Proper Sample Size for Representing Population
Discuss about the Quantitative and Qualitative Job Insecurity.
The sample size that is required in order to represent a population is based on the population size of the subject which is of interest to a study. A sample of size larger than the required size will involve greater cost and will consume larger time for the data collection process. This will be inconvenient for the study purpose. Again a sample of size smaller than the required sample size will not be able to represent the total population properly (Hopkins, 2017). Thus, the results obtained will not be reliable.
In this study, the size of the population, that is the total number of employees in all the 63 Belgian Banks is 69,000. A sample of size 15,000 which is approximately 21 percent of the whole population is selected for the study. But this sample size is too large to study a population of size 69,000. The sample size that can represent this population is 384 (Chow et al., 2017). A sample of size 400 can be considered for the survey, but more than that will cause inconvenience to the study.
In the current scenario, there are 63 Belgian Banks and it was compulsory for all the banks to participate in the study. 21 percent of the total number of employees working in these 63 Belgian Banks have been selected as the sample. The sampling was done in such a way that roughly 21 percent employees were selected from each of the banks. 21 percent of the employees were selected from within each bank randomly from all the employees in that particular bank. No preference was given to any other factors such as gender of the employees, age of the employees or the level of the employees at the time of sampling. Thus, a sampling of this type is known as stratified random sampling (Robinson, 2014). Here, all the 63 banks were the 63 strata to which the population was divided and 21 percent of the employees were selected randomly from each stratum.
Stratified random sampling is a particular type of probabilistic random sampling (Hill & Brierley, 2017). There are certain advantages and disadvantages of this particular type of sampling. The respective advantages and disadvantages are discussed as below:
- This method usually gives better estimates than random sampling. This is because, the strata being homogeneous, the estimates obtained from them are likely to be satisfactory even when the sizes of the samples from different strata are small. These good estimates, when suitably combined, give a good estimate for the whole population (Robinson, 2014).
- The sampling procedure here ensures that some members from each stratum are included in the study; so that the combined sample observation serve as a good representative of the population under enquiry (Hill & Brierley, 2017).
- Stratified Random Sampling supplies not only an estimate of the whole population, but also separate estimates for the individual strata (Robinson, 2014).
- Sometimes Stratified random sampling is administratively very convenient. The existing administrative set-up at different zones may be used to facilitate the organization of the field work (Hill & Brierley, 2017).
- In some situations, different segments of the population demand different sampling procedures. For example, in human populations, people living in hostels, hospitals, prisons, etc. should not be treated in the same way as those living in their homes. Stratified sampling is very useful in such cases (Avery & Burkhart, 2015).
- A stratified random sampling can only be conducted if the total population is available for the study. Amy missing units in the population will result in error in the sampling procedure. There are several problems that can be faced to collect the complete list of the population. Access to the list might not be available due to some privacy policies. This will consume a lot of extra time to gain access to the data and then run the research (Sekaran & Bougie, 2016).
- The strata for the sampling must be well defined such that each and every population unit falls under a distinct stratum. There should not be any population unit present in the study which can fall in more than one strata. In this case, the strata are strictly defined as banks. There are no employees who can work for simultaneously for two different banks. Each employee will work for only one bank and thus, there is no way of overlapping the two strata (Avery & Burkhart, 2015).
- If the sample requirements are exceeded, then for the necessity of the study to be conducted, the predefined strata might have to be redefined and some more strata might have to be constructed. Thus, at the time of sampling, this will result in more sampling units and this will in turn lead to greater cost and more time for the research to be carried out (Bryman & Bell, 2015).
Cronbach’s alpha is a measure which gives an idea about how much reliable the question is to the study. The more the value of the Cronbach’s alpha is close to 1, the more reliable the data is (Bonett & Wright, 2015). In this study, for the measure of qualitative job insecurity, it can be seen that the Cronbach’s alpha value is 0.87 which is quite close to 1. Thus the data on qualitative job insecurity is quite reliable.
