Sampling and Data Collection Plan
The sampling shall be done using the randomized sampling strategy. However, it is necessary to point out that technically, sampling will involve clustering, as Banana Inc. is a multinational supplier of electronic merchandize and the event that was the initial reason behind the research concerned mostly United States. In this way, the sample shall randomly be chosen exclusively from people living in the United States. The technique to be used in obtaining research data is going to be data mining. The size of sample to make valid assumptions on the regard of the research question is needed to be significant, exceeding 1000 respondents by the prior rough estimation.
For this research we shall first calculate a size of sample required to correctly represent a population of all possible respondents who leave feedback on company’s products. Basing on the retail study conducted by Bettencourt (2016), who used the sample of 27 000 respondents, this research shall take this number as one representing the population as it is impossible to find out its real size. In this way, according to calculations presented in Appendix 1, the sample needed for adequate representation of population is 1292.
As for the problems, concerning validity and reliability, the following measures shall be taken. The validity of research shall be preserved according to the requirements of external and internal validity. As for the external validity, the research question is formulated as “What the correlation between negative feedback is from surveys and dollars made?” this particular type of validity is going to be secured by transparency of research’s objective for potential participants. In this way, the concern about the discrepancies between the declared purpose of research and the measures taken as the main threat to external validity vanishes. As for the internal validity, the research is testing the hypothesis about the influence of company’s image on sales. To address the hypothesis, the adequate measures must be chosen. In this research, measures of choice are the quality of client’s feedback, evaluated as mean ranks instead of nominal variable, and the volume of money obtained during the studied period evaluated in US dollars, presenting it in an interval scale. Reliability shall be achieved by preserving the raw data obtained from respondents so any person could get access to them by request and estimate the quality of research design and calculation. In this way, the possibility to replicate the results as the main requirement of reliability is met. Human subjects shall be given a form of concern to sign prior to starting a survey, approved by the university’s commission on problems of research ethics.
Data collection plan shall include searching through the company’s client base and contacting those people with the request to take part in survey. After obtaining their consent, they will be asked to give some feedback on the company’s public image regarding the privacy issues related to using company’s products. General context of being either semantically positive or negative is going to be …