Weekly Question Lists: You must submit two questions each week on the class web page in the indicated locations.  One question will cover specific details of one or the other (or both) lectures.  The second question will be a big picture question, concerning the relationship of the material in the current section to other material previously covered, the relationship to material covered in other classes (eg, other research methods classes), or the relationship to material read form other sources, including online sources or sources I have made available on the reserve desk.

Specific Question:  Ask about specific mathematical formulas, specific interpretations, specific software implementations, or specific examples discussed in one or the other (or both) lectures.  The questions should be detailed and technical and exhibit engagement with the course material. 

Big Picture Question:  Ask about how the ideas discussed in one or the other (or both) lectures fit with other material we discussed earlier in this class, or about related material in other classes, or about material found through internet searches.  The questions can be technical or philosophical.  This is a good time to ask questions about technical material that you may still not understand from previous material in the class, or from other classes, as long as you make the connection between that material and the material from the last two lectures.  Make sure that your question can be understood by everyone: if you discuss material from another class, or from an external source, make sure you provide enough context so that the question is self-contained.

Submit the questions to the web page sometime after Thursday's class and before Friday, midnight.  Set your electronic calendars!  30% is a lot of grade!  Late questions will be graded down, depending upon how late. 

I will answer all questions.  

Question Grading:

Specific Question: (a) Clarity (the question should be well-stated) (b)  Specificity (the question should not be too vague), (c) Relevance (the question must concern the specific material of last two classes), and (d) Engagement (the question should show clear engagement with the lecture material; evidence that you have wrestled with the concepts presented and desire clarification).  Students typically lose the most points on engagement.

Big Picture Question:  (a) Clarity (the question should be well-stated) (b)  Self-containment (the question should be understandable to all readers, regardless of their backgrounds), (c) Integration (the question must integrate material from the last two classes with the other material), and (d) Engagement (the question should show evidence that you have wrestled with the concepts presented and desire clarification).  Again, students typically lose the most points on engagement.

At the bottom of my answers, you will see four numbers, like 95 100 100 90, meaning that those are the grades given for each of (a), (b), (c), and (d).   See http://merlin.tltc.ttu.edu/forums/showthread.php?t=1054 (password: 1bridge34) for examples of questions and answers, and how I graded them, in my regression analysis class.  But note that the bulletin board system is different this semester.

When deciding a question to ask, think to yourself, "What material am I being held responsible for learning?"  This question should help to focus your attention on questions that are specifically relevant.  See old exams on the website for examples of such material.

Late questions will be marked down by a percentage depending on how late.  You must send me an email telling me that you have submitted it late; otherwise I won't even look, and you'll get 0 credit.

Notes:  To preserve  anonymity, use your student ID.  Put the ID in the "From" box along with brief, but descriptive "Subject" headers.  Choose the subject headers last, after you have written your question; that way others who read the questions and responses will have a better idea what the question is about.

For example:

Subject

A5347628 I.  What is the correct formula and interpretation for a standard deviation?  II. Is higher standard deviation better in research?

Change "A5347628" to your specific student "Personal Test Number."  You have access to this though your eraider account: Personal test numbers are under a student's Techsis account, where they check grades and sign up for classes, under the "Student Records" heading in the left-hand column.   The "I." is the subject header for the "specific" question, and the "II." is the subject header for the "general" question.

Examples of a "specific" question:

Bad Question:  "What is a standard deviation?"

Better Question: "I thought a standard deviation was s/sqrt(n), where s=(1/(n-1))Sum(Xi-Xbar)**2. As I understand it, standard deviation is the margin of error for the confidence interval for the mean, in other words, the true mean lies within plus or minus two standard deviations of the sample mean.  Johnson and Wichern's text, shows this formula in Section 5.4. However, in class on Tuesday, you presented a formula without the sqrt(n) in the denominator. Is my understanding incorrect, or are there different ways that one can calculate standard deviation, depending upon the circumstances?" 

The "better question" exhibits a confusion between standard deviation and standard error, but that is fine.  I expect that you should be confused about many things.  It's natural.  Even your instructor is confused about many things!  So there is no stupid question, and I will definitely NOT grade on how "intelligent".  The point is to LEARN. You LEARN by asking questions! So ask about whatever is not clear, especially if you perceive that the concept is important.   A good, honest question is worth 100%, even if others might think it is too "easy."   As long as you provide enough context, I don't care if the question is easy.  In other words, you should write a fair amount to explain exactly where your confusion lies.  Again, most of the points marked down on the question grading occur because of lack of engagement.

The student's description of standard error (the one with the sqrt(n) in the denominator) is correct, but the interpretation of the standard deviation involves the actual data, not the sample mean.  Specifically, roughly 95% of the actual data lie within plus or minus two standard deviations of the sample mean.  I would explain all of this in my answer.

Example of a good "general" question:

In one of my research methods classes, we discussed a paper that used regression methods.  One of the independent variables was discarded because, according to the researcher, the "standard deviation was too small." I thought that small standard deviation was good!  Greater standard deviation, more accuracy, right?  The confidence intervals are narrower with a smaller standard deviation, and we like narrower confidence intervals, right? And since the standard deviation is in the denominator of the t -statistic, the t-statistic is larger when the standard deviation is smaller, meaning results are generally more significant.  This is all good, right?  So why did the researchers state that the variable was discarded because of low standard deviation?  It seems backwards! 

(As far as the answer to the question goes, if you have had regression analysis you should know the answer:  higher standard deviation is good for predictor variables, not good for the response variable.  This can be understood by examining the formula for the standard deviation of the estimator of the slope coefficient.)

If everyone asks similar questions over lecture notes, then I have a sure indication that more time is needed on a particular concept. So please ask honest questions - I will not be critical of any honest questions. 

All questions and answers will be required reading - many questions on the exams will be taken directly from them.