Benefits and Drawbacks of Stratified Random Sampling
Similarly, for the quantitative job insecurity measure, it can be seen that the Cronbach’s alpha value is 0.89 which is quite close to 1. Thus the data on quantitative job insecurity is quite reliable.
For the measure of psychological distress, the Cronbach’s alpha value has been obtained to be 0.89, which indicates that the data on psychological distress is also quite reliable.
One of the most important characteristics of the population that has to be studied are the demographics of the population. Different characteristics such as age, gender, occupation, income, race, etc. are some examples of population demographics that are usually considered for a research. Thus, it can be said that when a research is conducted, it is important to evaluate the population of interest for the particular study, how the responses of the survey will be broken down to form meaningful groups. Both of these two evaluations are made based on the demographic profile of the respondents of the survey (Aelenei, Lewis Jr. & Oyserman, 2017). The importance of some demographic questions are discussed below:
Gender: One of the most common question which is used almost in most surveys is the sex or the gender of the participants. It is known that the gender of a person has a huge impact on their way of thinking. This affects the survey results. The left hemisphere of the brain of men is stronger than that of the right hemisphere and both the hemisphere can be equally balanced by women. For this reason, the common emotions are more common to women while the thinking method of men is much more task oriented, from the point of view of an objective management (Ingalhalikar et al, 2014).
Age: Another important question that can be seen in almost every survey is the age of the participants. Age makes a huge difference in surveys while it may not make any difference in other events in life (Omar & de Belder, 2016). For example, in a research study about films, a teenager will be more comfortable with information about recent films than a film released 10 or 20 years back. Similarly, a senior citizen will have more information about the films released in their teen ages rather than the films that has been released recently. Thus, age does make a huge difference in opinions.
Education: Education is another common question asked in surveys. The education level of the respondents also makes a huge difference in their opinions. The point of view of a person who is highly educated will always be different from the point of view of a person who is illiterate or very little education (Conde et al., 2016).
Demographic Questions in Surveys
The research design that is used for this study is survey design. It is not always possible to study the whole population as it will be costly and time consuming as well. Thus, sampling is required. For the purpose of different studies, different types of information are required from the participants of the study. Collecting information from the participants of the study is known as survey (Patten, 2016). In this study, all the 21 percent participants from each bank were provided with a questionnaire which contained questions that the employees had to answer. Based on the answers given by the employees, analysis will be performed.
There are several merits for this survey research design. Along with the merits, there are several demerits to this method as well. Both the merits and the demerits of this type of research design are discussed below:
- High Representativeness: The sample that is selected for the purpose of the study is more capable of representing the population from which the sample has been drawn in case of a survey design. The survey usually involves a huge number of people answering several questions that are asked relative to the study. Thus, the data that is collected as a result of the survey is usually a better representative of the characteristics that are relative to the population that is involved in the study. This method can extract data that are more appropriate to the original attributes much more successfully than any other research design (Shipman, 2014).
- Low Costs: In case of survey designs, the only cost that is involved for the process to run is the cost for the production of the questionnaires required for the survey. In specific cases, where a larger unit of the population is required as a sample, incentive methods can also be introduced in kind or in cash. This can be a very little amount also. On the other hand, other data collection methods such as personal interviews or with the help of focus groups involves a much higher cost (Heeringa, West & Berglund, 2017).
- Convenient Data Gathering: There are a variety of ways with the help of which administration of the surveys can be conducted with the participants. The questionnaires can be distributed to the participants with the help of e-mail, fax, mail or with the help of the internet which is very common nowadays. This online method of data collection is convenient as well as less costly and involves participants from all over the world (Bryman, 2015).
In this study, data from participants all around the world is not required. Data is only required from 63 Belgian Banks. Thus, the participants were provided with the questionnaires through the regular mail as all the participants were accessible to the researcher. The completed survey forms were collected by the researcher by visiting each of the banks.
- Good Statistical Significance: As the survey method represents the population much more accurately, it becomes easier for the researcher to obtain statistically significant results from the data collected with the help of this method rather than all other data collection methods. Analysis of multiple variables can be analyzed effectively with the help of the data collected using the survey design (Denscombe, 2014).
