Sampling Methods: Types of Sampling, Types of Random Sampling, Types of Non-random Sampling

Sampling Methods

Types of Sampling

The sample methods are divided into two categories:

• Random sampling (Probability sampling)
• Non-random sampling (Non-probability sampling)

Probability Sampling

Every unit of the population has a given chance of being chosen for the sample in a probability sampling. It provides a significant amount of representation.

Condition for probability sampling

• Must specify desired sample size.
• There must be an equal possibility for each element to be chosen.
• Various sampling units have different chances of being chosen.
• The sample size is inversely related to the probability of a unit.

Non-probability Sampling

Non-probability sampling is defined as sampling in which the population's units have no specific probability of being chosen for the sample.

Types of Random Sampling (Probability Sampling)

• Simple random sampling
• Stratified sampling
• Systematic sampling
• Cluster sampling
• Multistage sampling

Simple Random Sampling

Every unit of the population has an equal probability of getting chosen using this methodology. As a result, simple random sampling is a technique for choosing 'n' units from a population of size 'N' units while providing each unit an equal chance of being chosen.

Procedure of Simple Random Sampling:

• Lottery method: The identical lottery tickets are meticulously cleaned before being packed in a bag, each bearing a number of the sold tickets. Then, without opening the bag, one ticket is taken out. The remaining tickets in the bag are properly shaken and mingled after the first ticket is drawn. Then, another ticket is drawn. Until the required number of tickets equal to the number of prizes to be awarded is drawn, the procedure of alternately shuffling and drawing a ticket is resumed.
• Use random number tables: Using random number tables is the most practical way to choose a random sample. These random numbers are derived from random number tables or computers. There are many different varieties of random number tables, some of which include:
  • Tippet's random number tables
  • Fisher and Yates tables
  • Kendall tables
  • A million random digits.

There are two types of simple random sampling.

• Simple random sampling without replacement (SRSWOR): If an unit is selected and is not returned to the population before the next drawing, this procedure is called simple random sampling without replacement.

• Simple random sampling with replacement (SRSWR): If an unit is selected, noted and returned to the population before the next drawing, the procedure is called simple random sampling with replacement.

• Merits of Simple Random Sampling:

  • Compared to judgment or purposive sampling, it is more representative of the population.

  • As the sample size grows, estimations become more accurate.

  • The technique saves time and money from an economic standpoint.

Demerits of Simple Random Sampling:

• It needs the most recent population list from which samples can be taken.

• In comparison to stratified random sampling, it typically calls for a higher sample size.

• The population may not produce accurate results if its makeup is diversified.

Stratified Sampling

When a population's characteristics are varied, stratified random sampling is employed. This sampling technique divides the population into strata, which are various classes or groups, in a way that makes the attributes of the units homogeneous within the strata and heterogeneous between the strata. Then, using a straightforward random sampling procedure, samples are taken from each stratum. As an illustration, suppose that we want to research Nepal's crop production.First, we split the entire country of Nepal into the three zones of mountain, hill, and terai so that the characteristics of crops are homogeneous within the strata but heterogeneous between the strata. Then, using a straightforward random sampling procedure, samples are taken from each stratum, possibly representing the production of crops throughout Nepal.

Merits of Stratified Sampling:

• From an organizational standpoint, it is superior than other sampling strategies.
• Taking small samples from the stratum helps lower the cost of data collection and analysis.

Demerits of Stratified Sampling:

• In the case of a non-homogeneous stratum, the outcome of a stratified sampling approach may not be valid.
• To divide a heterogeneous population into a homogeneous group, specialists in statistics were required.

Systematic Sampling

When there is a finite population, it is used. It entails creating the sample in some organized fashion by collecting materials at predetermined intervals. The units are first listed in serial order and put in either alphabetical or numerical order. Therefore, using the following relation, the sample interval is determined.

K = N/n

Where, K = sampling intervals, N=population size, n=sample size. The first sample is selected from the first interval, and then other samples are automatically selected according to pre-assigned manner.

For example; if there are 100 items and if 10 items are to be selected, therefore population size (N) = 100, sample size (n) =10, the sampling intervals (k)= N/n = 100/10 = 10

Now the available data can be arranged in the following way: 0-10, 10-20,20-30, 3040, 40-50, 50-60, 60-70, 70-80, 80-90 and 90-100.

If a first sample 7 is selected randomly by the help of lottery method, then other samples 17, 27, 37, 47, 57, 67, 77, 87, 97 are selected automatically.

Merits of Systematic Sampling:

• Applying is straightforward, affordable, and practical.
• The process of gathering samples from the population requires relatively little time and labor.
• When compared to random sampling, it is more effective if the population list is provided.

