Academic+Correctness

Academic Correctness (DQE Prep) (GWU EMSE-271)
Index | Topics (Logical Lectures) | Lectures | Problems | Readings | Nomenclature | Concepts

Top Concepts | Fill in the blank calculations | Academic Correctness | Notes | Other Oohs and Aahs | Test Question Bank | Problem-Solving Oriented Checksheets | Q for VD | Definititions (TBD) | Symbols (TBD) | Potential Question Analysis (Mulitvariate Analysis) | Classical Statistical Inference | 2011 Study Plan | Problems Index ]

Although it is not how I thought about explaining statisitcal concepts at the start, it is clear we need to express things with doctor's degree academic correctness for the DQE, topics identifed include:
 * Realization- TBR - needs review for DQE completeness; we need Van Dorp's words
 * [See also EMSE 208 Realization, mostly [|Wikipedia Realization]]
 * Hypothesis Test conclusions [ **Someone should check this page carefully)**
 * PCA: Loading Analysis Interpretation; Dependency Loading - example in problem - TBR
 * Model adequacy and model validation? How does VD word it when being Ph.D ish.
 * Definition of a random variable
 * A random variable is really just an expression that relates a function from the total sample space to “real life”. – PhD Quals question!
 * Adjusted R-Squared "How much varaince in the residuals explain the sample variance." "How much varaince in the residuals explain the Sum of Squares."
 * Appropriate Interpretation
 * ** Normality (or other distribution) goodness-of-fit **.
 * From a Minitab (normal) probability plot. "All points fall within the confidence boundaries hence the normal distribution seems to be a good fit. In addition, the p-value of the Anderson-Darling test equals 0.20, hence we fail to reject the normality hypothesis test a significance levels 1%, 5% and 10%." - 2005 Solution
 * Not tested or unable to test: "Normality of the data has to be tested, by example using probablity plots." - 2005 Solution
 * **Confidence Interval and Credibility Interval:**
 * "Interpretation: {the} 90% confidence interval is a random interval that has a 90% probability of capturing the mean mu. The calcuated confidence interval (above) is a realization of such an interval and it either contains the value mu or it does not." - 2005 Solution
 * "A 90% credibility interval for the (cadence) is also a random interval with a 90% random interval with a 90% probability of contain{ing} the value of X. The calcuated interval (above) is a realization of that interval and Pr(X member of [LB,UB]) is approximately 90%." - 2005 Solution
 * Two more quotations from Prof van Dorp regarding Confidence Intervals (Fall 2009) - DQE explanation of a Conf Int:
 * "the random interval capturs the true mean with (1-alpha)% confidence.
 * "There is a (1-alpha)% chance (probability) that the random interval captures the true mean."
 * The key is to remember that the Confidence Interval is a random variable and changes width and can shift when the number of data points sampled changes.
 * The Credibility Interval is a fixed interval that does not change with then number of data points sampled changes.
 * Confidence Intervals Versus Credibility Intervals (with / Graphics)
 * Hypothesis Testing: Q4g. Is this an academic wording or another point that needs to be understood.
 * Fail to reject the null hypothsis (not accept)

Sources:
 * EMSE 271, Fall 2009

Contributors: Sisson, Schrader