Sample Size Determination - 05-19-2009 by James Lani - Statistics Solutions - http://www.statisticssolutions.com
Sample Size Determination https://www.statisticssolutions.com/sample-size-5/ Sample size has been regarded as a study plan which can influence and control the recognition of important distinctions, relationships, or dealings. Gathering samples of appropriate sample size representing the population or other collectives is a regular goal for the researcher. In this method, the researcher determines the sample size by ignoring the sampling error. Sample size determination and the relation with the non-response bias are essential statistics. For a free consultation on sample size determination for your research project or dissertation, click here. While a researcher conducts a simple survey on any given product, the survey is most likely to uncover a large number of errors. Thus, it is important for the researcher to check his approach by making a suitable sample size selection. The common technique of sample size determination for simple and random samples profits most researchers through real life documents that illustrate the techniques. These documents or real life manuscripts consist of sample size issues that have been determined to solve certain drawbacks. Cochran (1977) has given a modus operandi for sample size determination. In order to decide upon the sample size, according to Cochran, the researcher has to be able to make out the boundaries of mistakes and errors in the items which have been considered crucial in the survey. Cochran holds that an approximate guess of the required sample size is made disjointedly for each item in the survey. The researcher who is undertaking the task will then use the help of a wide range of sample sizes which includes smaller sample sizes for dichotomous categorical variables. Sampling decisions should be made by the researcher based on the data acquired. The researcher uses the largest sample size if the range of the sample size is close to the variable of interest. When the researcher does not have direct influence over the variance, he must take in the variance estimates. This is a serious component in sample size determination. This is because the estimation or approximation of the difference in the important variables of interest under the study is an essential module for sample size determination. To estimate the population for sample size determination, Cochran followed four steps. In the first step of estimating the population variances and differences for sample size determination, the researcher obtains the samples in two steps. He uses the results of the first step in order to settle on the desired number of extra responses to achieve an appropriate sample size based on the differences studied in the first step. Secondly, while determining the sample size, the researcher estimates the population variances for sample size determination by using the results of the pilot study. Next, the data from prior studies of the population is used by the researcher to determine the sample size. Finally, the researcher makes the required estimation for sample size determination by the formation of the population using the assistance of some logical mathematical results. Another developed mode of determining the sample size for the categorical type of data is that of Krejcie and Morgan’s (1970). For the determination of sample size, these formulas provide identical sample
1/2
Sample Size Determination - 05-19-2009 by James Lani - Statistics Solutions - http://www.statisticssolutions.com
sizes in instances where the researcher modified the charted or tabulated value established on the size of the population which should be below or equivalent to 120. The researcher should, however, take care while using these formulas for the sample size selection. While these are the two important and more popular formulas amongst many others in sample size determination, the researcher always has to be cautious with the process of determining the sample size. _______________________________________________
2/2 Powered by TCPDF (www.tcpdf.org)
