The population is the whole of the research object which is the center of attention and is the source of research data. The object of research can be humans, animals, plants, symptoms, values, events, attitudes to life, and so on. The sample is part of the population selected by using certain rules, which are used to collect information/data that describes the nature or characteristics of the population.
If we look at the definition, the definition of population can be very diverse, so we must define the population clearly and precisely. On the other hand, a sample that is representative of the population must be able to describe the characteristics of the population because the sample is used to generalize a population. Thus, the sample must be truly representative so that it can represent and reflect the characteristics of the population from which the sample was taken.
Population
The following is an example of a population:
- Padjadjaran University (Unpad) Student Population
- Student Population of the Faculty of Agriculture (Faperta)
- Agrotechnology Student Population, Faperta, Unpad
- 2009 Class of Agrotechnology Student Population, Faperta, Unpad
- Class A Agrotechnology Student Population, Class of 2009, Faperta, Unpad
If we look at the population example above, the definition of population there is relative, the definition depends on the researcher, whether he wants to know the Unpad Student Population as a whole or is only interested in the 2009 student population of Agrotechnology.
We must be careful in defining a population. The population must be clearly and precisely defined. For example, we want to know the average GPA of Unpad students. It means that the parameter/nature/character you want to know is the average GPA of students and the object being studied is Unpad students. If we formulate a population like this, the formula is clear but not precise. The meaning is clear: (1) the parameters to be studied are clear, namely the GPA of Unpad students and not other parameters, such as height, IQ scores and so on (2) the population is only Unpad students, not the GPA of students from other universities. The meaning is not quite right, when we talk about Unpad students, the scope is quite broad. Are we going to record the GPA scores of all Unpad students from all generations, whether they are still active, inactive, dead, dropped out, or those who have graduated?
Thus, the limits of the scope of the population to be studied must be clearly and precisely defined, because all conclusions that will later be obtained from the results of sampling (samples) only apply to the population in question, not to populations that are outside the given scope limits. .
Consider the following population definition:
"The population in this study is Agrotechnology Students Class 2009, Faperta Unpad, who are still active"
The definition of a population like this is clear in scope, so any conclusions that are given to a sample taken from that population only apply to the population limited by 2009 Agrotechnology Students, Faperta, Unpad, who are still actively studying and do not apply to other students who are currently studying. is outside that scope. So it only describes the state of the average student GPA in that scope.
Populations can be divided based on circumstances (complexity) and based on size. According to the circumstances the population can be divided into two parts, namely Homogeneous Populations , and Heterogeneous Populations . Based on its size, the population is also divided into two parts, namely the finite population, and the infinite population .
Population by state:
Homogeneous Population : the population is said to be homogeneous if the elements of the population studied have relatively uniform properties with each other. These characteristics are often found in the exact field, such as water, solutions, etc. If we want to know whether a cup of coffee is sweet or not, it is enough to try a drop of the coffee liquid. A drop of coffee liquid can represent the sugar content of a cup of coffee.
Heterogeneous Population: the population is said to be heterogeneous if the elements of the population studied have relatively different properties from each other. These characteristics are often found in social and behavioral research, where human objects or phenomena in human life are unique and complex. For example, if we want to know the average IQ of Unpad students batch 2009 (meaning the average of all faculties). Obviously, the average IQ of students between Faculties is likely to vary, the IQ of Medical Faculty students is relatively higher than the average IQ of other Faculties students, so we can say that the population is heterogeneous. To overcome heterogeneous populations in conducting research, it is necessary to group them based on their characteristics, so that the existing population is grouped into several groups, which later these groups will be homogeneous in their groups, but these groups are very heterogeneous among their groups. In the previous example, the group is identical to the Faculty.
Population by size:
Finite population: A population is said to be finite when the number of members of the population can be estimated or known with certainty, in other words, the limits are clear quantitatively, for example:
- Number of Class A Agrotechnology Students, Class of 2009, Faperta, Unpad
- High population in a certain city
- Length of fish in a lake
Infinite population : the population is said to be infinite if the members of the population cannot be estimated or the number can not be known, in other words, the limits cannot be determined quantitatively, for example:
- Water in the ocean
- The amount of sand on Pangandaran Beach.
- The number of children who suffer from malnutrition
- The depth of a lake measured from various points
However, in the practice of everyday life, we often encounter a finite population which is considered an infinite population , and things like this are statistically justified, for example, the number of Indonesians who smoke, the number of Indonesians now, and so on.
Sample
In inferential statistics, we want to know the description of certain characteristics of a population, but sometimes it is sometimes impossible and impractical to observe all objects/individuals that make up a population. Rice retailers only examine a handful of rice to determine the current quality of the rice. Gold traders only examine the rubble of the jewelry to determine the quality of the gold jewelry. Environmental researchers only examine a few milliliters of water to determine the quality of water in a river or lake. The question is, why not research as a whole, won't the results be better and more precise?
Considering that a researcher in conducting research is full of limitations both in terms of cost, time, and so on, the research conducted to collect the desired information or data according to the problem under study is taken by taking part of the population, taking into account the existing limitations of the researcher. Part of the population as a place to collect information is called a sample.
Thus, the sample is part of the population selected by using certain rules, which are used to collect information/data that describes the nature or characteristics of the population .
From this definition it is clear that the sample we take is used to describe the characteristics of a population, or in other words, the sample is used to generalize a population. Thus, the sample must be truly representative so that it can represent and reflect the characteristics of the population from which the sample was taken.
Take a look at the representative sample picture in the image below. The lighter colored areas represent low concentrations/values and darker colored areas represent high concentrations/values. A representative sample must be able to represent the value of the population so that the probability of taking a light, medium, or dark color must be equal or proportional.
A researcher, rarely observes the entire population for two reasons:
- Cost is too high and
- The population is dynamic, that is, the elements of the population can change from time to time.
There are three main advantages of sampling:
- lower cost,
- Faster data collection, and
- This is possible to ensure uniformity and to improve data accuracy and quality because the data set is smaller.
Sample types
In the sample selection process there are two determining factors that play a role, namely:
- The presence or absence of a randomization factor, and
- The role of the person who chooses (takes) the sample.
In the sampling process using a randomization factor it includes elements of opportunity, while the role of the sample voters can be objective and can also be subjective.
The objective nature of selecting a sample is a way of selecting a sample that uses a certain clear method, so that if the sampling is carried out by someone else, it will obtain results that are not much different from the previous sampling, in estimating the nature or characteristics of the population. So by taking a sample using a certain and clear method, a consistent sample will be obtained, meaning that if the sampling is carried out repeatedly on the same population, the results are still controlled in the sense that it still describes the nature or characteristics of the population, even though the results are not exactly the same between the two groups. one with the other.
The subjective nature of selecting a sample is a sample selection involving personal considerations from the sample taker to take a good sample according to his own version (the researcher's version). Thus, the sample obtained is a biased sample, moreover the people who choose the sample have less background on statistical concepts, especially the concept of sampling theory.
Reference:
- Bungin, Burhan. 2006. Quantitative Research Methods on Communication, Economics, and Public Policy and Other Social Sciences. Prenada Media Group. Jakarta.
- http://en.wikipedia.org/wiki/Sampling_(statistics)
- Walpole, RE 1992. Pengantar Statistik. PT Gramedia Pustaka Utama, Jakarta.
- Other sources
