Principles of Economic Statistics

phenomena or socio-economic processes. But because we do not stop at numbers but must read their meaning and draw conclusions about phenomena, people classify statistics as social sciences. Statistics is a social science that studies not just one method but a system of methods: collection - processing - analysis, in the analysis there is analysis and prediction. On the basis of analyzing statistical numbers, people can draw out the nature and regularity of the phenomenon. Therefore, statistics is also a quantitative science.

1.1.3. Research object of statistics

Statistics is a social science. However, unlike other social sciences, statistics does not directly study the qualitative aspect of a phenomenon, but only reflects the nature and regularity of the phenomenon through numbers and quantitative expressions of the phenomenon. That means that statistics uses numbers about the scale, structure, proportional relationship, comparative relationship, level of development, level of popularity, etc. of the phenomenon to reflect and express the nature and regularity of the research phenomenon in specific conditions and circumstances. Thus, statistical numbers are not general or abstract, but always contain certain economic, political, and social content, helping us to perceive the nature and regularity of the research phenomenon.

According to the philosophical point of view, quality and quantity are two inseparable aspects of all things and phenomena, and there is always a dialectical relationship between them. In that relationship, the change in quantity determines the change in quality. The law of quantity - quality of philosophy has shown that each specific quantity is associated with a certain quality, when the quantity changes and accumulates to a certain extent, the quality changes accordingly. Therefore, studying the quantitative aspect of a phenomenon will help to perceive the nature of the phenomenon. It is possible to evaluate the production performance of an enterprise through statistics on the total number of products produced, the achieved production value, the rate of completion of the production plan, the unit cost of products, labor productivity and income of workers, etc.

However, in order to reflect the nature and development law of the phenomenon, statistical numbers must be gathered and collected on a large number of individual phenomena. Statistics considers the totality of individual phenomena as a complete entity and takes it as the object of research. The quantitative aspect of individual phenomena is often affected by many factors, including both inevitable and random factors. The level and direction of impact of each of these factors on each individual phenomenon are very different. If only collected on a small number of phenomena, it is difficult to draw out the general nature of the phenomenon, and many times people only find random factors, not

Maybe you are interested!

nature. On the contrary, when studying a large number of individual phenomena, random factors will compensate and cancel each other out, and then the nature and development laws of the phenomenon will be revealed.

The law of large numbers is a law of probability theory, the meaning of this law is: synthesizing observations of large numbers to a sufficient extent of individual random events, the inevitability of the phenomenon will be clearly revealed, thereby revealing the nature of the phenomenon.

The main research object of statistics is large-scale socio-economic phenomena, which include many individual units or phenomena. Through studying a sufficiently large number of these individual units, we will draw conclusions about the nature and regularity of things and phenomena. This conclusion may not be true for each individual phenomenon, but it reflects the large-scale phenomenon. For example: According to the results of the Population and Housing Census at 0:00 on April 1, 2009, in the total population of our country today, the male/female ratio is 98.1/100. This ratio may not be true for each family but is true for the majority of families in Vietnam today. But which number is large enough depends on the characteristics of the phenomenon.

Statistics uses the law of large numbers to quantify the nature and laws of socio-economic phenomena through statistical regularity. Statistical regularity is one of the forms of expressing the general relationship of phenomena in nature and in society. When studying statistical documents on a fairly large number of individual units, statistical regularity is clearly expressed. As in population statistics, through studying a fairly large number of families in many different localities and countries, people found that the rate of female births did not exceed 49%.

In nature, statistical regularities, like laws in general, reflect inevitable causal relationships. However, these relationships are often not of a general nature but depend on the specific development conditions of the phenomenon.

Statistical regularity is not the result of the impact of one cause, but of all causes combined together. It is the synthesis of many causal relationships, a characteristic of large number phenomena synthesized through statistical populations. In general, the wider the time range and the increase in the number of units of the statistical population, the more statistical regularity is manifested.

