AllFreePapers.com - All Free Papers and Essays for All Students
Search

Business Statistics

Autor:   •  March 8, 2011  •  Essay  •  466 Words (2 Pages)  •  1,936 Views

Page 1 of 2

From the research conducted it was found that the customers with a balance greater than $100,000 did not conduct their banking in the morning, rather, in the afternoon. This conclusion can be made by the use of the techniques, of descriptive analysis and hypothesis testing. In this example it was noted that a two tailed test was needed as the statistical hypothesis was directionless, meaning the alternate hypothesis can be either greater than or less than. That is,the null (Ho), cashed up = morning and alternate is (Ha) ≠ morning. In other words, the null hypothesis states there is no difference between means.

Through the analysis of the given data, the sample average balance for customers in the morning was $99,138 while the afternoon had an average of $101,616. Although afternoon customers had a higher average, the difference was minimal as it approximately equated to 2.44%. Furthermore the coefficient of variation of both morning and afternoon balances were similar with values of 57.82% and 56.98% respectively and differing by only 0.84%.In other words, the variability between both sets of data are virtually identical. Through a descriptive statistical point of view, it suggests that there are no significant differences between the mean of the morning and afternoon. However, this inference is not entirely concrete as the varying sample sizes have to be accounted for as the morning had an observation value of 477 while the afternoon was 483.

To confirm the inferred conclusion via descriptive statistics, the comparison of the mean through the technique of hypothesis testing can be utilized. This methodology considers the different sample sizes meaning that the conclusion can be deduced at a more quantified level.

That is, standard error is conducted to determine how accurate the sample mean is compared to the actual population mean. The calculation of the standard error will act as a precursor to determine the confidence interval which

...

Download as:   txt (2.8 Kb)   pdf (83 Kb)   docx (10.7 Kb)  
Continue for 1 more page »