I. INTRODUCTION
A research of course has a relation with the conclusion because a research has a purpose to take a conclusion which can give useful information. To get the information about data of this research will be related to take the sample and calculate the statistic value. Whereas the statistic value depend on values of sample which is observed, therefore the statistic value will be various from the sample to the other sample.
In order to the conclusion that we take is useful and appropriate with our desire, so at the sample pulling must précis to get the information that we need. At the next part we will explain about technique in the sample pulling at a research.
II. DISCUSSION
At the statistic we know some method of the sample pulling that may be more efficient, in the meaning that can give information that more great quantities about this populations without extra cost. There are three methods at the sample pulling, that are systematic sample pulling, random stratified pulling, and assemble sample pulling.
1. Systematic Sample Pulling
Systematic sample pulling take every substance to-n in the population, to become sample with the starting point that is fixed random between n substances at the first. For example every thing to-10th is taken as a sample to investigate, or every person that to-25th in the phone book is interviewed, and etc.
Systematic sample very easy to get and it is used with assume like random sample. Systematic sample can produce the conclusion that more appropriate about population parameter, because research values of the sample spread evenly in the whole population. But, the fact dangerous in the systematic sample pulling is if the interval sample pulling that we choose is same with period that hidden in the population.
2. Random Stratified Sample Pulling
This method is taken the simple sample from every stratified population. In this method population is partitioned become some stratified, so relative homogenous in the every stratified. For example, to research teacher’s income at a city may be will be better if this population is partitioned according to profession and next in every profession is taken the simple random sample. Thus, we will get the income variety in every profession more little than the income variety in the whole population. Decrease of the variety in every stratified will produce value that more appropriate to mid value of population, so a stratified random sample that has fixed measure more efficien than simple random sample with the same measure.
3. Assemble Sample Pulling
Assemble sample pulling take some assemble random from population and next take the whole, or choose random a part of substance from every assemble that is choosed to become sample. This method more efficien in cost if the population is spread extensively. For example, at the research about sleep at the afternoon habit of the person at the city, will more cheap if we interview and collect the data from the people that stay closed in assemble or the area that take random, than take simple random sample from whole the people in this city.
III. CONCLUSION
From this explanation we can take conclusion that the three methods are more efficien in cost problem and can give information more great quantities, so in pull the conclusion can more appropriate.
Three methods above, can use together in a research. For example if we want to research the society opinion about international school, we can see the area as assemble, and then take election area with the random stratified. Next, we can use systematic sample pulling to interview the people in the election area that is elected.
REFERENCE
Walpole, R. E. 1995. Pengentar Statistika Edisi ke-3. Jakarta: PT Gramedia Pustaka Utama
Admojo, Suwondo dan Darsono. 2005. Kamus Lengkap. Semarang: CV Widya Karya