Statistics only study large number phenomena? The answer is no. Statistics mainly study large number phenomena and combine research with research on individual units and phenomena, usually those that are typically advanced or typically backward. For example: In a factory A, production team B has continuously had the highest labor productivity in the factory for many consecutive years; then, study production team B alone to draw conclusions, why this team has high labor productivity, due to age, level of workers, skill level, overtime... from there draw lessons in management experience to improve productivity.

total factory labor productivity

The research object of statistics always exists in specific time and place conditions. We all know that the quantity of economic and social phenomena often changes over time and space. When time and space conditions change, the nature of things and phenomena may also change. Therefore, when researching, it is necessary to determine when and where the phenomenon occurs. For example: Gold prices at different times and places are different. Even at the same time but in different locations, different stores, gold prices are also different.

From the above analysis, we can draw the following conclusion: The research object of statistics is the quantitative aspect in close relationship with the qualitative aspect of large number phenomena, under specific conditions of time and place.

1.1.4. The role of statistics

Since its inception, statistics have played an increasingly important role in social life. Through the discovery and reflection of quantitative laws of phenomena, statistical numbers help to check, monitor, evaluate programs, plans and orient future socio-economic development.

Nowadays, statistics permeate every activity and every area of life and statistical information becomes one of the invaluable resources to assess the nature and development trends of phenomena.

Statistical information also suggests to users measures to promote better production processes or predict the possibility of achieving them in the future. Therefore, Lenin said: Statistics is the most powerful tool for social cognition.

Depending on different purposes, statistics serve different aspects. Statistical numbers can be used many times with many different goals. Because of its objective nature, easy influence and spread, statistics is one of the important tools, playing a role in providing information for management at both micro and macro levels.

Nowadays, along with the development of human society, the progress of science and technology, statistical science is more and more perfect in theory, methods and information. Statistics are diverse, rich, widely used and increasingly meet the requirements of users.

1.2. Some commonly used concepts in statistics

1.2.1. Statistical population and population units

1.2.1.1. Concept of statistical population and units of population

A statistical population is a set of units (or elements) belonging to the research phenomenon, which need to be observed, collected and analyzed in terms of their quantity according to one or several

certain criteria. The units (or elements) that make up a statistical population are called population units.

The population unit is the starting point of research, because the quantitative aspect of the population unit is the data that the researcher needs to collect. Thus, to determine a statistical population, we need to determine all the units that make it up. Or the essence of determining a statistical population is to determine the population units.

Statistical population is an important concept of statistics. Determining the population is to limit the scope of research for researchers. Thereby, it shows us which units we have to collect documents from and where. For example, when we want to study the industrial production situation in Hanoi, the statistical population will be the population of industrial production enterprises in Hanoi, each enterprise is a population unit. Correctly determining the statistical population is important in statistical research. If the statistical population is not correctly determined (that is, including units that are not actually in that population), the conclusions drawn will be wrong, and the research purpose will not be achieved.

1.2.1.2. General statistical classification

- Based on the recognition of units in the whole, it can be divided into two types: revealed whole and latent whole.

The revealed population is the population that we can directly observe or recognize (For example: the population population, the population of Vietnamese universities...). The latent population is the population that we cannot directly observe or recognize. To determine it, we must use one or more intermediate methods (For example: the population of people who like reformed art, the population of superstitious people...).

This division is directly related to the determination of the whole. Normally, determining the units of an exposed whole is not difficult because they are clearly defined, have definite boundaries with other units, ... Meanwhile, finding the complete and accurate units of a latent whole is more difficult because there is no clear and precise distinction between them and units not belonging to the whole. Therefore, confusion and omission of units in the whole easily occur.

- Based on the research purpose, it can be divided into two types of aggregates: homogeneous aggregates and heterogeneous aggregates.

A homogeneous population consists of units that share the same main characteristics relevant to the research purpose. A heterogeneous population consists of units that differ in type and main characteristics relevant to the research purpose.

This division is very important in determining the representativeness of the calculated statistics. These numbers are only meaningful and representative when calculated from a homogeneous population. If they are calculated from a heterogeneous population, their meaning and representativeness for the population are greatly reduced, and they cannot even be used. For example, when studying income, we often use the statistical indicator of average income. However, average income is only meaningful and only ensures high representativeness when calculated from a population that includes only people with the same working conditions, nature of work, etc.

