived values obtained not by simple summation, but by comparing absolute values, and express
a unique quantitative ratio of social phenomena and processes.
The numerator of the relative value indicates the compared quantity, which is called the
reference or current, and the denominator is the comparison base. Thus, the calculated relative value
shows how many times the compared absolute value is greater than the absolute value or what part of
it makes up. If the base is equal to 100, then the relative value is expressed in percentages (%), in the
case of 1.000 - in promils (‰), and in the case of 10.000 - in prodecimiles (o/ooo).
In statistical practice, all the relative values can be divided into the following types:
1. Dynamics;
2. Planned task;
3. Implementation of a plan;
4. Structures;
5. Coordination;
6. Intensities
7. Comparisons
Relative Dynamics Indicator (RDI) expresses the change in the levels of phenomena over time. It
is calculated as the ratio of the levels of different periods or moments of a given phenomenon. Relative
dynamics indicators are expressed in coefficients or percentages.
𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑖𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 𝑌𝑖
RDI= = %
𝐵𝑎𝑠𝑒 𝑖𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 𝑌0
Where: Current indicator (Yi), Baseline indicator (Y0)
The calculated value shows how many times the current indicator exceeds the base indicator, or
what part of it it is. If the indicator is expressed as a ratio, it is called a growth factor, and when it is
multiplied by 100%, it is called the growth rate.
For example, in 2022 the factory produced 300 machine tools, and in 2023 – 330 tools. RDI is 1.1
(330/300), and the growth rate is 110%, that is, the output increased by 10% over the year.
All entities in the financial and economic sphere, from small businesses to large corporations,
compare the results of actual achievements with the expected level when drawing up long-term
activity plans. For this purpose, we can determine the planned task and plan implementation.
29
The planned task indicator is the ratio of the planned level (Ypl) for the upcoming period to the
level of the previous period indicator (Yi).
Ypl
PTI = (%)
𝑌𝑖
The plan implementation indicator is the ratio of the actual level of achievement (Yi+1) of the
current period of time to the level of the planned indicator (Ypl) in the same period.
Yi+1
PII = (%)
Ypl
For example, we have the following data on the company's enterprises (see table 4.1):
Table 4.1 Examples of calculating the planned task and plan implementation indicators
Company Actual volume of sold The planned task of Actual volume of
product product sales growth sold product
in 2022, mln AMD in 2023, % in 2023, mln AMD
A 30 104 32.6
B 48.5 106 52.7
C 60 102.5 63.0
Determine (for the company as a whole):
1. The volume of the planned task according to product sales growth in 2023
2. Percentage of plan implementation according to the volume of product sales in 2023.
3. The indicator of the dynamics of the sold products.
Thus,
Ypl 31.2+51.41+61.5 144.11
1. PTI = (%) = = =1.0405 (104.5%),
Yi 30+48.5+60 138,5
Yi+1 32.6+52.7+63 148.3
2. PII = (%) = = = 1.029 (102.9%),
Ypl 144.11 144.11
148.3
3. Relative Dynamics Indicator (RDI) = = 1.0707 (107.7%),
138.5
There is a mutual relationship between PTI, PII and DRI.
RDI=PTI×PII
30
The relative indicators of a structure (RIS) are the relations of the parts of the structure of the
studied object to the whole.
The relative indicator of a structure characterizes the structure of a population or object. It is
calculated as the relation of the size (scope) of a component to the total, which is a single object or a
whole population:
𝐹𝑖
RIS = x 100%
∑ 𝐹𝑖
Where: Fi is the scope (size) representing a part of the population (object), ∑ 𝐹𝑖 is the whole (the
total) population.
When calculating the relative index of the structure, we take the whole ∑ 𝐹𝑖 as a comparison
base and find the specific weight (RIS) of individual parts in relation to the whole:
For example, there are the following data on the income of employees of the organization by
quarter (see Table 4.2).
Table 4.2 The income of employees of the X organization by quarter (thousands AMD)
By quarter
Year Total
I II III VI
2022 120 150 200 250 720
2023 150 180 230 280 840
Calculate the relative indicators of the income structure of the X organization by quarter.
Solution: Let's calculate the relative indicators of the income structure of the X organization for
2022 and 2023.
In 2022 In 2023
1. RIS1=120/720x100%=16.7% RIS1=150/840x100%=17.9%
2. RIS2=150/720x100%=20.8% RIS2=180/840x100%=21.4%
etc. etc.
The calculated relative indicators of the income structure are presented in Table 1.3.
