This post is about statistical observation as one of the methods of collecting information about socio-economic phenomena.

##### Table of Contents

- What Is Statistical Observation
- Forms of Statistical Observation
- Types of Statistical Observation According to the Completeness
- Types of Statistical Observation According to the Time of Registration of Facts
- Statistical Observation: Ways to Collect Primary Data
- Methods of Survey Data Collection
- Observation Schedule
- Observational Errors

## What Is Statistical Observation

**Statistical observation** is a scientific tool and method of primary data collection.

Also read post “Definition, Role and Brief History of Statistics”.

## Forms of Statistical Observation

From an organization’s point of view, there are two forms of statistical observation:

- reporting,
- specially organized statistical observation.

**Reporting** is a form of statistical observation in which a company provides documents and reports of its activities on time and at certain addresses.

The reporting is subdivided into *standardized* and *specialized*.

On the frequency criteria, reporting is divided into:

*current*– daily, monthly, quarterly.*annual*.

**Specially organized statistical observation **is collecting data from the census and surveys.

Examples of such surveys are sociological surveys, population censuses, censuses of industrial equipment, remains of raw materials, etc.

With the help of specially organized statistical observations, researchers study phenomena and processes that fall outside the scope of statistics reporting. Also, observations are used to check the reporting data.

## Types of Statistical Observation According to the Completeness

According to the completeness of population units, the statistical observation is divided into **complete** and **incomplete**.

### Complete Observation

**Complete observation** involves the survey of all units of the population under study without any exception.

Examples of complete statistical observation:

- in the case of determining industrial production, when the main task is to carry out a full accounting of all industrial production at all enterprises in the region or the country;
- or when determining the number of capital investments, all investors without exception should be covered.

Complete observation is widely used in statistical practice. It is mainly carried out in the form of statistical reporting, which is provided by all enterprises without exception (units of the population under study).

### Incomplete Statistical Observation

**Incomplete observation** is an observation in which not all units of the population, but only a part of them, are subject to survey and registration.

For example:

- a study of cargo handling productivity,
- a study of tariffs and freight for export-import,
- a study of the quality of transport products,
- budget research.

Incomplete observation is very common in statistics. It has some advantages over complete statistical observation:

- Incomplete observation is cheaper and therefore requires less effort and money.
- Makes it possible to use a more detailed program and more quickly obtain research results.

In some cases, incomplete observation is the only one possible for research, for example, when studying the quality of transport products.

#### Types of Incomplete Observation

In turn, depending on the research tasks and the nature of the object of study, **incomplete observation** can be classified into the following types:

1. **Sample observation.** Sampling consists of the fact that only a certain part of the population, selected according to established rules, is examined. Then the obtained data are distributed to the whole population.

2. **Observation of the main array.** Observation of the main array is a kind of incomplete observation in which only that part of the units of the population that has the greatest extent of the studied attribute is examined and considered.

3. **Monographic observation**. It provides for a deep and detailed study and description of individual typical elements of the population (for example, shipping companies, ports, transshipment complexes, regions).

The purpose of such research is to identify existing or just emerging trends in social phenomena and their patterns of development. From this point of view, it is especially important to study the experience of organizing the work of advanced enterprises (shipping companies, ports, etc.). How enterprises have achieved success deserves to be described in detail to make them accessible to a wide range of households.

Also, monographic research can be used to identify shortcomings in the work of individual enterprises.

4. **Questionnaire**. This type of observation provides for obtaining the necessary data by sending out special questionnaires. The addressee fills in and returns them on a completely voluntary basis. For example, the study of public opinion about the work of urban transport, communications, working conditions, etc.

The correctness of filling out the questionnaires depends entirely on the person who fills it out. Such information is difficult to verify.

For example, questionnaires are sent out with a request to characterize the timeliness of mail delivery. This request will be answered mainly by those citizens who have complaints about the work of post service. Therefore, an incomplete observation carried out using questionnaires may produce anomalous results.

This method of observation can be applied in cases where needed not high accuracy of the information, but only approximate data.

## Types of Statistical Observation According to the Time of Registration of Facts

According to the time of observation (time of registration of facts), statistical observation is divided into:

**continuous**(or current) is a type of statistical observation, in which changes in the phenomena under study are recorded as they occur.**discontinuous**, which is divided into periodic (repeated at certain intervals) and one-time (collection of information about the phenomenon under study is carried out directly in the process of research).

## Statistical Observation: Ways to Collect Primary Data

Primary data are collected by researchers directly from main resources.

No matter what form or type of statistical observation is used, primary data can be collected in the following ways:

- direct observation,
- documentation,
- survey.

**Direct observation**: data are obtained by measuring, weighing, and counting. Then the obtained values are entered into the observation forms. This method is used to conduct inventory, register or monitor prices, etc.

**Documentation**: the source of information is legal documents, for example, accounting data, tax accounting data, inventory lists, technical data on fixed assets.

**Survey.** In this case, the source of information is the answers of the interviewees. This method of obtaining data for statistical observation is used in the population census, to form the rating of political leaders, to study public opinion on various issues.

