However, it is quite not really possible for a researcher to make use of the entire large population for his study; this could be as a result of limited financial resources of the researcher, too large area to cover which may not be easily covered by the researcher or it could be as a result limited period allocated for such study. Hence, there is need to draw out some fraction known as sample size out of the entire population which can be easily managed by the researcher. Basically, Popoola (2011) posits that the major components of research methodology especially in Library and Information science or in any research work include: research design; the population of study; sampling procedure and sample size; research instrument(s); validity and reliability of the research instrument(s) and methods of data analysis.
Nevertheless in this paper, we shall be dealing with how to choose the best sampling technique and sample size for a study. The paper shall lay more emphasis on population of the study, also educate our readers on how to choose a sample technique and sample size for any study they are presently undergoing or intend to carry out in future.
Population of study
Population of study for any research work has been variously defined by different scholars and their definitions pointed toward the same direction. Avwokeni (2006:92) refers to population of study as the “set of all participants that qualify for a study” while Akinade and Owolabi (2009 :72) defined population as “the total set of observations from which a sample is drawn”; Adeniyi et al (2011:49) see it as the “total number of large habitations of people in one geographical area, for example, the population of a country”; besides, Popoola (2011:2) defines population as the “totality of the items or objects under the universe of study. It often connotes all the members of the target of the study as defined by the aims and objectives of the study”. We concur with the definitions of the above scholars. But we conceptualize population to mean the whole body of items, objects, materials or people that fall within the geographical location in which a researcher intends to investigate for his study. That is the whole participants of a study. The constituents of population have certain attributes in common; the number may be large or small. What should a researcher do if the population is large or small?
However, as earlier pointed out in this paper, it is quite impossible for a researcher to study the entire population due to some reasons. In objecting this view, Popoola (2011:2) points out some reasons under which a researcher may study the entire population, these include:
- When the size of the population is small.
- When there is no time constraint in carrying out the study. That is if there is enough time for the researcher to carry out his study.
- When the resources (human, money and materials) available for the study are adequate.
- When the sole objective of the study is to take census of the elements in the population.
Types of Population
According to Akinade and Owolabi (2009:73) and Adedokun (2003:43-44) population of a study can be classified to include:
- Listed population: All the elements or subjects are known and can be identified by name or number e.g. students who registered for a particular course or all employees in an organization.
- Homogeneous population:Population that consists of discrete elements that have common property with very little variation such as same age, sex, level of education and rank in a job, or people of the same skin colour (black or white).
- Heterogeneous population: Here the population consists of different categories of participants or elements that have uncommon characteristics e.g. children in all classes in a school or all categories of workers in an organization (lecturers, accountants, clerks, secretaries, managers, directors and artisans). This type is further subdivided into:
- Unlisted population:This is a very large and nebulous type. The elements cannot all be identified by names or numbers e.g. all the men or women in a country.
- Group population:This is a structured and listed population e.g. students offering a course of study may be classified into male and female; local and foreign; single or combined honours students; workers (junior, senior) management and supervisory staff.
- Scattered population:Population of listed members who are found in different geographical locations e.g. members of Nigerian Library Association or lecturers of private universities of a country or lawyers in a country.
- Clustered population:Population of unlisted individuals who are known to exist together in different locations e.g. Nigerian footballers in European and Asian countries.
- Target population:This consists of all members of a group of under the investigation to which the result of the investigation can be applied. They are population of interest (e.g. university students) but all of whom are not of interest to the researcher.
- Accessible population: This refers to the members of a target population, which are within the reach of the researcher, work with and obtain their sample for his study.
Furthermore, it is pertinent to note that characteristics of a population of scores are called parameters and characteristics of a sample of scores from a larger population are statistics. The mean of an entire population of scores is a parameter and the mean of a sample is a statistics (Akinade & Owolabi, 2009). However, when the population is relatively too large to be covered within the period allocated for the study and also with the limited human and material (fund) resources of the researcher, it is necessary for the researcher to take sample size from the entire population for the study.