- Based on the scope of research, people also distinguish between the general whole and the partial whole. The general whole includes all units of the research phenomenon; the partial whole contains only a part of the general whole.

Statistical populations can be finite or infinite (it is impossible or difficult to determine the number of units such as the population of newborns, the population of products produced by a type of machine...). Therefore, when determining statistical populations, it is necessary to limit not only the entity (what kind of population is the population), but also the time and space (in what time and space does the population exist).

In statistical research, the boundaries of the population are often ambiguous and difficult to determine precisely. People have to agree that some types of units are included in the population, and others are not considered as population units.

1.2.2. Statistical criteria

1.2.2.1. Concept of statistical criteria

Statistical criteria are characteristics of the overall unit selected for research depending on different research purposes.

Thus, statistical criteria are not all the characteristics of the overall unit but only the characteristics selected for study.

For example: In the total number of shareholders of company A, each shareholder is a total unit. These shareholders are identified according to different characteristics such as: full name, age, gender, occupation, number of shares held, holding ratio... Each of these characteristics, when selected for research, is a statistical criterion.

In statistical research, statistical criteria are also called variables. For example, when studying the characteristics of students in a school, it is necessary to collect information about the variables: year of study, gender, major, etc. The manifestation of these variables for each student is different. One student may be a second-year student, male, studying Business Administration, while another student may be a first-year student, female, studying Accounting. Criteria help to clearly identify each total unit, thanks to which we can distinguish one unit from another.

1.2.2.2. Classification of statistical criteria

Statistical criteria include the following types:

- Entity criteria: This type of criteria reflects the content characteristics of the overall unit. Depending on the way of expression, there are two types:

+ Attribute criteria: are criteria whose expressions are used to reflect the attributes of the overall unit and do not have direct numerical expressions. For example: criteria of gender, occupation, ethnicity, economic sector,...

There are two types of attribute criteria: First, attribute criteria have direct manifestations, for example, gender has two manifestations: male and female, these manifestations are used to indicate that one person is male and the other is female. Second, attribute criteria have indirect manifestations, for example, the criterion of living standards has indirect manifestations through income, housing area, etc., the indirect manifestations of attribute criteria are called statistical criteria.

+ Quantitative criteria are criteria that reflect the quantitative characteristics of the overall unit and are directly expressed in numbers, each number is called a variable. For example: number of people in a family, monthly salary of workers, labor productivity, etc. Variables are the basis for performing statistical calculations. There are two types of variables: discrete variables (expressed in integers) and continuous variables (expressed in both integers and decimals).

- Time criterion: is a type of criterion that reflects the research phenomenon according to its appearance at what time. For example, if there is data on the number of international tourists to Vietnam by quarter in the past five years, then "quarter" is the time criterion.

- Spatial criteria: is a type of criterion that reflects the territorial scope and location of the research phenomenon. For example, the criterion "province/city" in the data reflects the industrial production value of Vietnam by province/city...

1.2.3. Statistical indicators

1.2.3.1. Concept of statistical indicators

If statistical criteria reflect the characteristics of the overall unit, statistical indicators reflect the characteristics of a large number of overall units or the entire population. Statistical indicators are obtained by synthesizing the quantitative characteristics of many units and individual phenomena into numbers of a large number of phenomena in specific time and space conditions to clearly express the nature and laws of the research phenomenon.

Thus, statistical indicators reflect the quantitative aspect in close relationship with the qualitative aspect of large number phenomena in specific time and space conditions.

For example: According to the 2010 statistical yearbook, Vietnam's gross domestic product (GDP) in 2010 (preliminary calculation) at current prices was 1,980,914 billion VND; the country's population density in 2010 was 263 people/ km2 .

