Business Research Design : Health Care Decision Making

1: Research design

This particular study used a cross-sectional design. What are the advantages and disadvantages of using a cross-sectional design in this research? Suppose they were to extend this study to become a longitudinal design, what are some of the issues that should be addressed?

2: Self-completed questionnaire

In this study a Self-completed questionnaire was created and send out to the respondents. Discuss some of the problems and limitations of Self-completed questionnaires that would have to be addressed in this study

3: Secondary data

In any quantitative study we want to believe that the sample selected is a representative sample of the true population. One of the methods to check representativeness is to use Secondary Data. What secondary dataset can be used to check the representativeness of the sample and how can it be used?

4: Sample size

The sample size for this study is fifteen thousand employees selected from a total of sixty-nine thousand bank employees (about 21% of the employees). What factors should be considered in decision on sample size? Discuss the advantages and disadvantages of having a sample of this size.

5: Sampling method

To find the relationship between job characteristics and job satisfaction at employee individual level, this study randomly selected from employees from each of the banks. What are the advantages and disadvantages of the current sampling method? What are your suggestions to improve the sampling methods?

The case study highlights the job satisfaction of the employees working in the banking industry with special focus on Belgian banks. The theoretical framework carried forward in the research is quantitative in nature. However, the research consortium was selected based on cross sectional design, self-completed questionnaires which were send out to respondents, with a sample size of 15,000 selected from 69,000 respondents using random sampling method.  Conversely, the design, approach, sampling as well as data collection method followed needs to be reviewed in the study based on the questions below.

Cross sectional design used in the study is to evaluate the subset of the employees working in Belgian banks. The population considered for the study was 69,000 but due to the inefficiency of questioning all the employees, the study had taken on a fixed sample of 15,000 employees, which considered 21% of the sample.

The cross sectional study is undertaken because of following reasons because it is an observational study that not only collects data from the study population of Belgian banks at a single point but also examines the impact of job characteristics on stress and wellbeing. Secondly, the researcher has adopted this research to study the prolonged effect of job characteristics resulting in stress as well as wellbeing. However, this cross sectional study is analytical in nature because it is studying the association between variables to draw valid conclusions from the result.

The possible advantages undertaken in the study constitute that the study is less costly as it is a representative of the population rather smaller sub population. Moreover, it is preferably good for both analytical as well as descriptive analysis as in the case study stated. In addition, it is termed to be good for generating hypothesis as well.

On the other hand, the possible disadvantages can be stated to be susceptible to biasness as well as the risk of validation with adherence to less reliability (O’Sullivan, 2015). Secondly, it does not relate to causation but only associations whether the two variables – wellbeing and stress are related to job characteristics or not. Thirdly, the differences seen can be also due to effects of time or age depending on how much time it takes to collect the study and perform the results (Pedersen et al., 2016). Fourthly, it is unequally distributed between groups as from 69,000 population, 15,000 have been selected but the sample does not represent equal numbers of employees from different Belgian banks. Lastly, there is existence of variability in inter-subject because stress is a negative concept whereas well-being is positive concept as they are not segregated in carrying out results.

Nonetheless, if the study is extended to longitudinal study then there are certain issues that needs to be considered. Firstly, longitudinal study requires a lot of time because it has embarked on equally long periods to collect the data (Gelman et al., 2014). Secondly, the risk gathering data cannot be considered 100 percent reliable because observation periods are not taken into account. Moreover, there is a probability that respondents can change their answers to better suit the objective of the observer. Thirdly, there will be risk in panel attrition where there is a high possibility that the respondents/ subjects would no longer be able to participate because of refusal of contact details, death and incapacity (Twisk 2013). Fourthly, they need to have large sample size, in this case it might be considering whole of the sample for instance and lastly, it can be considerable more expensive than cross sectional studies.

Self-completed questionnaires and structured interview are more or less similar methods that are used in social research (Butler et al., 2013).

The following problems as well as disadvantages in self-completed questionnaires can be discussed with reference to the recent case study enlisted that the respondents are liable to answer the questions based on their understanding, as there is no interviewer or researcher to help the respondents. Moreover, if the questionnaire remains incomplete then the study might be affected (Bryman, 2015). Moreover, the researcher would not be able to complete additional data, as it will not be feasible based on postal questionnaires. The other problem that can be highlighted is that the data can be missed at important variables because respondents depending on their comfortableness will answer some questions and leave out the rest (Britt, 2014). Although, the questionnaire is sent by post, it cannot be justified whether the right respondents from Belgian bank is answering the questionnaire or but if not right respondents then the questionnaire can be delegated to a junior officer if sent in an organization or to another family member if sent at home. However, it is impossible to have any control over the intrusion of non-respondents.

