Evaluating F Tests

EVALUATING F TESTS 3

EvaluatingF Tests

Inthe analysis of data, F-test and t-test are usually consideredhowever, the two tools of analysis are utilized under differentcircumstances. From the hypothetical data provided, it is apparentthat some of the results became analyzed using the t-test whileothers considered the application of the F-test. This report willdiscuss why there was a difference in the use of these two tools.

Thet-test is used in the estimation of population parameter that is,population mean. Also, it is utilized for hypothesis testing of thepopulation mean, where the standard deviation of the population isnot known (Keller, 2012). In the hypothetical data provided, thet-test was applied to test a hypothesis where the standard deviationof the population was not presented. Furthermore, a t-test is usedwhen one desires to compare only two sample distributions. In thecase under consideration, the t-test became applied since there was aneed to contrast two means.

Alternatively,the F-test can be used where there is a need of finding out whetherthere exists any variance within samples (LeBlanc, 2004). This is thecase since F-test is considered to be the ratio of the variance oftwo samples. Unlike in the t-test, where only two sampledistributions are compared, the F-test can be used in instances wherethere are more than two sample distributions being taken into account(Timm, 2002). In the hypothetical data, there was a scenario whereone independent variable had more than two levels, or means was beinganalyzed. This scrutiny necessitated the use of the F-test.Therefore, in the hypothetical data provided, there was a need toapply the F-test tool since the application of multiple t-tests couldhave resulted in consuming a lot of time and making errors.

References

Keller,G. (2012). Statisticsfor management and economics.Mason, OH: South-Western Cengage Learning.

LeBlanc,D. C. (2004). Statistics:Concepts and applications for science.Boston: Jones and Bartlett.

Timm,N. H. (2002). Appliedmultivariate analysis.New York: Springer.