Sample and Sample Size
A sample is a manageable section of a population but elements of which have common characteristics. Also, it refers to any portion of a population selected for the study and on whom information needed for the study is obtained ( Awoniyi; Aderanti & Tayo, 2011; Akinade & Owolabi, 2009; Adedokun, 2003). It is the elements making up the sample that are actually studied and generalizations or inferences about the population are made. This generalization of result based on the sample to the population is the major purpose of sampling and also a major concern in any scientific investigation. It can be reemphasized here that, to study the entire population may be cumbersome, time consuming and of course very costly, hence, a sample takes a fair portion as representative of the entire population. In statistical analysis, population of the study is being denoted with ‘N’ while sample size is denoted with ‘n’
Sample size is the number of elements that can be selected for a research. This number varies from one study to another. In homogenous population (where there is little variation) requires a small size. Experimental studies tend to use relatively small sample size. But for heterogeneous population (where there is a wider variation) requires a larger sample size. This is common in survey research as in education and behavioural sciences (Akinade & Owolabi, 2009).
Types of Samples
The following types of samples are adopted from Akinade & Owolabi (2009: 74), they include:
Random Sample: Is one in which all the elements have equal chances of being selected for a study. It is the basis of most of the other types of samples. For example, all the male students in a classroom.
Subjective Sample: This refers to a subset of an accessible population. It is deliberately or purposefully chosen. On the basis of researcher’s predetermined intention, selecting experimental subjects this way is open to bias. For instance, if a researcher wants to find out the reading habits of students in a university and uses the students he/she teaches. This sample is not representative of all the students in the university not to talk of the world. It is a sample that is neither random nor representative. It is a sample that the researcher has allowed his/her preferences to influence his choice of items to be selected for the study. It may not be a captive sample. However, generalization cannot be made from the results. Captive samples and biased samples are good for quick collection of data from the respondents.
Captive Sample:it is a sample obtained from a captive population such as students in a classroom or workers in an industry, patients in a psychiatric hospital ward. The elements have no population and so one cannot generate finding from it. It is not used in serious research. It may be useful for pilot study or where a researcher wants a quick result before commencing the actual study.
Purpose of Sampling
Basically, sampling is done to ensure that there is no bias or subjectivity in the selection process. It also helps the researcher to work with reasonable size of elements since it is difficult to do so with the entire population. It thus saves time spent on each research and also reduces cost of research operations (Akinboye & Akinboye, 1998 in Akinade & Owolabi, 2009).
However, according to Adeniyi et al (2011: 50) other reasons for sampling include:
- In a large population, there could be some similarities and uniformities along the line of research investigation, drawing up a sample along these similarities will give a general result for the whole population.
- It makes the researcher more thorough and affords more time for better study.
- Sampling makes data analysis easier and more skillful with qualitative results.
- Since sampling enables us to deal with a part of the population, it is obviously cheaper to study a sample rather than the entire population.
- It enables us to obtain quicker results than covering an entire population with its attendant problems.
- At times, it is practically impossible to take a complete and comprehensive study of the population because of the nature and pattern of distribution or dispersion of the population elements.
- For research study involving practical enumeration of subjects, sampling is the only best option to achieve it.
- Sampling helps the researcher to guard against incomplete and inaccurate instruments such as questionnaires.
- Sampling makes it possible to study infinite population.
Nevertheless, limitations of sampling include the following. Some samples may be too small or heterogeneous to collect a representative sample. In some cases, the researcher may not be equipped with sufficient knowledge of diverse or relevant sampling methods in that case there may be bias (Akinade & Owolabi, 2009: 75).
Sampling techniques can be classified into two main groups, namely: probability and non-probability sampling.
- Probability Sampling Technique:
Here all the items/units in a population have equal chance of being selected as sample for a study. The advantages of this method of sampling is that it only when the items have been selected with known probabilities that one is able to evaluate the precision of the sampling result (Popoola, 2011; Adeniyi et al, 2011). The examples include:
- Simple Random sampling
- Systematic sampling
- Stratified sampling
- Cluster/Area sampling, and
- Double sampling.
