Thursday, August 2, 2012

DATA SOURCES : POPULATION AND SAMPLE

Understanding the source of the data in the study is the subject of where the data is obtained. When the researchers used questionnaires or interviews in data collection, the data source is called the respondent, that is, those who respond to or answer the research questions, either written or oral questions.
When the researchers used a technique of observation, it can be a source of data objects, motion, or process anything. Researchers who observed the growth of rice, then the source data is rice, while the object of research is the growth of corn. When researchers use the documentation, then the document or record that is the source of data, while the contents of a record variable is the object of research or study.
To facilitate identifying data sources, then use the classification of data sources shortened by 3 P, namely:
1. Person:
If the source data is a person. Person is the source of data that are biased to give an oral answer in the form of data through interviews or written responses through questionnaires.
2. Place:
If the source data is a point. Place the source of data that presents the appearance of being still and moving.
Silence, for example the room, the completeness of the tool, object shape, color and others.
Moves, such as: activity, performance, vehicle rate and others. In general, the display still and motion of an object to the use of observational methods.
3. Paper:
If the source data is a symbol. Paper is a source of data that presents signs include letters, numbers, pictures, symbol or other symbol. Definition of paper is not limited to paper, but also can be either stone, wood, bones, palm leaves and so on, which is suitable for the use of methods of documentation.
What sources of data described above are viewed from a subject in which the data stick. In the following section the data sources are also discussed in relation to the coverage (area) source of data taken as a research subject.
With respect to the data sources that serve as research subjects, it is known there are three types of research are:
1. The study population
2. research Samples
3. case study 

POPULATION
Population is a group of people, events or anything that has certain characteristics. Population is the entire subject of study population members is called the elements of the population.
If someone wants to examine all the elements that exist within the study area, the research is the study populations or so-called population studies, or studies are also census.
A population is a set (or collection) of all elements possessing one or more attributes of interest. (In the Encyclopedia of Educational Evaluation)
Population studies conducted if researchers wanted to see all the attributes contained in the population. Since all the subjects covering all contained in the population, it is also called a census.
Objects in the populations studied, then the data obtained and the results analyzed, summarized, and conclusions are valid for the entire population.
Judging from the amount, the population can be divided into two, namely:
1. Finite number,
This means that the number of members (elements) of the population can be calculated or a certain amount. For example, want to know the current UMB student achievement in 2003. In this case the number of students it can be seen from the academic records of the bureau.
2. Infinite number,
This means that the number of members of the population can not be determined amount. For example, research on student achievement UMB. In this case we do not know how many UMB students, as not all active students. Therefore, in the study population should have a first restriction, so that the resulting conclusions can describe the actual condition of the population.

SAMPLE
If you only want to examine a portion of the population, then penelitianya called the study sample.
Sample is representative of the majority or the population studied. Based on the sample and then generalize the results of the study investigators. which
referred to generalize the conclusions raised in the conclusion of the study sample as applicable to the population.
Sampling reason
Some of the factors is the reason why researchers conducted the study sample rather than census (population) is:
1) if the number of elements of the population too much, researchers would not be possible to collect all because it takes effort and costs are relatively expensive.
2) The quality of data generated by the study sample is often better than the census results.
3) The study sample is relatively faster.
4) Alas another's, is when do the research that require destructive testing.
Or when we should conduct research sample?
Research a new sample may be implemented if the state of the subject in a truly heterogeneous population.
If the subject population is not homogeneous, then the conclusion should not be applied to the entire population (the results are not generalizable).

How to take a sample?
To determine the sample there is some way or sampling techniques.
Sampling must be done in such a way that the obtained samples can really describe the actual state of the population or representative.

Sample Selection Criteria
Research using a representative sample will give results that have the ability to generalize. Representative sample depends on two interrelated aspects, namely:
- Accuracy of the sample, ie the extent to which statistical sampling can estimate the population parameters appropriately.
- Accuracy (precision) samples, is how far can the results of research based on a sample population accurately reflect reality.
The sample selection procedure
In order to obtain a representative sample of researchers have used a systematic sampling procedure. Stages are as follows:
1) identify the target population
2) selecting a sample selection framework
3) determine the method of sample selection
4) plan the procedure of determining the sample unit
5) determine the sample size
6) determine the sample unit
Some ways of making the study sample were as follows :
1. Random sample or a random sample
Random sampling technique named because researchers assume that all members of the population are the same so that each member of the population have the same opportunities to choose from.
A simple way to capture samples are as follows:
a. each member of the population (subjects) are listed and numbered consecutively. Suppose there are 100 members of the population, meaning there is a serial number to serial number 100.
b. Determine the amount of sample is desired by the researcher based on certain considerations or according to the formula recommended in the statistics. For example the number of sample 40.
c. Based on the serial number of the population, researchers created a lottery number order. Serial number that came out of the lottery is chosen as a sample. Eg serial number 7, 3, 9,4,23 and so on until all are 40. Number of randomly selected 40 This is a research sample to be analyzed and summarized.

2. Sample stratified or stratified sampling
Stratified sample is used when investigators believe that there are differences in traits, or characteristics of strata exist, and these differences affect the variable strata.
If members of the population is divided into levels or strata, then sampling should not ignore the existence of strata, and each stratum must be represented in the sample.
For example, research that aims to determine the turnover rate of employees in manufacturing industries. In this case we can see that there are strata of the population, namely, that firms in the manufacturing industry has a different company sizes.
Differences in firm size can affect the level of employee turnover. Therefore, the population is divided into three strata, namely, large corporations, small and medium enterprises. Sample must be selected from each stratum or in other words each strata should be represented. Then the selection of samples for each stratum can be done at random.

3. Region or area probability sample sample
Just as stratified sample is performed when there is berbedaan strata, then we can also choose a sample based on region (area) because of differences in the characteristics (traits) between regions with each region other. Thus the sample is the sampling technique is done by taking a representative from each region contained in the population.
For example research on the level of labor skills in Java. Because of factors can affect the level of skill areas, eg in Jakarta workforce more skilled labor than in Magelang then the sample must be selected from each region, in this region can be used by the province or city.

4. Proportional sample proportion or sample
Sampling techniques are used to refine the use of the proportion of sampling or stratified sampling area.
Sometimes the number of subjects included in the strata or the amount is not the same. Therefore, to obtain a representative sample of the research subject of any decision or any territory specified strata balanced or proportional (proportional) to the number of subjects in each strata or regions.
Suppose in the example of research on employee turnover in the manufacturing industry, then every perususahaan small, medium and large respectively taken 50% of the subjects contained in the strata.

5. Samples, or purposive sample aimed
Samples aiming is done by taking the subject of research is not based on random, but the strata or regions based on the presence of a particular purpose. This technique is usually done because of some specific considerations.
Although this is allowed but there are conditions that must be met are:
a. sampling should be based on traits (characteristics) or a particular trait that is characteristic of the subject population.
b. The subject is taken as the sample is really a subject that most contain many features contained in the population or is a key subject.
c. Determination of the characteristics of the population must be done carefully in a preliminary study.
 Thanks

Ditulis Oleh : Unknown // 12:37 AM
Kategori:

0 comments:

Post a Comment

Enter your email address:

Delivered by FeedBurner

 
Powered by Blogger.