- Little or No Observer Subjectivity: In most of the scientific research studies, survey research design is used as with the help of this design, the participants of the survey are provided with a stimulus which is standardized. Thus, the data becomes extremely reliable and the biases that could have taken place from the researcher’s end gets eliminated (Heeringa, West & Berglund, 2017).
- Precise Results: The data gathered is extremely precise as the questions provided in the questionnaires are usually approved after careful scrutiny of each of the questions that the participants are supposed to answer (Denscombe, 2014).
- Inflexible Design: The survey and the method of administering the survey that has been chosen by a researcher at the beginning of the research cannot be altered at any point in the data collection process. This can be considered as both a weakness and a strength of the method as the study can be more precise and fair due to this inflexibility (Bryman, 2015).
- Not Ideal for Controversial Issues: In cases when the questionnaire contains questions, answers to which are controversial, the participants may not be able to give precise answers to such questions. Many participants might face difficulties in remembering the proper information that is related to the question. The true facts might not be disclosed by the participants accurately when using a survey method. They might disclose the facts when some alternate method of data collection can be used such as focus groups or face-to-face interviews (Shipman, 2014).
- Possible Inappropriateness of Questions: The questions that are to be used for the surveys are usually standardized before sending them to the participants. Thus, the researcher has to keep in mind that the questions in the questionnaire has to be general so that the general population ca give answers to all the questions (Sekaran & Bougie, 2016).
Aelenei, C., Lewis Jr, N. A., & Oyserman, D. (2017). No pain no gain? Social demographic correlates and identity consequences of interpreting experienced difficulty as importance. Contemporary Educational Psychology, 48, 43-55.
Avery, T. E., & Burkhart, H. E. (2015). Forest measurements. Waveland Press.
Bonett, D. G., & Wright, T. A. (2015). Cronbach's alpha reliability: Interval estimation, hypothesis testing, and sample size planning. Journal of Organizational Behavior, 36(1), 3-15.
Bryman, A. (2015). Social research methods. Oxford university press.
Bryman, A., & Bell, E. (2015). Business research methods. Oxford University Press, USA.
Chow, S. C., Shao, J., Wang, H., & Lokhnygina, Y. (2017). Sample size calculations in clinical research. Chapman and Hall/CRC.
Conde, D. A., Colchero, F., Silva, R., Syed, H., Jongejans, E., Jouvet, L., ... & Steiner, U. (2016, October). Exploring Data Gaps at the Species Level: Starting with demographic knowledge. In TDWG 2016 ANNUAL CONFERENCE.
Denscombe, M. (2014). The good research guide: for small-scale social research projects management. McGraw-Hill Education (UK).
Heeringa, S. G., West, B. T., & Berglund, P. A. (2017). Applied survey data analysis. Chapman and Hall/CRC.
Hill, N., & Brierley, J. (2017). How to measure customer satisfaction. Routledge.
Hopkins, W. G. (2017). Estimating Sample Size for Magnitude-Based Inferences. Sportscience, 21.
Ingalhalikar, M., Smith, A., Parker, D., Satterthwaite, T. D., Elliott, M. A., Ruparel, K., ... & Verma, R. (2014). Sex differences in the structural connectome of the human brain. Proceedings of the National Academy of Sciences, 111(2), 823-828.
Omar, S. A., & de Belder, A. (2016). Expert Opinion Percutaneous Coronary Intervention in Older People: Does Age Make a Difference?. Interventional Cardiology Review, 11(2), 93.
Patten, M. L. (2016). Questionnaire research: A practical guide. Routledge.
Robinson, O. C. (2014). Sampling in interview-based qualitative research: A theoretical and practical guide. Qualitative Research in Psychology, 11(1), 25-41.
Sekaran, U., & Bougie, R. (2016). Research methods for business: A skill building approach. John Wiley & Sons.
Shipman, M. D. (2014). The limitations of social research. Routledge.