Table 4.3. The relative indicators of the income structure for 2022-2023 by quarter (%)
Quarter Share of employees' income (%)
2022 2023
I 16.7 17.9
II 20.8 21.4
III 27.8 27.4
IV 34.7 33.3
The data in the table show that the percentage shares of incomes increase consistently from the
first to the fourth quarter.
31
The relative indicators of coordination (RIC) characterize the ratio of different parts of the whole.
One of them is taken as a base for comparison. Relative indicators of coordination are the ratios of
absolute values to the same unit of measurement, expressed as percentages, per mille (ppm)
𝐼𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 (𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟𝑖𝑧𝑒𝑠 𝑡ℎ𝑒 𝑖−𝑡ℎ 𝑝𝑎𝑟𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛)
RIC= =
𝐼𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 (𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟𝑖𝑧𝑒𝑠 𝑡ℎ𝑒 𝑝𝑎𝑟𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑠 𝑎 𝑐𝑜𝑚𝑝𝑎𝑟𝑖𝑠𝑜𝑛 𝑏𝑎𝑠𝑒)
As a result, it is determined how many structural units correspond to one unit of the base
structure.
For example, in Armenia, as of 2023, the average annual number of economically active
population was 1 355.9, including the employed population was 1 184.7 and the unemployed was
171.2.
Calculate how many unemployed people there were for every 1 000 employed in the economy of
the Republic of Armenia.
Solution:
171.2
RIC= x 1 000=144.5
1 184.7
This means that for every 1 000 people employed in the economy, there are 145 unemployed
people.
Relative Intensity Indicators (RII) characterize the prevalence or degree of development of a given
phenomenon in a certain environment.
𝐼𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 (𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟𝑖𝑧𝑒𝑠 𝑝ℎ𝑒𝑛𝑜𝑚𝑒𝑛𝑜𝑛 𝐴)
RII=
𝐼𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟 𝑐ℎ𝑎𝑟𝑎𝑐𝑡𝑒𝑟𝑖𝑧𝑒𝑠 𝑝𝑟𝑒𝑣𝑎𝑙𝑒𝑛𝑐𝑒 𝑜𝑓 𝑝ℎ𝑒𝑛𝑜𝑚𝑒𝑛𝑜𝑛 𝐴
Table 4.4 Natural movement of population of the Republic of Armenia for 2021-2023
year Population Natural movement of population of RA
(thousand people) (thousand people)
Number of Births Number of Deaths Natural growth
2021 2 963.3 36.623 34.388 2.235
2022 2 961.4 36.375 26.692 9.683
2023 2 977.1 36.590 24.313 12.277
On the basis of the above data, it is required to determine the relative indicators of the intensity
of birth, mortality and natural growth.
Thus, in 2021 the relative indicator of birth intensity will be 36.6/2 963.3x1000=12.30 ‰
The relative indicator of mortality intensity: 34.4/2 963.3x1000=11.60‰
The relative indicator of natural growth intensity: 2.2/2 963.3x1000=0.75‰
For other years, relative intensity indicators are calculated according to the same principle (see
Table 4.5).
32
Table 4.5 Relative Intensity Indicators
Natural movement of population of RA
Year (in promils (‰))
Number of Births Number of Deaths Natural growth
2021 12.30 11.60 0.75
2022 12.28 9.01 3.27
2023 12.290 8.17 4.12
The relative indicator of comparison is used in territorial (spatial) comparisons and is defined as a
ratio of similar indicators that characterize different objects (populations) or territories (regions) and
belong to a particular period or moment in time:
𝑨
RIComp= , where A is the reference indicator and A is an indicator that characterizes a
𝑩
comparison base.
The comparison base is chosen randomly. The interpretation of a computed indicator depends on
the choice of a comparison base. For example, in a country the number of visits to theaters per year is
6.2 million, while 17.5 million visits are made to museums. If the number of theater visits is taken as a
comparison base, then we can see that the level of museum visits is 2.8 times higher than the level of
theater visits (17.5/6.2 = 2.8). If the number of museum visits is the comparison base, then the level of
theater visits achieves only a third of museum visits (6.2/17.5 = 0.35). Additionally, a normative or
standard value can be used as a comparison base. For example, the average daily consumption rate for
a man is 3500 kcal. If actual consumption is 3000 Kcal, then consumption is 14.3 % less than the
average (3000/3500 = 0.857).
33
4.4. The concept of the average in statistics and kinds of average values
In statistics, social, including economic, scientific and experimental, production phenomena are
studied using summary indicators, among which there are average values. The most common method
of assessing statistical indicators in socio-economic research is the average value, which is a general
quantitative characteristic of a statistical population under certain conditions of time and place.
The average value reflects the total that is specific to all units of the studied population
In statistical practice, the following types of average values are used:
• arithmetic,
• harmonic,
• geometric,
• square,
• structural median, mode.
