Sampling Method and Sample Size
A convenient sampling method was employed. Every staff who met the inclusion criteria outlined in the eligibility checklist was included in the study until the desired sample size was met. Convenient sampling was to be adopted in the research in order to eliminate voluntary response bias as well as to eliminate under coverage bias. The research population was made up of the entire 220 staff (170 junior staff and 50 senior management staff) employed by the organization while the sample size was made up of 22 senior management staff and 75 junior staff. Both the senior and junior management staff were considered for the study because they were both responsible for the daily running of activities of the organization, and they all interacted with customers at one level or the other. The senior management staff comprised of heads of departments and strategic business unit managers who oversaw the various departments of the hospital while junior staff comprised the nurses, assistant nurses, pharmacy technicians, records personnel and other operational staff.
Staff was included if they had worked in the hospital for at least five years. Staff who had worked for less than five years were excluded.
Data Analysis Approach
Analysis of data would be carried out in descriptive statistics with a five (5) point Likert scale. The advantage of Likert scales is that they produce a cumulative response process, i.e., they make it possible to sum up, the scores and produce a cumulative score. Likert scales enhance speed, confidentiality, and collection of honest feedback. However, it is possible for the respondents to give false information in an attempt to maintain privacy. Lack of understanding of the question asked may also make respondents give wrong feedback. Another major shortcoming of Likert scales is difficulty in achieving the internal consistency of the scale, which can affect result interpretation. Also, intervals between the different points on the scale may not be a true representation of the changes in individual attitudes.
In view of the hypotheses stated above, nominal and ordinal data would be collected for the research because they were both classified as categorical data, i.e., non-quantitative and could be represented by string variables. Nominal data do not have quantitative values and are used for the ‘naming’ of variables, as the name implies. Ordinal data, on the other hand, have their variables arranged in an ordinal form but are not measured mathematically. They are in the form of labels used form representing opinions. Similarly, a nonparametric test would be conducted since the hypothesis was aimed at determining how senior management and junior staff contributed to organizational performance as well as the impact of Training and development on the organization’s competitive advantage. A nonparametric test was considered because the analysis does not require data to fit a normal distribution. Nonparametric tests rely mostly on ranking and orders in which data are sorted.