Chemistry Sampling
Using GPower3, the sample size was 100. This sample size is viable as it meets the threshold of normality. The sample size is greater than 30 hence the justification that it is normally distributed.
T-test is to be used for testing the difference in means of the two groups. The study should be designed in a manner whereby there is intrinsic pairing of the corresponding items in one sample with those of the other sample. This will test the mean of the two groups effectively. One of the samples would act as the control while the others act as the treatment.
The aim of the testing will be determining whether the populations from which the samples were drawn are significantly different. A descriptive design is to be used as two samples are drawn from a similar population.
In this test, the assumption was that the data was normally distributed. The type one error, alpha level was set at 0.05 levels. The statistical power level for calculating of the sample size is to be set at 0.8. The effect size in the test is set at 1. The sample size was equal for business unit one together with business unit 2. The sample size in the two samples was 200 with each sample having 100.
Sampling comes with its costs; therefore, the research design should be designed in a manner that consumes the least cost. However, the data integrity and authenticity should not be compromised. Samples should be selected randomly from the business units without bias. The test is ethical as it has no bias; different potential subjects were accorded equal chances of being incorporated in the study (Liu, & Raudenbush, 2000).
Second post
With a margin error of 5%, confidence interval of 95% and a population size of 200, the recommended sample size is 132. This sample size arrived after an online calculation and it is justifiable by being more than 30 hence it meets the threshold of normality.
A random sample will be selected from the business units. The samples are selected concurrently from business unit one and business unit two. Selecting them alongside one another will ensure equal selection, as well as avoidance of bias. Simple random sample would be obtained by allocating random numbers to different units randomly. The units allotted with different numbers are then selected one after another and included in the sample survey. This design is efficient in including the units in the sample design.
It is assumed that sample one and two have a bivariate independent variable. The samples are continuous dependent variables. Each observation of the dependent variable has independence of the other observations of the dependent variable. The dependent variable is also assumed to have a normal distribution with similar variance and standard deviation. The alpha is assumed to be 0.05, with the statistical power assumed to be 0.8. Taking the standard deviation to 14 the sample size was found to be 125.
A controlled study will be reliable in this study. This is because the subjects are assigned randomly to a given treatment or another. This diminishes any inherent bias amongst the subjects. The calculations needed are on how many participants are required to determine any relations. The design is ethical in that it avoids any form of bias (Kirk, 1995).
References
Is this your assignment or some part of it?
We can do it for you! Click to Order!
