Sampling is studied in the probability section of mathematics. Likewise in the research method sampling plays an important role. It is clearly evident that not whole population can be involved in any observation. Thus according to the demand and process involved, a certain amount of people are taken and it is called sampling. W.G Cochran defined sampling methods as the most efficient methods for low-cost budget and high observation skills.
TYPES OF SAMPLING
There are two basic categorizations viz. Probability sampling and Non-Probability sampling. In Probability sampling, there is a fair chance for every section of the population. It permits the freedom of likelihood activities with like people. This is a more preferred method but in some cases contrast method viz. The nonprobability method is implied. For example in IIT top scorers are needed, not every student is given priority to study there.
TYPES OF PROBABILITY SAMPLE
- SIMPLE RANDOM METHOD– It is the synonym to hit and trial method. In this method, every person is assigned a particular number and then they are selected by picking the chips placed in a hat or bowl randomly.
- Systematic Random Sampling: In this, absolute randomness is not there as sample is selected on the basis of some pre-defined criterion. For example, every tenth of a population may be picked for a sampling exercise.
- STRATIFIED SAMPLE METHOD– This is a certain modification of Simple Random method by using Monte Carlo method. In this method, there is division according to the needed arrangement. For example, a group of the 10th and 12th standard is required for board details fulfillment, then there will be two subgroups in which the number of students will be decided randomly till the need ends.
- CLUSTER SAMPLING– Clusters are either spatial or temporal. It can be multistage also as in the case of state level sampling followed by district level samlping, followed by block level sampling and so on. This method is specially designed to save cost. It is pretty evident that experimenting nondeserving candidates will be less efficient and expensive. Thus every candidate will be interviewed and based on the results; this sampling process will be done. For example and job comes under this sampling.
DISADVANTAGES OF PROBABILITY SAMPLING:-
- As probability sampling consumes every element of the population, it is more time consuming and expensive than non-probability sampling.
- It is not mandatory that every person will be interested in the research process. Thus the less response will act as a disadvantage.
They are generally not favored as compared to former method but in some of the cases we cannot involve every element of the society, thus it becomes important to use this method too. Some of its types are as follows:-
- Accidental Sample– It is the most convenient form of sampling, also known as Grab Sampling. In this method, a handful amount of people are selected according to the need. The whole population is not needed which makes it less time-consuming.
- Purposive Sample– It is termed as judgmental kind of sampling. For example, if the researcher wants to study how people behave in ganging up with friends vs. alone. Then he will take anybody irrespective of his behavior. The whole analysis is based on the perspective of the researcher, Quota sampling and Snowball Sampling are essentially subtypes of purposive sampling.
- Quota Sample– It is just opposite to random method. The researcher is clear with the number of people and key factors to select. For example, if a football team requires 4 boys and 4 girls, then the subgroups of both gender will be made and the trial session will be done till the required amount of people are selected
- Snowball Sample– This sampling can be associated with the way internet works. In simple language, it deals with referral of the same category of people once the experiment is over. For example, a football competition is organized and 16 members are selected. Once the competitions are over, the coach or say researcher will ask for any sportsperson involved in any kind of sports. Being in the same genre it always helps to decide the sample size accordingly. It could be biased many times because in the sports case, many people who are good in games will lose the chance to perform due to less contact.
FACTORS AFFECTING THE SAMPLE SIZE
We want the process to be more efficient, less expensive and more convenient. The samples are advised to be small because more than required will hamper the efficiency. Now we will see the factors which can affect the size of a sample.
- Different places consist of a different population. Thus every place is not homogeneous. The selection of sample size will be directly affected by the number of elements.
- The subgroup is also an important factor. Different analysis demands different subgroups. For example, a national level selection will demand large sample size, to the contrast for a local research; the sample size will be small.
3. The kind of research equally matters. For example, if the research period is large, simultaneously the sample size will grow in size. Intensive researches usually comprise of large sample size.
4. The amount of time, skilled labors, and available money are the key factors of this method. If the skilled labor number decreases, time will increase and thus sample size will increase. There is the direct proportional relation between time and sample size.
5. If the amount of questions asked in research is large, then the sample size should be kept small to save the time. For example, if an interview comprises of 1-hour question asking session, the number of appointees will be less.
6. The type of sampling involved is important in size determination. For example in the stratified method, the size would be small due to the less random pattern.
PROCEDURE OF SELECTING A SAMPLE
It is always preferred to select a sample which can be used efficiently in generalizing the society. The selection entirely does not depend upon size, but the reliability and accuracy do matter. The chances for everyone to be included in the sample should be fair and not biased. The pattern of selection does matter. For example, if you select a smaller group with less observation will be less productive. On the other hand, if a large amount of sample size is selected with keen observation, it will be more precise. For example like the universal set involves every element, it is important that sample should be a universal concept.
If a sample does not account for a fair chance, then it is called as a biased sample. There are many methods of sample selection viz. simple random sampling, stratified random sampling, regular interval sampling, area sampling etc. This procedure is an integration of processes which are sometimes time-consuming too.
A hypothesis can be seen as a mere guess for further methodology. The previous happenings are taken into account first then a relationship is established. A hypothesis is generally verified to check whether it can be formulated as a theory or not. This system is a most accurate method as compare to other ones.