Statistical indicators include two aspects: the concept and the level of the indicator. The conceptual aspect of the indicator includes the definitions and limits of entities, time and space. The level of the indicator is the values reflecting the scale, comparative relationship or intensity of the phenomenon with appropriate units of measurement. For example, with the indicator of the national population density in 2010 being 263 people/km 2 , the conceptual aspect of the indicator is "national population density in 2010" and the level of the indicator is 263 with the unit of measurement being people/km 2 .

1.2.3.2. Classification of statistical indicators

Based on different criteria for classification, statistical indicators can be divided into the following types:

- According to the form of expression, divided into two types:

+ Physical indicators: are indicators expressed in natural units or conventional measurement units. For example: population of a locality (unit of people), production output (unit of meters, tons), ...

+ Value indicator: is an indicator expressed in currency units such as Vietnamese Dong, US Dollar,... For example: GDP, industrial production value (unit Vietnamese Dong), FDI (unit US Dollar),...

- According to the nature of expression, divided into two types:

+ Absolute indicator: is an indicator reflecting the scale and volume of the phenomenon. For example: The number of 12th grade students in high school A is 50, which is an absolute indicator reflecting the scale of the phenomenon. The fabric output in June 2011 of textile factory X is 1 million meters, which is an absolute indicator reflecting the volume of the phenomenon.

+ Relative indicators reflect the comparative relationship between levels of phenomena. For example: the revenue growth rate of enterprise A in 2016 compared to 2015 is 110%.

- According to time characteristics, divided into two types:

+ Period indicator: reflects the quantitative aspect of the research phenomenon in a certain period, depending on the length of the research period. When it is an absolute indicator, it can be added together to calculate the indicator in a longer period.

For example: The revenue of enterprise A in the first quarter of 2015 was 1.2 billion VND, the second quarter of 2015 was 1 billion VND, the third quarter of 2015 was 0.7 billion VND and the fourth quarter of 2015 was 1.1 billion VND. So the total revenue target of enterprise A in 2015 can be calculated as 4 billion VND.

+ Time point indicator: Reflects the quantitative aspect of the research phenomenon at a certain point in time, regardless of the length of the research period. Usually the indicator

This reflects resources such as labor, capital, etc. They cannot be added together to calculate indicators over longer periods.

For example: Vietnam's population on April 1, 2013 was 89,759.5 thousand people. Vietnam's population on April 1, 2014 was 90,728.9 thousand people. Vietnam's population on April 1, 2015 was 91,713.3 thousand people. Then, it is impossible to add the above three indicators to get the Vietnam population indicator for the period 2013-2015.

- According to the content of the feedback, it is divided into two types:

+ Volume index: reflects the scale and volume of the research phenomenon according to specific time and location.

For example: The population of Vietnam at 0:00 on April 1, 2009 was 85,846,997 people.

+ Quality indicators: express the level of popularity and comparative relationship in the whole. Quality indicators can be relative numbers or average numbers, not expressed in absolute numbers.

For example: Vietnam's GDP per capita in 2009 was 1,064 USD/person.

1.2.3.3. Statistical indicators system

The statistical indicator system is a set of many indicators that reflect the most important characteristics, properties, and main relationships of the phenomenon being studied.

The main relationships here include the relationships between aspects of the phenomenon and the relationships between the research phenomenon and related phenomena. The system of statistical indicators plays a very important role: it is the basis and foundation for conducting statistical research.

Meaning: The system of statistical indicators allows quantifying the most important aspects, quantifying the structure and basic relationships of the research phenomenon so that the specific nature and regularity of the phenomenon can be perceived.

But for the indicator system to have quality, effectively reflect the nature of the phenomenon and be feasible, it must comply with certain requirements.

Some basic requirements for building a statistical indicator system:

- Must originate from specific research purposes and characteristics of the research phenomenon.

- Must reflect the most important characteristics, properties, and basic relationships of the research phenomenon.

- Must be feasible, meaning that data can be collected for calculation.

In statistics, the quantitative aspect always goes hand in hand with the qualitative aspect of the phenomenon being studied. However, not all things and phenomena can be quantified. Therefore, to quantify, especially when dealing with attribute criteria, one must use scales. But what scales are there and how are they used for quantification?

Principles of Economic Statistics - 2

Comment