Each of the listed probability sampling techniques shall be discussed in turn:
- Simple Random Sampling:
In simple random sampling, every member of the population has equal chance of being selected for a study. It is a method that gives each member of the population non-zero probability of being selected. This type of sampling technique is used when the population has similar characteristics (homogeneous population), the sampling frame is available and the population size is determinate or finite. To ensure a random sample, the selection of samples may be done through balloting, table of random numbers and computer simulation. In most cases, the sampling is done without replacement (Popoola, 2011).
- Systematic Sampling:
This involves the selection of every nth subject or item from serially listed population subjects or units. Where n is any number determined from the population. For example, obtaining a systematic sample of 100 from a population of 800. The will be as follows:
- Number the items serially up to 800.
- Divide 800 by 100 i.e. N/n = 800/100 = 8
- Randomly select a starting point, say number ‘8’ of the population list.
- Then select every 8th unit after the first. The list will include the following: 8th, 16th, 24th, 32nd, 40th, 48th, 56th, 64th, 72nd etc. on the population list (Adeniyi et al, 2011).
- Stratified Random Sampling:
This is a method in which the heterogeneous population is first stratified by dividing it into a set of mutually exclusive or non-overlapping sub-populations or strata, and thereafter random samples are then selected from each stratum for detailed study (Popoola, 2011). It must be noted that the stratified random sampling may involve stratifying the respondents into infected and uninfected; males and females; old and young; residents and non-residents; employed and unemployed; literates and illiterates; rich and poor. The author reiterates that four basic methods could be employed to determine the sample size per stratum or total sample size for all the strata. These include probability proportionate to size (PPS), optimum allocation, Neyman allocation and equal allocation.
- Cluster/Area sampling:
This type of sampling technique is similar to stratified random sampling that had been discussed in this paper. Cluster or area sampling is otherwise known as multistage sampling in that the population is subdivided into units and a random sampling of smaller units; that is, it is a population that exists in clusters over a geographical area. The selection of individual cases in the group may also employ simple random technique (Awoniyi et al, 2011). This method is used when the population is very large and covers a large geographical area. For example a country can be divided into regions or states which constitute sampling units called clusters.
- Double sampling:
A double sampling method is a sampling design in which some information is gathered from the whole sample and additional information is either at the same time or later collected from sub-samples of the full sample. With only one sub-sample, the technique is called two-phase sampling or double sampling (Popoola, 2011:8).
- Non-probability Sampling Technique:
Here the elements or items of the population to be studied have no equal chance to be selected for the study. Rather, randomness if it occurs is by mere chance. Examples of this sampling technique include:
- Quota sampling;
- Purposive sampling;
- Accident/Availability sampling;
- Convenience sampling;
- Snowball sampling; and
- Event sampling.
Each example of the non-probability sampling techniques listed above shall be briefly discussed in this write up:
- Quota sampling:
Quota sampling method is a subjective method of selecting samples from a population of study. It is always difficult to determine the probability of drawing samples from a study population. Also, quota sampling is the type of sampling scheme in which deliberate control factor is used to draw samples from a study population on the assumption that the chosen samples have similar characteristics with the sampling population (Popoola, 2011; Hammed & Popoola, 2006).
- Purposive/Judgmental sampling:
This is otherwise called judgmental sampling. Judgmental sampling is a non-probability sampling techniquewhere the researcher selects units to be sampled based on their knowledge and professional judgment. It is a sampling procedure that is characterized by a deliberate effort to obtain representative samples from a study population (Explorable.com, 2009). Purposive sampling is used in cases where the specialty of an authority can select a more representative sample that can bring more accurate results than by using other probability sampling techniques. The process involves nothing but purposely handpicking individuals from the population based on the authorities or the researcher’s knowledge and judgment. For example, in a study wherein a researcher wants to know what it takes to graduate summa cum laude in college, the only people who can give the researcher first hand advise are the individuals who graduated summa cum laude. With this very specific and very limited pool of individuals that can be considered as a subject, the researcher must use judgmental sampling (Explorable.com, 2009). However, two major drawbacks of this method are that of reliability and the bias that accompanies the sampling technique.