Data collection is the raw material of investigation.
For example in a science laboratory, students can predict the amount of current passing at zero position of a galvanometer. On that account further readings are taken if the hypothesis is true.
FUNCTIONS OF HYPOTHESIS
1. As the amount of facts is discussed in detail, thus there is hardly any chance for missing any point out there. The case of half information is negligible.
2. The information line is not complicated, thus it does not increase the time in petty things, instead, observations and results are directly connected to each other.
3. As mentioned earlier, this method directly deals with the facts, unlike random sample collection. It decreases the amount of money and time both at the same time.
4. Merely data collection if found true leads to terms and theories which accounts for standardization.
5. Northrop has seen hypothesis as a solution to many problems. The paths are direct and thus information coming out of hypothesis can itself be the solution.
CONDITIONS FOR A VALID HYPOTHESIS
- A hypothesis is either fully mathematically defined or refused. There is no in between because in that case, it will behave as a supposition. For example, if we consider a hypothesis, its information is pure facts based.
- A hypothesis is mainly generated for the solution, thus if it is not yielding any result, then the process should be stopped at any moment.
- Sometimes a hypothesis can come out as a false one, but it helps in finding a new hypothesis with modifications. For example, if we have considered the age of students in a 10th class to be 14 years and the average student’s age is 15 then we will realize that the assumption is wrong and thus it helped in a modification.
- Sometimes we have more than one hypothesis and thus the selection process should be based on the weight of the hypothesis on the ground of powerful facts and results.
- There are also cases when two consecutive hypotheses come into the picture but then the choice is based on the simplicity of the hypothesis. A simple hypothesis contains lesser assumptions, more generalizing terms, and less independent variable.
- Generally, the hypothesis stands tall with the conventional knowledge. But some time conventional initials can be wrong too, thus a valid hypothesis removes the wrong from the beginning.
- Although hypothesis at starting is merely an assumption. But the ideation should not be an image in an air castle. The assumption should be made in order to perceive a definite motive.
FORMULATION OF HYPOTHESIS
There is no set pattern of norms to decide what hypothesis should be taken into account for any problem. It is generally formulated from the previously failed hypothesis. A failed assumption thus plays an important role in the success of any hypothesis. Stebbing’s view on the origin of the hypothesis is the collection of knowledge and wisdom possessed by the group of literate people. There is the scientific approach to the formulation of hypothesis in which facts are studied intensively. The group in which hypothesis is the formulation, each and everybody sense of experiencing thing matters. Thus it is a collective analysis. For example, the election process of any country involves a collection of people thinking about the pros and cons and then getting agreed to a common contender.
VERIFICATION AND PROOF OF HYPOTHESIS
Verification of Hypothesis in simple words means analyzing the truth through facts and theories. The validity of hypothesis depends upon the accuracy between inferences and the facts. Greater the accuracy of these factors, greater the strength of hypothesis. There are two ways of checking the validity of hypothesis viz. direct method and Indirect Method. Indirect method there is a direct transmission of observations with theories and if they are complementary to each other, a hypothesis is verified. In the indirect method, a direct approach does not go to ensure the validity of the hypothesis. In this method, the results are checked to ensure that it does not contradict the hypothesis or say assumptions.
There is also one condition which is very important. If the hypothesis is made out of certain facts then the single hypothesis should stand for every fact and conclusion. No other hypothesis should be taken for the validity.
The involvement of two hypotheses can be seen in many cases and then be deciding one of them is crucial for further investigation. For instance, if in a robbery case two groups are involved then deciding which one was involved is crucial. There can be chances that the other group was not present near the location then it should be the deciding link.
As mentioned earlier a hypothesis should be clear and specifically point driven. The hypothesis undergoes two process which helps in seeking how accurate or precise an assumption is. The two processes are validity and reliability.
Reliability in simple words means yielding constant results no matter how many times the procedure is repeated. For example, if someone is a kind-hearted man, if he is asked 100 times for a charity, he will not refuse it. This counts how reliable a person is. In research method and analysis, reliability accounts for consistency of any test. The major advantage of reliability is that it ensures the quality of the procedure.
It is the most basic way to check the reliability. For example, if someone drives a car four times and it operates smoothly, for that person the vehicle will be reliable because he has checked it to his utter satisfaction. Likewise in interviews, the reliability of a person is checked constantly by asking him the same question in different ways. Accurate the answers, more reliable are the person.
The methods are jumbled up to check how reliable measurements are. Qualitative and Quantitative approaches are integrated. For example, in ancient time coal was used as the only source for generating electricity but the modern world is trying new sources with the addition of coal. This method can be called as triangulation.
The selection of sample size for any method plays the most important role. The ways of choosing sample size are based on many assumptions and raw materials. The raw materials are manpower, sample units expenditure and the amount available. The three factors are aligned with each other. Those factors will change accordingly to the surveys and analysis methods. The main objective of these factors is to generate a pattern to know what kind of result in sample size is expected.
There are other factors too on which sample size is dependent.
1. Population or census acts as a major factor. Population differs from place to place, thus the size also differs for sampling.
2. The amount of assumption which can be estimated. Different surveys permit a different amount of assumption.
3. The intensity at which research is examined.
also read →