 The major issues considered is the low response rates that is most threatening as well as damaging limitation because lower the response rate, more questions would be raised representing the sample (Williams, 2014). This is likely to be an issue of randomly selected sample because as seen in the case study only 15,000 employees have been taken from 69,000 employees of Belgian Banks such that number of employees from one Belgian Bank will differ from the other. Moreover, then there is likeliness of biasness in the data. However, if the rate of response is less than 50% then the risk of bias will be greater on the findings, which can lead to further variability. Moreover, the further suggestion that could be made is that the researcher could have mixed self-administered questionnaire with telephonic or face-to-face interview wherever possible. This could have been done based on two parts such that general part of the survey could have been constituted to questionnaires whereas the specific part could have carried forward using telephonic or face-to-face interview, which could have majorly solved the issue of response rates (Bryman, 2015).

Secondary Data analysis is the data collected by a third party that are examined to answer a research question other than the research question that was selected based on the initial collection of the data set. However, secondary data analysis is an empirical analysis that carries those studies utilizing the primary data and the further the steps followed in the research method (Bryman & Bell, 2015). Moreover, secondary data analysis is not only considered a viable method in the process of inquiry but also follows a systematic process before carrying out the results (Thomas, Silverman & Nelson, 2015).

On the other hand, secondary data analysis is a flexible as well as validated as the research remains to be under-used technique in many fields. Moreover, the area of investigations highlights the way the researcher collects the data from other source, analyzes as well as interprets the data in the study. However, the importance lies when according to research questions, the data is identified and through evaluation is performed (Johnston, 2014).

The data set can be identified based on the supportive research question as well as supported literature that instills the data to be collected. The data is mostly collected from large government funded datasets, school records, journals, college records as well as different authors websites. Conversely, after collecting the data set from different sources evaluation is performed in which the research is able to evaluate its initial requirements of the research followed by appropriateness of the research topic (Stangor, 2014). The secondary data analysis constitutes to have benefits like saving time and money because there is international access to cross-historical data that would take years to complete. Secondly, the data extracted from other sources can be of high quality because studies funded by government involve larger samples. Thirdly, the data sets accompany many variables enabling validity of the study (Fowler Jr, 2013).

According to Richey & Kelvin (2014), there are following methods that can be enlisted in ensuring appropriate match of a dataset as well as congruency, representativeness, quality of the primary study and the resulting dataset are given as the following. Firstly, to evaluate the purpose of the study with the dataset. Secondly, to ensure the questionnaire is written so that the comparison can be made with the survey. Thirdly, the information of the period depends on the actual dataset collected whether weighed, discarded or used the dataset, as it is (Zhang et al., 2014).

The consequential research undertaken for determining the sample size studies the statistics that drive sample size decisions (Berger et al., 2014). However, there are certain factors that needs to be determined before evaluating the study. They are level of precision, confidence level and degree of variability.

The degree of precision decides that how close an estimate is to the actual population characteristics. It can be often termed as sampling error. However, according to Lemy & Lemeshow (2013), the desired precision enables the number of errors that can be accepted in the proposed sample. It relies on upon the measure of risk an analyst is willing to acknowledge while utilizing the data to decide. It is frequently stated in percentage.

The confidence level is ascertained from the normal distribution probability model (Haberman, 2014). The model can be given as:

Moreover, if the sample statistics is drawn more from population then it tend to deviate from the population parameter. When the sample size is large, that is above 30 then the distributionis normal distribution. However, sample estimate is a true estimate for population parameter. Nevertheless, the confidence level depicts the level at which the error toleration in the sample statistic does not go beyond the specification of planned precision. For example if the population parameter needs to analyze on 95% confidence level then it can be said that 95 out of 100 sample will evaluate the precision set by the researcher.

The degree of variability attributes on the distribution of the population such that more varied a population, larger sample will be needed whereas less population, smaller samples will be worked on (Wolf et al., 2013). However, when it comes to analysis, 50% depicts a larger variability because proportion of 0.5 will be determined more often in conservative sample size.

The advantages of the case study is that the sample size being 15,000 samples depict 21% of the sample, which depict less variability as the sample size under taken is of the same distribution of population that is Belgian banks. Moreover, the anticipated sample proportion undertaken is 20%, which comes under absolute sample precisions undertaken on true value rather on population parameter (Feder & Pfeffermann, 2015)

The main disadvantage of the case study is that the number of sample is not evenly distributed based on the number of employees from each Belgian bank. However, the result may be correct for the whole population but not when analyzed on individual sample from each Belgian bank. However, it is recommended that the data set should be followed based on the factors mentioned above such that not the conventional rather the practical response rate is followed that constitutes the response rate of 30% that can be term to be effectual in answering responses. Moreover, it is required a proportion of employees are taken from each Belgian bank to validate the study more efficiently and effectively (De Vaus, 2013).

