1A. I would like to ask you to expand your response on how would you use inferential statistics in a practical situation. Generally speaking, in what type of situation would inferential statistics be more useful than descriptive statistics? 

Before you answer this question, there is a good article that summarizes two popular, easy to read stat books.

 

“Statistics” by Mark J Lauer , Performance Improvement, ISSN 1090-8811, 04/2005, Volume 44, Issue 4, p. 48.

 

1B.  I agree that car companies could use survey research as part of their product development cycle,  however, when I read your response, I do not see a strong correlation with use of mean and standard deviation for supporting business decisions.  Please explain how would you use mean and standard deviation to “to compile what the consumer wants”. 

Also, I am not sure how you propose to use inferential statistics in a practical situation. Generally speaking, in what type of situation would inferential statistics be more useful than descriptive statistics? 

Before you answer this question, There is a good article that summarizes two popular, easy to read stat books.

“Statistics” by Mark J Lauer , Performance Improvement, ISSN 1090-8811, 04/2005, Volume 44, Issue 4, p. 48. 

 

1C.  I agree that car companies could use survey research as part of their product development cycle,  however, when I read your response, I do not see a strong correlation with use of mean and standard deviation for supporting business decisions.  Please explain how would you use mean and standard deviation to “to compile what the consumer wants”. 

Also, I am not sure how you propose to use inferential statistics in a practical situation. Generally speaking, in what type of situation would inferential statistics be more useful than descriptive statistics? 

Before you answer this question, There is a good article that summarizes two popular, easy to read stat books.

“Statistics” by Mark J Lauer , Performance Improvement, ISSN 1090-8811, 04/2005, Volume 44, Issue 4, p. 48. 

 

1E.  You stated: “Inferential statistics would look at their past performance to anticipate their potential performance”.  I do not believe that inferential stats can be used to predict future (forecast future performance). Regression models could be used for that purpose.

Generally speaking, what are some of the things that we can do with inferential statistics that could not be done by descriptive stats alone?

 

2A. I agree with your approach (plot a histogram of your data) set.  A graphical presentation is the easiest and quickest way to identify the distribution of your data set. Can you provide a specific example on how would knowing the distribution shape impact the analysis of your data set? We certainly would like if our empirical distribution would be normal, but what would you do if you would determine that your distribution is too skewed to be considered normal? How would you analyze such a data set?

 

2B. I am not sure what you meant by “… we can make the appropriate business decision to make the data more symmetrical”? Even though researchers can manipulate data, why would they try to make them more symmetrical? Please clarify.

We certainly would like if our empirical distribution would be normal, but what would you do if you would determine that your distribution is too skewed to be considered normal? How would you analyze such a data set?

 

3.  I agree that mean/median are not suitable for qualitative data. Is there any measure of central tendency that can be used for all four levels of measurement (nominal, ordinal, interval and ratio)?