Average values (except for the median and mode) are calculated in two ways: simple and
weighted. The most common and applicable form in statistical practice is the arithmetic mean.
Arithmetic mean, its properties
The most common type of average value is the arithmetic mean, which is calculated when the
total volume of the averaged attribute is formed as the sum of the individual values of the studied
statistical population.
The attribute for which the average value is calculated (denoted by the symbol 𝑋̅) is called the
averaged attribute. The value of the averaged attribute for each unit of the population is called the
individual value and is denoted by Xi: X1 X2 ...Xn. The frequency is the number of times an individual
value occurs and is denoted by Fi.
Depending on the nature of the source data, the arithmetic mean can be simple (ungrouped) or
weighted (grouped). The simple arithmetic mean is used when each attribute value (Xi) has the same
frequency as the ungrouped data.
If the attribute values of a given population Xi are equal to X1 X2...Xn, then the arithmetic mean is
determined by the ratio of the sum of the attribute values and the number of units:
𝑋1 +𝑋2 +….+𝑋𝑛 ∑𝑛
𝑖=1 𝑋𝑖
𝑋̅= =
𝑛 𝑛
where 𝑋̅ is the arithmetic mean, Xi is the value of the i-th attribute, and n is the number of units.
For example, Table 4.6 presents data on emissions of harmful substances into the atmosphere
from stationary emission sources in the Republic of Armenia (2018-2023, thousand tons).
34
Table 4.6 Atmospheric emissions from stationary sources in 2018-2023. (thousand tons)
Year 2018 2019 2020 2021 2022 2023
Atmospheric
emissions 25 27 28 24 29 23
Based on the data in the table, let's calculate the value of the arithmetic mean.
𝑋1 +𝑋2 +….+𝑋𝑛 ∑𝑛
𝑖=1 𝑋𝑖 25+27+28+24+29+23
̅=
Solution: 𝑋 = = =26
𝑛 𝑛 6
Thus, the average annual emissions into the atmosphere from stationary sources of harmful
substances amount to 26 thousand tons.
Weighted arithmetic mean
The weighted arithmetic mean is calculated in cases where one or more individual values of an
attribute are repeated. For calculation purposes, the data are grouped into discrete or interval
distribution series.
If we have a discrete distribution series:
xi x1, x2, …, xn
fi f1, f2, …, fn
Then the weighted arithmetic mean is equal to the ratio of the sum of the products of the atribue
values and their corresponding frequencies to the sum of the frequencies:
𝑋 ×𝐹 +𝑋 ×𝐹 +….+𝑋𝑛 ×𝐹𝑛 ∑𝑛 𝑋𝑖 ×𝐹𝑖
𝑋̅= 1 1 2 2 = 𝑖=1∑𝑛
𝐹1 +𝐹2 +⋯+𝐹𝑛 𝑖=1 𝐹𝑖
Xi is the value of the i-th attribute, 𝐹𝑖 is the number of frequencies (weight, frequency).
For example, the distribution of deposits placed in the bank during the month by size is as follows
(thousand AMD) (see Table 4.7).
Table 4.7 Amount of bank deposits
Deposits (thousand AMD) 10 15 20 25 30 35 40 45 50
Depositors (number) 4 10 15 19 22 14 7 6 3
Based on the data in the table, calculate the average amount of deposits placed in the bank
during a month.
Solution: To determine the average value, it is advisable to use the weighted arithmetic mean
formula.
35
To simplify the calculations, we will draw up a table (see table 4.8), in the first column of which
the amounts of deposits (Xi), in the second column - the number of depositors (Fi) and in the third -
their products: Xi × Fi.
Table 4.8 Calculation table
Xi Fi Xi × Fi Fi/Σ Fi Xi × (Fi /Σ Fi)
10 4 40 0.04 0.4
15 10 150 0.1 1.5
20 15 300 0.15 3
25 19 475 0.19 4.75
30 22 660 0.22 6.6
35 14 490 0.14 4.9
40 7 280 0.07 2.8
45 6 270 0.06 2.7
50 3 150 0.03 1.7
Total 100 2815 - 28.15
The weighted arithmetic mean:
𝑋 ×𝐹 +𝑋 ×𝐹 +….+𝑋𝑛 ×𝐹𝑛 2815
𝑋̅= 1 1 2 2 = =28.15
𝐹1 +𝐹2 +⋯+𝐹𝑛 100
As a result, the average amount of deposits placed in the bank during one month is equal to
28.15 thousand AMD.
