Levels of measurement scales and ways of measuringMeasurement scale is the way of ranging values according to different properties of a concept by assigning specific numbers to each characteristic in a reliable and valid method (Zikmund, Babin, Carr, & Griffin, 2013). Measurement scale is critical to research methods because it shows the importance of independent variables to a concept and supports the decision process. Measurement has four levels: Nominal, ordinal, interval, and ratio (Zikmund et al., 2013). In my unit three, ten survey questions were given to get employees’ expectations about labor issues with the aims of improving their loyalty. Questions from 2 to 8 are nominal scales because they classify issues into different categories and there are no orders among choices. The question 1 is an ordinary scale because it ranks the numbers of working years of employees and helps researchers evaluate the loyalty levels of workers to the company. Questions 9 and 10 are interval scales because they present the orders of scales and the distance between choices (Agriculture and Consumer Protection, n.d.). For above survey questions, the summated scale scores will be used to measure the roles of variables to employees’ loyalty. To nominal scales of questions from 2 to 8, each choice of questions will be given a label and choices which are most frequently chosen will be have highest percentages. To the ordinal scale of the question 1, researchers will ask respondents more than one time. Each choice of the questions will be presented in a number. The total results of different responding times will be analytical numbers. To the interval scales of questions 9 and 10, the levels of agreements will be given in digits from 4 to 1. 4 equals to strong agreement and 1 equals to strong disagreement. The chosen numbers will be analytical data.Measure reliability, validity, and sensitivityTo calculate the reliability of collected information, the test-retest method should be applied. Survey questions should be raised to respondents more than one time. Thanks to that, researchers can check the consistency of information provided by respondents. Validity means whether scores exactly show the characteristics of the concept or not (Zikmund et al., 2013). The higher score results are, the more important roles of independent variables to employees’ loyalty. Using questions to evaluate employees’ loyalty has been popularly used by CEOs. Concerning about short and long term goals of employees has been one of suggested questions by Belosic (2015). Plus, variables which are mentioned in survey questions are very realistic to the case of Sao Mai JSC. Through given survey questions, managers will deeply understand about plans and expectations of employees. That information will provide a comprehensive picture about the situation of employees’ loyalty in Sao Mai JSC. Sensitivity is the tool to count the variability of a concept (Zikmund et al., 2013). A research has a high level of sensitivity if variables are widely placed and collected data present different ranks of respondents. To measure sensitivity if this research, more questions should be added to get more detailed information and an issue will be viewed from many aspects. Ethical dilemmas of research processEthical dilemmas in research are critical concerns to both researchers and research clients to avoid potential ethical disputes. Researchers will have to strictly follow legal regulations and laws such as human rights or property rights; therefore, respondents will not raise any lawsuits. Plus, following ethical dilemmas will help researchers reach the targets of researches. In that way, personal information of respondents is secured and their trusts about the aims of researches are established. Respondents feel reliable to provide truthful information and errors of researches can be minimized (Resnik, 2011). In addition, existing laws are not sufficient, it is necessary to apply ethics when dealing with humans.To secure confidentiality, information which has been provided by employees will not be freely provided to the third parties without pre-approval of clients and research participants. To ensure anonymity, research participants do not need to provide their names to survey sheets. Besides, to reduce conflicts of interest among research clients, researchers, and research respondents, research aims, plans, and process must be initially approved by decision makers so that research implementation will not face difficulty of money or time (Zikmund et al., 2013). Researchers should carry out their works at the expected levels of research clients. Researchers also must be honest to provide information and errors. Plus, survey correspondents must be informed about the aims of researches, time, and place which they will be invited for surveys. Respondents have the rights of not answering any question if they are not willing and they can withdraw from the surveys at any time.
đang được dịch, vui lòng đợi..