Population and Sample Differences in Research

Population and sample are one important part of the research that must be determined from the beginning. By determining the type of object of this research, we can determine the research method that is more by the conditions and needs.

In administering government statistics, the sample and population have their roles in each study. Various National Statistical Office (NSO) as providers of basic statistics use both in providing data.

However, consideration of the use of populations and individual samples certainly has a strong methodological basis.

What is Population?

A population is a group, collection, or set of all objects that we will examine.

The values ​​calculated and obtained from this population are called parameters.

Population-based data collection has begun a long time ago, even since the days of colonialism. The UK, the Netherlands, and even Greece have started a complete data collection on the population by collecting various data such as name, gender, date of birth, marital status, employment, etc. which will be used in development planning.

What is a Sample?

The sample is part of a population that has characteristics similar to the population itself.

The values ​​calculated and obtained from this sample are called statistics.

In the sample selection process, of course, it must refer to the rules of sampling methods to produce unbiased statistics.

Population and Sample Differences in Estimation

In the use of samples, there are 2 types of errors that we usually know about.

1. Sampling error is a type of error that occurs due to errors in the sampling method itself.

2. Non-sampling error is a type of error that occurs because of errors outside the sampling method. An example of this non-sampling error is a human error (mistakes made intentionally or unintentionally by humans.

By using samples, researchers must be careful in managing sampling errors. In the sample selection process, the characteristics and conditions of the study should be fully understood so that this error can be minimized as little as possible.

With a small sampling error, then, of course, the resulting statistical value will approach the actual parameters or values.

In contrast to the sample, in studies with objects, the population is very vulnerable to non-sampling error, especially human error.

Sampling error in the population should be zero because there can be no error in the selection of research objects. This is because all objects are recorded completely.

However, population data collection is very vulnerable to human errors that may occur in the process of data collection.

An example of this non-sampling error is, there is a data collection officer who is less thorough in collecting data so there are questions that are missed. This could be due to the very heavy burden of researchers so that it drains concentration and causes accidental errors.

population-and-sample-difference

Example of Use of Population and Sample

Suppose we have data of 30 students with height as follows.

Based on the data above, the parameters of student body height are 173.9 cm. Remember, this is a mean value with parameter types.

population-in-research

To get this data, of course, we need quite a lot of time and energy. We have to do data collection with a height measuring devices to these 30 students.

Meanwhile, if we use a sample, of course, we can produce data with a relatively faster time. Suppose, we use a particular sampling method, obtained 10 student samples with the following calculation results.

sample-in-research

See? With just 10 samples, we can already estimate the average height statistic in that class is 172.9 cm.

When you see the difference between parameters and statistics which are only around 1 cm, of course, the results of this study are valid enough to be trusted. The difference is a natural thing in research. This is a risk of using the sample itself.

But, by choosing the right sampling method, these statistics and parameters will have values ​​that are getting closer and even (in some cases) almost the same.

Why Do Many Studies Use Samples?

There are several main reasons why studies are more likely to use samples than populations.

1. The population size is too large

If the population size is too large, it seems very difficult to do research. In this case, the use of samples will be far more effective and efficient to produce the required data.

2. Cost efficiency

In the process, it could be that the use of the population will incur enormous costs. For example, suppose we want to get the latest population indicators for national, provincial and district levels.

When using population, can you imagine how much it would cost to reach the entire population in a country? How many millions of questionnaires must be used to obtain data on the entire population?

3. Faster time

Suppose we have sufficient funds to research with population objects. The next question is, how long does it take in the process of collecting until presenting the data?

Research using samples allows the presentation of data in a relatively faster time than the population.

4. More efficient resources

The resources referred to here are other supporting matters such as data processing information technology equipment. To process population data with millions, of course, requires data storage devices and computers with high specifications.

This will be very different if the data collection is done in the form of samples.

5. Research that is not possible to use population

In some research examples, it may be that the use of populations is not possible and will endanger the object of the population itself.

For example, are we sure that the color of human blood is all red? Are we sure that the blood on the head and the legs is red?

To ensure this, we do not need to suck all the blood and check the results.

Simply take a few drops on each part of the body and of course, we can conclude it.

This is what is meant by research that is not possible to use population because it can bring unwanted things.

Does research still need population objects in this era?

With the development of statistics, the increasingly diverse availability of data sources, and the demands for the provision of data in a faster time, the population-based data is certainly becoming increasingly avoided.

Besides requiring large costs and resources, time is one of the most essential reasons why the population is no longer the first choice in meeting data needs.

However, the United Nations Statistics Divisions still provides recommendations for the implementation of several Censuses around the world. Among them:

1. Population and housing census

Population and housing census is a complete population data collection to meet the availability of basic population data to the level of the smallest regional unit in a country.

As we know, the availability of data up to the smallest area becomes a challenge for data producers. The increasing need and curiosity of the community for data makes data providers must be able to present data to the smallest level of the region, and this is still difficult to obtain with sample population objects.

Population and housing censuses allow a country to obtain basic population data and various demographic questions parameters up to the level of the smallest regional unit with a very small error rate.

Population Censuses are conducted every year ending in 0. In 2020, 54 countries implement Population Censuses worldwide.

2. Economic Census

The Economic Census is a complete data collection on economic activities in all regions without exception.

With the Economic Census, it is expected that data on every economic activity in all regions of Indonesia will be available without exception.

3. Agricultural Census

The Agricultural Census is a complete data collection on agricultural activities throughout the region without exception.

Although the title is the Agricultural Census, the population coverage of this data collection includes animal husbandry and fisheries.

4. Tax Census

The tax census is a complete population data collection that contains detailed information about the condition of the population in terms of taxation. In this census, of course, the type of work, the amount of income, the number of taxpayers, etc. will be asked which will be useful in increasing state revenues.

5. Village Potential Census

In some countries that consist of many villages, complete village data collection is still recommended. This is to support the availability of complete and up-to-date information on the condition of the village.

Summary

A population is a group, collection, or set of all objects that we will examine and the sample is part of a population that has characteristics similar to the population.

The population and sample have their advantages and disadvantages in the study. By understanding the advantages and disadvantages, a researcher will certainly be wiser in determining the types of data analysis.

If you have all the resources, why do not use the population as the research object? But, when everything is limited and restricted, using a sample is the best option we have.

Do not hesitate to leave a comment!