The properties of the arithmetic mean
The properties of the arithmetic mean are:
[Link] algebraic sum of deviations in all individual values from the average is equal to null:
∑(𝑋𝑖 − 𝑋̅)=0
2. With the increase or decrease of each individual value by any constant value, the average
∑(𝑋𝑗± 𝑎) ×𝐹𝑗
changes by the same value: = 𝑋̅ ± 𝑎;
∑ 𝐹𝑗
[Link] the division or multiplication of each individual value by any number, the average
∑ 𝑋𝑗 × [÷]𝑎) ×𝐹𝑗
decreases or increases by the factor of that number: ∑ 𝐹𝑗
= 𝑋̅ × [÷]𝑎;
4. The average does not change by increasing or decreasing the frequencies of all individual
∑ 𝑋𝑗 ×(𝐹𝑗 × [÷]𝑎)
values by the same factor: = ∑(𝐹𝑗 × [÷]𝑎)
= 𝑋̅;
5. The sum of the square deviation of the individual values from the average is smaller than for
any other value: ∑(𝑋𝑖 − 𝑋̅)2 min
36
Methods for calculating average values
In economic research, when processing statistical materials, problems arise that require different
average values for their solution. Mathematical statistics derives different average values from the
exponential average. The exponential average has the following formula:
𝑛 𝑘 𝑘
∑ 𝑋 ×𝐹𝑖
𝑋̅=( 𝑖=1
∑𝑛
𝑖
)
𝑖=1 𝐹𝑖
Given k different values, we get different average values. Consider the following cases:
1. In case k = -1, we get the weighted harmonic mean:
𝑛 −1
∑ 𝑋 −1 ×𝐹𝑖 ∑𝑛
𝑖=1 𝐹𝑖
𝑋̅=( 𝑖=1∑𝑛 𝑖
) = 𝐹𝑖
𝑖=1 𝐹𝑖 ∑𝑛
𝑖=1 𝑋𝑖
2. In case k = 0, we get the weighted geometric mean:
∑𝑓 ∑𝑓
𝑋̅= √𝑥1 𝑓1 × 𝑥2 𝑓2 × … 𝑥𝑖 𝑓𝑖 = √∏ 𝑥𝑖 𝑓𝑖
𝑖 𝑖
3. In case k = 1, we get the weighted arithmetic mean:
∑ 𝑛
𝑋𝑖 ×𝐹𝑖
𝑋̅= 𝑖=1
∑𝑛
𝐹 𝑖=1 𝑖
4. In case k = 2, we get the weighted square mean.
∑ 𝑛 2
𝑋𝑖 𝐹𝑖
𝑋̅=√ 𝑖=1
∑𝑛 𝐹 𝑖=1 𝑖
The arithmetic mean is used when distribution regularities are studied; the mean square is used
when variations are studied; and the geometric mean is used when dynamics is studied. In
mathematical statistics, different kinds of average are calculated from the same data to give different
values. The correlation between them is called a rule of majority and is expressed as
̅̅̅̅̅̅̅̅̅
𝑋 𝑠𝑞𝑢𝑎𝑟𝑒 > ̅̅̅̅̅̅̅̅
𝑋𝑎𝑟𝑖𝑡ℎ > ̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅
𝑋𝑔𝑒𝑜𝑚 > 𝑋 ℎ𝑎𝑟𝑚
As such, the harmonic mean is smaller than the geometric mean, which, in turn, is smaller than
the arithmetic mean, and the latter is smaller than the square of the mean. Other forms of average
(third, fourth degree, etc.) also exist in the theory of statistics.
The higher the degree of an indicator, the larger the average value. In socioeconomic statistics,
the calculation of various averages for the same population is not advisable, so the choice of the type
of average in every specific case of research is of importance.
37
Table 4.9. Formula for degrees of averages
Calculation formula
k Name of average
Simple Weighted
∑𝑛𝑖=1 𝑋𝑖 ∑
𝑋̅= 𝑖=1
𝑛
𝑋𝑖 ×𝐹𝑖
∑𝑛 𝐹
1 Arithmetic mean
𝑛 𝑖=1 𝑖
𝑛 ∑𝑛𝑖=1 𝐹𝑖
1
-1 Harmonic mean
∑𝑛𝑖=1 ∑𝑛𝑖=1
𝐹𝑖
𝑋𝑖 𝑋𝑖
𝑛 ∑ 𝑓𝑖
0 Geometric mean √∏ 𝑋𝑖 √∏ 𝑥𝑖 𝑓𝑖
∑𝑛𝑖=1 𝑋𝑖2 ∑𝑛𝑖=1 𝑋𝑖2 𝐹𝑖
√ √ 𝑛
2 Mean square
𝑛 ∑𝑖=1 𝐹𝑖
38