Demerits of Systematic Sampling:

• When the population size is limitless or unknowable, it cannot be employed.
• Units are correlated when they vary at regular intervals and the variation is periodic. As a result, the estimates are biased and the sample becomes lopsided.

Cluster Sampling

In a cluster sample, the population is separated into distinct groups, or clusters, so that the characteristics of the units within the cluster are heterogeneous and the qualities between the dusters are homogeneous. So that there are about equal numbers of sampling units in each cluster, a cluster is then chosen as a sample using simple random sampling. Let's say, for instance, that we want to research the economic situation of residents in the Katmandu metropolitan area. First, the metropolitan area is divided into many wards with varying economic conditions within each ward and uniform economic conditions between wards. Then, a ward is chosen as a sample using a straightforward random sampling technique, allowing us to investigate the economic situation of people in the Katmandu metropolitan city.

Merits of Cluster Sampling:

• When the sampling frame is not accessible, it is a suitable way of gathering samples from the population.
• If it's difficult to locate the basic units, cluster sampling is a suitable strategy for gathering samples.

Demerits of Cluster Sampling:

• In comparison to other probability sampling, it exhibits greater mistakes for comparable sample sizes.
• It requires the capacity to uniquely identify each member of the population before include them in the cluster.

Multistage Sampling

This method involves performing the san..:ng operation in stages. The population is first separated into sizable groupings known as primary stage units. The second stage units are subsequently divided into third stage units, and so on, until we reach the final units of sample size. These primary stage units are then further divided into smaller groups known as second stage units. A sample of primary stage units is initially selected using any practical approach. The process is then repeated from stage to stage till we reach the final units of sample size. Next, a sample of second stage units is chosen from each of the selected primary stage units.

For instance, a VDC can be used as the primary sampling unit (PSU) in agricultural surveys to estimate the yield of a crop in a district, followed by villages as the second stage units, crop fields as the third stage units, and a plot of a specific size as the final sample unit.

Merits of Multistage Sampling:

• It makes it possible to utilize already-existing division and sub-division, saving time, money, and labor.
• When the investigation area is quite large, it is more convenient.

Demerits of Multistage Sampling:

• Samples from the various stages should be properly taken in order to avoid incorrect results.
• When the chosen sampling units are reduced, there is a strong likelihood that sampling error will occur.

Types of Non-random Sampling (Non-probability Sampling)

• Judgment or purposive sampling
• Convenience sampling
• Quota sampling

Judgment Sampling

The selection of sample items in this sampling procedure is left up to the investigator's discretion. To put it another way, the investigator uses his judgment when making the selection and includes those things in the sample that he believes are most significant in relation to the qualities being studied. For instance, if we were to research the heart patients in the Katmandu Valley, we would first choose the most well-known cardiac experts there, from whom we could gather the necessary data.

Merits of Judgment Sampling:

• Simple sampling technique for rapid decision-making.
• When the sample size is small, the results are more favorable.

Demerits of Judgment Sampling:

• If the investigator is personally biased, the results are unreliable.
• Since sampling error is not based on random sampling, it cannot be estimated.

Convenience SaMpling

By choosing "convenient" population units, one can obtain a convenience sample. The chunk is another name for the convenience sampling approach. A chunk is the portion of the population under investigation that is chosen not based on probability or judgment but rather out of convenience. For instance, if someone wishes to get the public's perspective on the new fiscal policy of the Nepalese government, they can conduct interviews with several economists through phone or mobile device, depending on what is most convenient for them.

Merits of Convenience Sampling:

• Simple sampling technique for rapid decision-making.
• When the sample size is small, the results are more favorable.

Demerits of Convenience Sampling:

• If the investigator is personally biased, the results are unreliable.
• Due to the fact that it is not based on random sampling, sampling error cannot be predicted.

Quota Sampling

Judgment sampling includes quota sampling. In a quota sampling, quotas are organized based on certain predetermined features within the quotas, and the selection of sample items is left up to individual discretion. For instance, in a radio listening study, the interviewers may be instructed to speak with 500 residents of a specific area, and that out of every 100 people they speak with, 60 will be housewives, 25 will be farmers, and 15 will be young children. The interviewer is able to choose who will be interviewed within these quotas.

Merits of Quota Sampling:

• Compared to other sampling methods, it is less expensive and time-consuming.
• As a result of using a stratified-cum-purposive sample approach, the investigator gains from both.

Demerits of Quota Sampling:

• Due to the investigators' personal ideologies and preconceptions, it can be prejudiced.