Sampling Methods : Nonprobability Sampling and Online SamplingTechniques

Nonprobability Sampling

Reasearch Design Methods Unlike the probability sampling, in which researchers are already familiar with all population elements, in non-probability sampling researchers do not yet have the list of population elements. Non-probability sampling means every member in a population being measured does not have an equal chance of being selected into the sample because the sample selected based on researchers' subjective assessments (personal judgment). Non-probability sampling is appropriate for exploratory research (although it is also widely used for conclusive research) and when members of the population are homogeneous. Non-probability sampling is an easy and inexpensive sampling technique as we are not required to choose sample according to some statistical calculations, in spite of this technique has a greater possibility of bias.

In general, non-probability sampling consists of convenience sampling, judgment sampling, quota sampling, and snowball sampling.

1. Convenience Sampling
   Convenience sampling technique is a sampling technique in which samples were selected because it was at the right time and place when sampling was conducted. This technique is the most widely used in research because it is the easiest and cheapest method. Sample selection criteria are based on the researcher's subjective opinion. For example, a researcher wants to measure consumer satisfaction on the services provided by particular mall, the researcher then takes a random visitor whom he coould meet without having to know the visitor first. Although this method is the most convenient and inexpensive, but the method has a lot of potential for bias. Convenience sampling method was not suitable for the research which has a population inference.


2. Judgemental Sampling
   Judgmental sampling is a part of convenience sampling in which the population elements were purposely selected based on an assessment researcher using certain criteria (usually qualitative criteria). Several studies using this method include: any tests to determine the market potential of a new product, the Consumer Price Index, and so on.


3. Quota Sampling
   Quota sampling dikenal juga sebagai two-stage restricted judgmental sampling. Langkah Quota sampling is also known as two-stage restricted judgmental sampling. The first step is to make a category of control, or quota, for each element of the population. To make the quota, the researchers first create a list of characteristics for a sample (such as age, sex, race, etc.) and determine the distribution of these characteristics. There are two techniques in determining sample proportion: proportional and non-proportional. In the proportional quota sampling, the quota of sample should be the same as the proportion of the population. For example, in a population of 1000 people where 48% was male and 52% female, were taken to sample as many as 100 people. Thus, men are chosen to sample as many as 48 people and for women by 52. However, in non-proportional the number of quota member in each category is based on researcher interest. The next step, after the quota sample has been determined, we then choose respondents or samples by using convenience sampling or judgmental sampling.


4. Snowball Sampling
   In snowball sampling, the first sample is selected randomly. Then, the second sample will be selected based on recommendations from the first sample. The third sample will also be selected based on recommendations from the second sample, and so on. Snowball sampling is commonly used as the member of population is hard to find, such as: patients of a rare disease or gay groups.

Online Sampling

Today, improvement in research study has developing so fast. The presence of internet enables us to take sample without interact directly with the respondent. When we use of internet to choose sample - called online sampling, the selection of respondents is automatically done by the computer. Usually this method is widely used to research on the population of internet users. Generally, there are two internet based sampling techniques:

1. Online Intercept Sampling
   Pada metode ini pengunjung suatu situs/website “ditangkap” dan diberi kesempatan untuk berpartisipasi dalam survei. Pemilihan sample bisa dilakukan secara random maupun non –random. Jika menggunakan non-random sampling maka semua pengunjung situs tersebut “ditangkap” (biasanya akan muncul Pop-up yang menawarkan untuk berpartisipasi untuk survei). Cara ini cocok untuk situs yang memiliki traffic rendah. Sebaliknya, jika traffic suatu website tinggi sebaiknya menggunakan random sampling dimana software akan memilih secara acak pengunjung tertentu untuk dilibatkan dalam survei.


2. Recruited Online Sampling
   Recruited online sampling method consists of panels and non-panel. If using panel, the survey will be conducted through email or telephone manually. Meanwhile, if researcher using non-panel, prospective sample will be sent an email or a brochure that contains the password to enter the site survey. For both types of online sampling recruited, the researcher should already have a database about the "potential" sample.

Online sampling technique offers an attractive benefit that is cost for research is cheaper that other techniques, and the response obtained is faster. However, we must consider the aspects of reliability and integrity of the sample when using online sampling technique. To maintain the reliability and integrity of online sampling you can read some tips below:

• Password protection . When you decide to send survey invitation to your prospective respondents by email, you should not forget to complete this email with security information and password to access the online questionnaire. This password is useful to ensure that only reliable respondent can access the questionnaire. It also helps you ensure your respondents will answer the questionnaire until the end.


• Reminder invitation. You need to remind your potential respondents so that they will not miss to join your survey. Getting respondents is not an easy job therefore you should not lose any potential respondents you have gained. Besides, reminding the respondent also helps you get the answer at the time - data processing will take quite long time so that we need to manage fieldwork process efficiently, especially in time spent.


• Summary of the survey findings. Usually, the researcher presents a temporary summary of the survey finding on the site in order to improve the response rate from the respondents. Some researchers somehow believe this action can motivate respondents to answer the questionnaire.