## Methods of Survey Data Collection

There are 5 methods of survey data collection:

- reporting;
- expeditionary;
- self-registration;
- correspondent;
- questionnaire.

**The reporting method **implies the obligatory submission by enterprises, institutions, or organizations of statistical reports about their work.

**Expeditionary method**. Employees of statistical bodies receive the necessary data from each unit of observation directly in the field. In other words, in this case, the designated person interrogates another (which is the unit of observation) and fills out the study form from his/her words.

In the case of **self-registration**, the relevant documents are filled out by the interviewees themselves. After which the documents are checked by the registrars.

With the **correspondent method**, the correspondents send information to the bodies that carry out the observation (survey forms are sent with instructions on how to fill them out to enterprises or individuals with a request to fill out and return to the address of the organization that sent them).

**Questionnaire**. Questionnaires are sent to respondents by mail, e-mail, or delivered personally.

## Observation Schedule

Statistical observation is always carried out according to a predetermined plan, which is an integral part of statistical research.

The schedule of statistical observation contains solutions to program-methodological and organizational issues.

#### Program and Methodological Issues

Program and methodological issues include determination of the purpose, object, unit of observation, observation program, accounting form and instructions, type, and method of observation.

**Any observation begins with a clear definition of its purpose.**

The purpose of statistical observation is the main expected result of statistical research. Its clear formulation makes it possible to avoid the collection of incomplete or redundant data.

2. **By the purpose, determine the object of statistical observation. **

The object of statistical observation is a set of units of the phenomenon under study, about which it is necessary to collect statistical data.

So, in the case of a census of industrial equipment, the *object of observation* is the totality of machines and mechanisms of industrial enterprises.

The success of the entire work largely depends on the accuracy of determining the object of observation, establishing the boundaries of the population under study. If an object is defined incorrectly, then it is likely that a certain part of it will not be examined.

3. **It is necessary to single out the unit of the population, and also to establish the unit of observation.**

**The unit of the statistical population** is the primary element of the object of statistical observation, which is the carrier of the signs that are subject to observation.

**The unit of statistical observation** is the primary unit from which the necessary statistical data should be obtained.

Thus, the unit of the population is what is to be researched, and the unit of observation is the source of information.

For example, during a population census, the *object of observation* is the population of people who live in the country at the time of the census. The *unit of the population* is an individual, so in this case, such a unit coincides with the unit of observation.

However, this does not always happen, and most often the two units do not match. For example, in the case of a census of industrial equipment, the *unit of observation* is the individual industrial plant from which the equipment data will be obtained. *The unit of the population* is a separate piece of equipment since the features that are recorded do not characterize the enterprise, but a separate piece of equipment.

The unit of observation is called an integral part of the object of observation, which forms the basis of accounting and has features that are subject to registration during observation. For example, when studying the composition, structure, and efficiency of the transport fleet of the shipping company, the *object of research* will be the navy, and each ship will be *the unit of observation.*

4. **After the purpose, object, unit of population, and unit of observation have been determined, it is necessary to write an observation program, which is the main part of statistical observation.**

**The observation program** is a list of clearly formulated questions that need to be answered; a list of descriptors and indicators that are subject to registration.

The observation program should be constructed in such a way that it gives an essential characteristic of the practical and theoretical significance of various aspects of the object under study.

Answers to the questions of the observation program are recorded in statistical forms, which are card and list.

The data in the forms can be given in a list or individually for each unit of the population.

In the case of card forms, a lot of time is spent filling out the title and address parts, and paper is used inefficiently. The list form is more convenient for machine processing, in which a whole series of work is performed with less time: encryption, typing.

#### Solving Organizational Issues for Statistical Observation

The statistical observation plan contains a wide range of organizational issues: supervisory authority, place, time and period of observation, sources, and methods of obtaining data, material, and technical base.

The **supervisory authority** is the organizer and executor of statistical observation.

In Ukraine, these functions are mainly performed by the State Statistics Committee with its wide network of departments and inspectorates. Here resolved questions about the time of the observation, including the establishment of the period and the critical moment of observation.

**The period of observation** is the time during which data about the object of observation is collected. The observation time is divided into subjective and objective.

**The critical moment of observation** is the moment at which data about phenomena are recorded.

**The observation period** is the interval in which the start and end of data collection are indicated.

## Observational Errors

For various reasons in the process of statistical observation occur errors – a discrepancy between collection data and the true values of indicators.

Depending on the nature, sources, and causes of occurrence, there are:

- registration errors,
- representativeness (representation) errors.

**Registration errors** occur due to the incorrect establishment of facts during observation or their incorrect recording. In turn, they are *random* (due to inattention, fatigue, for example) and *systematic*. It can arise in both complete and incomplete observation.

**Representativeness errors** are the discrepancy between the value of the indicators obtained from the sample and the value of the same indicators that would have been obtained during a complete observation.

Representativeness errors only occur when there is an incomplete statistical observation.

To eliminate the detected errors, it is necessary to check the collected data.

Image: freepik.com

P. S.: Did you like this article? Share it with your friends, thank you!