In: Math
Now that you have conducted a series of inferential statistics to validate your claim. Discuss any three (3) possible concerns in business or statistical aspects regarding the approach used here. Such that in retrospect, how would you have done this study differently again?
The three possible concerns regarding the approach would
be
1) If we didn't measure a population completely of the data
provided, we won't be sure about the correctness of the values and
the test statistics that we have calculated. Without a complete
data, there will always be an uncertainty about the estimates.
2) Some inferential tests may need an individual to make some random guesses to run the inferential tests which can create some uncertainty in the results of the inferential statistics
3) The p-values or the value of significance when conducting the inferential tests used can be a cause of concern. If an appropriate p-value isn't used, the test statistics can be influenced significantly causing us to misinterpret the data.
Keeping these problems in mind, a large sample would be a big help since through large sample size, we can easily approximate our estimates to a significant figure which would eliminate all the sampling errors in the population.
If we use an appropriate p-value for our population sample, the estimates would come generally closer to the actual values. The p-values can be important to test the significance of the H0 and hence, it would be important to use a p-value according to the sample size.