CHAPTER THREE
3.0 RESEARCH METHODOLOGY
3.1 Introduction
Procedures utilized in research activities are referred to as Research methodology. This covers data collecting methods, statistical techniques for establishing relationships between data and unknowns, and method accuracy for evaluating the correctness of the results acquired (Kothari & Garg, 2014). The study and sample strategy, data collecting, and data analysis methods used in this body of work are all discussed in detail in Chapter 3.
3.1 Conceptual Framework
Figure 3.1: Conceptual Framework
Employee Performance in ABC Biscuit Company – Sri Lanka is the study’s dependent factor, as shown in Figure 3.1. In the scope of institutional HRM procedures, three independent variables have been recognized: Staffing Practices, Training Practices, and Incentive practices. The primary idea behind the structure is to determine the influence of HRM strategies on firm employee performance. As per the independent variables, three hypotheses will be created.
3.2 Area of the Study
As per the firm’s yearly reporting, the total number of employees is 1280, including executives, non-executives, and other support workers. Nevertheless, due to limits and constraints, the research has been limited to personnel involved in the production procedure within the plant grounds.
3.3 Target Population
A population is a collection of all sets of units that fulfill a set of characteristics; it may be thought of as a comprehensive gathering of pertinent analytics or data pieces (Hair et al., 2016). According to the firm’s yearly reporting, there are 720 people involved in the production procedure. As a result, the survey’s targeted demographic is the 720 personnel of the ABC Biscuit Company in Sri Lanka.
3.4 Sample Size
The sample size refers to the number of participants included in a research study to represent a community. The sample group relates to the total number of active research participants, and it is typically subdivided by parameters like age, gender, and location to ensure that the total sample represents the whole community (Denscombe, 2014). The sample size for the research will be computed using the preceding equation, given that the sample size was a headcount.
n = N / 1 + Ne2
n: Size of the sample
N: Total population
e: Level of Precision (Margin of error)
3.5 Sampling Techniques
A sample is a set from which data is collected in a research study. Since the investigator does not have an opportunity to examine or send questionnaires to everybody in the area, the sample size is always smaller than the study population. As a consequence, the process of identifying persons to be incorporated in the sample is referred to as sampling. Furthermore, sampling is critical because the population of this research is broad, diversified, and dispersed along a large geographical area. Since there were no issues, simple random sampling will be used in this inquiry (Hair, Page, and Brunsveld, 2016).
3.6 Data Collection Instruments
A shuttered five-point Likert-type questionnaire will be used to collect the primary data. The questionnaire asked a sequence of queries precisely on the independent variables (Omolo, 2015). This type of data collecting has the benefit of being low-cost. It also allows participants adequate time to react with their best response or option; large samples may be delighted while the results are more dependable.
3.7. Reliability & Validity
The word “validity” refers to whether the results were about what they said they were about (Kimberlin and Winterstein 2008). An analyzer was employed well before the data was acquired to double-check the timeline during the early stages. In the appropriate fields of study, all inquiries have been pre-tested. Adjustments were performed before data collection to examine conceptual relevance, perceptions, and the coherence of vocabulary used to describe ideas, leading to a validity test. The pre-testing of the questionnaires also helped to detect unnecessary, confusing, or redundant items.
Feedback regularity, in which repeated measures provide the same result throughout time and among researchers, is the subject of reliability evaluation (Kimberlin and Winterstein, 2008). Retest reliability, alternate form reliability, and inner coherence reliability are the three parameters used to measure reliability testing (Kimberlin and Winterstein 2008). Cronbach’s alpha value is a commonly used statistic for measuring the reliability of measurement or set of equipment (Katanda, 2004). A 95 percent confidence range was determined during the design phase to meet appropriate data dependability standards. Furthermore, using the analytical program for social science (SPSS) and Spreadsheets, the investigator used Cronbach’s test to analyze the reliability of the revised data.
3.8 Data Analysis Methods
Based on data analysis, the investigator was able to determine the impact of HRM practices on the productivity of ABC Biscuits Company. Quantifiable datasets will be analyzed using the Statistical Package for Social Sciences (SPSS) version 23 and Microsoft Excel 2010. To explore and identify the influence of independent and dependent variables, a confirmatory factor, comprehensive, rigorous tools, bivariate linear econometrics, and multicollinearity analysis will be employed. This research will utilize multiple linear correlation coefficients (r) to examine the type and intensity of the relationship between HRM strategies and personnel productivity. The link between the independent and dependent parameters will also be investigated using the multiple linear regression techniques. MS Excel was utilized to examine the demographic data to determine the respondents.
3.9 Ethical Implications
Participants’ sensitive data is acquired and retained during the testing procedure to answer a particular research issue. Throughout the study process, it is critical to maintaining a high level of high standards. In research, integrity can be described as ethical behavior while engaging with persons involved in or affected by the inquiry (Saunders et al., 2007 ). According to Cooper and Schindler, ethics are “laws and conventions that impact the moral element of behavior.” Two philosophical perspectives dominate management and business studies.
The first is deontology, which states that deceit should never generate new empirical evidence. The teleological method attempts to arrive at this conclusion by weighing the advantages of research findings against the penalties of bad behavior. This strategy usually leads to a moral issue since it is hard to explain a loss for one group to benefit the other. To recap, variations ought to be shunned if feasible, and if they have to be utilized at any point during the research procedure, they have to be well described and justified. Due to time constraints, a respondent may not examine the topics and just ticks at random; it is also challenging to know how genuinely a respondent fills the surveys (Saunders et al., 2007 ). There are several ethical considerations to be made.
The confidentiality of the respondent’s oughts to be protected, and they ought to be notified that involvement is entirely voluntary, with the option to leave at any moment. The respondent has to be treated with dignity to avoid pressure, stress, pain, suffering, or injury. Those few who agree to participate in the poll will be asked to sign a permission form and told how long it would take (Saunders et al . 2007 ). When partaking in a scientific study, respondents have to be told about the eventual use of the data acquired and ensured that the data will not be used for any other reasons to conform with business research ethics. All data have to be validated in a private and protected setting. Personal data must be collected and stored by data security regulations.
3.10 Chapter Summary.
The entire research process, including data collection, analysis, and conclusion, is shaped by research techniques. Furthermore, data were cleaned to detect and eliminate mistakes such as missing information, incomplete information, and data out of bounds. After that, the data were classified into numbered classifications before the actual data entering began. Ultimately, the information was analyzed, and conclusions about the population were formed from the sample.