- Accident/Availability sampling:
It is a sampling procedure whereby a researcher selects available samples without any scientific bases. This type of sampling procedure must be avoided when conducting a research in any research activity.
- Convenience sampling:
A convenience sample is simply one where the units that are selected for inclusion in the sample are the easiest to access. This is in stark contrast to probability sampling techniques, where the selection of units is made randomly. For example in a population of 10,000 university students; where the researcher is only interested in achieving a sample size of 100 students who would take part in his research. As such, he would continue to invite students to take part in the research until his sample size is reached. Since the aim of convenience sampling is easy access, he may simply choose to stand at the main entrance to campus of the University where it would be easy to invite the many students that pass by to take part in the research (Lund Research, 2012).
Moreover, the merits of convenience sampling method include: it is very easy to carry out with few rules governing how the sample should be collected; relative cost and time required to carry out a convenience sample are small in comparison to probability sampling techniques and the convenience sample may help the researcher in gathering useful data and information that would not have been possible using probability sampling techniques, which require more formal access to lists of populations. However, the convenience sampling technique has the following limitations: it often suffers from biases from a number of biases. This can be seen in the example aforementioned and since the sampling frame is not know, and the sample is not chosen at random, the inherent bias in convenience sampling means that the sample is unlikely to be representative of the population being studied. This undermines researcher’s ability to make generalizations from his sample to the population he is studying (Lund Research, 2012).
- Snowball sampling:
This has to do with a situation when a researcher builds up a sample for his study through information from other people. He starts with one person who then suggests another people and so on until the researcher gets enough sample size for his study.
- Event sampling:
According to Awoniyi et al (2011), it involves a situation when a researcher uses the opportunity presented by a particular event to gather information (sample) for his study. For example if a researcher is investigating the perceptions of Nigerian people towards 2014 FIFA world cup tournament, he can get his sample size from the football fans during the ongoing world cup event.
Thus far in this paper, we’ve tried to do justice to population and various types of sampling techniques that can be used in any research activity. However, researchers are not restricted only to what we’ve presented, the choice of which sampling technique to use depends solely on the researcher and the investigation he intends to carry out. Therefore, conducting a scientific investigation in any human endeavor is a peculiar task globally; hence, the accepted process must be taken in order to come up with a valid result.
Adedokun, J.A. (2003). Basics of Research Methodology. Sagamu: New Hope Publisher.
Adeniyi, A.L.; Oyekanmi, A.O. & Tijani, M.O. (2011). Essentials of Business Research Methods. Lagos: CSS Bookshops Limited.
Akinade, E.A. & Owolabi, T. (2009). Research Methods: A Pragmatic Approach for Social Sciences, Behavioural Sciences and Education.Lagos: Connel Publications.
Avwokeni, J.A. (2006). Research Methods: Process, Evaluation & Critique.Portharcourt: Unicampus Tutorial Services.
Awoniyi, S.A.; Aderanti, R.A. & Tayo, A.S. (2011). Introduction to Research Methods. Ibadan: Ababa Press Ltd.
Explorable.com (2009). Judgmental Sampling. Retrieved from https://explorable.com/judgmental-samplingon Jun 27, 2014.
Hammed, A. & Popoola, S.O. (2006). Selection of Sample and Sampling Technique: In G.O. Alegbeleye; I. Mabawonku & M. Fabunmi (ed). Research in Education. Ibadan: Faculty of Education, University of Ibadan: 138 – 154.
Lund Research (2012). Convenience sampling. Retrieved from http://www.dissertation.laerd.com/convenience-sampling.php on 27/06/14.
Popoola, S.O. (2011, September). Research Methodologies in Library and Information Science. A paper presented at a training workshop on building research capacity for Library and Information Science professionals. Organized by the Nigerian Library Association, Ogun State Chapter, held at Covenant University, Ota, Nigeria on 18th – 22nd September, 2011