Sampling method is advocated in every research to study the sample from the given population. The sampling methods helps in determining the closeness of a sample to the population (Singh & Mangat, 2013). The sampling method undertaken in the given case study is simple random sampling method. However, according to the case study, the advantages can be large sample, which proves to be beneficial for the research. It did not require any additional information like gender, age, etc. as it was based on employees of Belgian banks and it is highly representative of the sample. The possible disadvantages on current sampling lead to poor presentation of the parent “population” because the study area is large. Moreover, it is costly as well as time consuming (Blair, Czaja, & Blair, 2013).

The suggestions that can be provided on the current sampling method is that the sampling method that needs to be followed is stratified random sampling method in a way the sample could have divided in sub-groups (Fowler Jr, 2013). This would be emphasizing on different employees from different Belgian banks and not as a total. However, this can make the study more representative of the population. Moreover, sample within the strata will be homogeneous but heterogeneous across the strata of samples (Bornstein, Jager & Putnick, 2013).

Berger, M. L., Martin, B. C., Husereau, D., Worley, K., Allen, J. D., Yang, W., …& Crown, W. (2014). A questionnaire to assess the relevance and credibility of observational studies to inform health care decision making: an ISPOR-AMCP-NPC good practice task force report. Value in health, 17(2), 143-156.

Blair, J., Czaja, R. F., & Blair, E. A. (2013). Designing surveys: A guide to decisions and procedures. Sage Publications.

Bornstein, M. H., Jager, J., &Putnick, D. L. (2013). Sampling in developmental science: Situations, shortcomings, solutions, and standards.Developmental Review, 33(4), 357-370.

Britt, D. W. (2014). A conceptual introduction to modeling: Qualitative and quantitative perspectives. Psychology Press.

Bryman, A. (2015). Social research methods. Oxford university press.

Bryman, A., & Bell, E. (2015). Business research methods. Oxford University Press, USA.

Butler, C. C., Simpson, S. A., Hood, K., Cohen, D., Pickles, T., Spanou, C., …&Kinnersley, P. (2013). Training practitioners to deliver opportunistic multiple behaviour change counselling in primary care: a cluster randomised trial. BMJ, 346, f1191.

De Vaus, D. (2013). Surveys in social research. Routledge.

Feder, M., &Pfeffermann, D. (2015). Statistical inference under non-ignorable sampling and non-response. An empirical likelihood approach.

Fowler Jr, F. J. (2013). Survey research methods. Sage publications.

Gelman, A., Carlin, J. B., Stern, H. S., & Rubin, D. B. (2014). Bayesian data analysis (Vol. 2). Boca Raton, FL, USA: Chapman & Hall/CRC.

Haberman, S. J. (2014). Analysis of qualitative data: Introductory topics. Academic Press.

Johnston, M. P. (2014). Secondary data analysis: A method of which the time has come. Qualitative and Quantitative Methods in Libraries, 3, 619-626.

Levy, P. S., &Lemeshow, S. (2013). Sampling of populations: methods and applications. John Wiley & Sons.

O’Sullivan, B. J. (2015). Adherence in HIV-positive women entering antenatal care on antiretroviral therapy: A cross-sectional study (Doctoral dissertation, University of Cape Town).

Pedersen, H. B., Helmer-Nielsen, M., Dieperink, K. B., &Østergaard, B. (2016). Exercise on Prescription: A Cross-sectional Study With Self-reported Outcome. Journal of physical activity & health, 13(4).

Richey, R. C., & Klein, J. D. (2014). Design and development research: Methods, strategies, and issues. Routledge.

Singh, R., &Mangat, N. S. (2013). Elements of survey sampling (Vol. 15). Springer Science & Business Media.

Stangor, C. (2014). Research methods for the behavioral sciences. Nelson Education.

Thomas, J. R., Silverman, S., & Nelson, J. (2015). Research Methods in Physical Activity, 7E. Human kinetics.

Twisk, J. W. (2013). Applied longitudinal data analysis for epidemiology: a practical guide. Cambridge University Press.

Williams, A. (2014). How to… Write and analyse a questionnaire. Journal of Orthodontics.

Wolf, E. J., Harrington, K. M., Clark, S. L., & Miller, M. W. (2013). Sample size requirements for structural equation models an evaluation of power, bias, and solution propriety. Educational and Psychological Measurement,73(6), 913-934.

Zhang, Y., Chen, M., Mao, S., Hu, L., & Leung, V. C. (2014). Cap: Community activity prediction based on big data analysis. IEEE Network,28(4), 52-57.