Another five days of data was collected to add to the first 10 days of data having a total of 15 days of observation. The following are the text thread data obtained for a total of 15 days. A larger sample is now present for the analysis of the text thread that I have for 15 days. A large sample is really important in statistics. As the sample become larger, the data will approach the data of the whole population. The data will become as close to the real data (Israel, 2009).
For example, the mean number of hours at work for the population of workers in United States is 8 hours. The sample mean of 50 workers 7. 5 hours. As the sample size increases, it is possible that the sample mean will become closer to the mean of the population. In this case, why should one use samples instead of the population? Or, why use large samples instead? Certain reasons are given to support the use of samples instead of very large samples or the population itself.
One of the common reasons is population are sometimes infinite. When one speaks of infinite population, it means that one cannot actually count the total number of the members of the population. Another reason is time constraints; one cannot collect data for a very large sample or the whole population because there are certain time limitations. After having 15 samples, the mean for the 15 text threads were computed. The mean obtained from the 15 sample is 8. 2 threads per day.
The mean for the 15 sample is less than the mean obtained for the data when there are only 10 observations included in the sample. The current sample of 15 observations is still insufficient in order to conclude something from the population. One still cannot determine whether the sample is already sufficient. Sufficient sample size can be determined through formula and different assessments such as the precision, confidence and variability a person wants on his sample.
Nevertheless, as long as it is not the population itself, any kind of samples will still have an uncertainty associated in it; the uncertainty associated with the sample is called the sampling error. The smaller the sampling error, the better the sample size one has obtained (US Census Bureau, n. d. ). Reference Israel, G. (2009). Determining sample size. Retrieved August 20, 2010 from http://edis. ifas. ufl. edu/pd006. US Census Bureau. (n. d. ). Things that may affect estimates from the American community survey. Retrieved August 20, 2010 from www. census. gov/acs/www/Downloads/ACS_Affect_Est. ppt.