Jay S. Mendell, Ph. D. 
Professor of Public Administration, College of Architecture & Urban and Public Affairs
Florida Atlantic University, 2912 College Avenue, Liberal Arts Building 470, Davie, FL 33314.
Home phone, 954.755-8928. Cell phone, 954.895-6364
mendelljay@gmail.com

PAD 4704, Research Methods for Public Management
Sequence Number 13495, LA 303D Monday 07:10PM-10:00PM Davie 

Provides a basic overview and application of quantitative decision-making methods for public management. It includes research design, some major management techniques and computer applications.

Posted 2006-01-15-- Please, if you send me e-mail, include in your subject line, "I am a student in your PAD class." It is "PAD" that stops my spam filter from trashing your e-mail.

Here is guidance in opening an account to log on with your own username and password (http://www.ecs.fau.edu/labs/about/easydirections.htm).

Posted 2006-01-18--Let me explain why, at the start of every evening, I call attendance and write on the board a "shun list," the names of everyone who fails to answer to roll call.

There will be a midterm exam and a final exam. In the case of each exam, I will give very strong guidance in advance as to the questions, and I will allow and encourage students to work in groups of three or four to prep for the exams.

Here is the significance of  the "shun list." If someone appears often on the shun list  (1) they might not turn up for your group meeting, ort they might not do the work they promise to do for the group; (2) there is a high likelihood that they will not be able to answer the exam questions, so they may be useless to your group (or worse); and (3) in grading quizzes, I will be especially critical of work turned in by groups who contain "shun-worthy: students. 

Posted 2006-01-22 at 10:14 a. m. 
Here is a picture of a bell curve.

Posted 2006-01-24 at 10:00

Here is a HW due in class on 2006-01-30. Please hand it in printed, except for your name in the upper right of each page, where you may write.

There are two questions.

(1) Let's see if you were paying attention in class.

I mentioned that we took a satisfaction survey of our MPA
student but did not survey students who chose not
to enroll at FAU. Write two reasons we did not do 
a survey of non-FAU students. Two sentences will suffice.

(2) We choose to give multivitamins to lab rodents
and then test to see if MV improve their ability
to navigate a maze. Explain why you think this is a one-tailed or two-tailed experiment. Two sentence should
suffice.

Posted 2006-02-01 at 4:00 a. m.

A graduate student wishes to infer or determine the mean age of all 100,000 members of the North American Suntan Society. The society offers to sell all 100,000 names and addresses for $20,000, or all names in Florida for $3000, or any randomly selected sample for $0.30 each.

One of these purchases is likely to bias the results and make it likely that any estimate of mean salary will be inaccurate. Which purchase would that be, and why?

(Don't bother looking for the NASS on the Intertnet, as I made it up. It is an imaginary group of men and women devoted to developing their suntans.)  

Mr. Montgomery has pointed out that in one place I ask about mean salary and in another about mean age. I really makes no difference if you use age, salary, or both, since the answer comes out the same. Feel free to call me when you think I have made a mistake in posting a question.

Posted 2005-02-07 at 6:45 p. m.

This is HW, due no later than 2006-02-20. See additional HW for that date, posted below.

  Go to http://www.socialresearchmethods.net/kb/sampstat.htm and scroll down to the diagram of the bell-shaped curve where it says in a green font that “About 99% of the cases . . .”

 Notice that the curve is symmetrical and centered over 3.75.

 Before you go on, ask yourself if you understand points (1), (2), (3), and (4) below. You don’t have to write anything; just make sure you know why the following are true. Scratch out the numbers on paper to reinforce your understanding,  if that helps.

 (1) If there is a 95% probability that a given sample from the population will fall inside the range 3.25 to 4.25 there is a 5% probability it falls outside the range. Hint: 100% minus 95% leaves 5%.

 (2) Another way to describe the range 3.25 to 4.25 is to say “3.75 plus or minus 0.50.” Hint: 3.25 is 3.75minus 0.50, and 4.25 is 3.75 plus 0.50.

 (3) To increase the probability from 95% to 99%, you have to expand the range to 3.0 to 4.5, which is the same as 3.75 plus or minus 0.75. Hint, read what is says in green font. Do the addition and subtraction.

 (4) A way to restate (3) is “To increase the probability from 95% to 99%, you have to expand the range by a factor of 1.5.”

Hint: That’s 0.75/.50.

 Now, and this is very important, that’s a general rule: To increase the probability from 95% to 99%, you have to expand the range by a factor of 1.5.

 Now, consider the following scenario. You are considering making a run for the county commission, so you hire a political pollster to discover if voters might support your election. The pollster reports that 0.56 of the public say they would vote for you. (Don’t worry about issues of bias, since the pollster knows polling backwards and forwards. Just focus on sampling error.)

 “Wait a second,” you say, “what is the range of uncertainty?”

 “Relax,” she says. “Your numbers are a .55 proportion of approval, plus or minus 0.04. In other words, you are a likely winner, because the low end of that range is 0.51 approval and the upper end is .59 approval. So you should win somewhere between a landslide at .59 or a squeaker at .51. But either the low or the high number or anything in between will elect you, since you will have over half of the vote.” She adds, “And that’s with a 95% probability.

 But you want to know what the numbers are at 99%. Calculate the range at the 99% level and tell me if it exceeds .50 at the low end.

 Posted 2005-02-07 at 6:45 p. m.

 Our final exam will be on Monday, 24 April 2006.

Posted 2006-02-14. 

HW due 2006-02-20.

You are the general manager of a TV station, and you are going to interview candidates for the jobs of (1) weatherperson and (2) sportscaster. I need you to write five questions you might ask each candidate in an interview.

Here are the ground rules.

(1)   Ask no generic questions such as “What do you want to do in your life?” or “Are you a people person?”

(2)   Ask the prospective weatherperson questions that clearly pertain to being a TV weatherperson, but not anything to do with reporting sports.

(3)   Ask the prospective sports reporter questions that clearly pertain to being a TV sports reporter, but not anything to do with weather reporting.

Be sure to consult http://www.citylinkmagazine.com/2000people.html and http://www.bls.gov/oco/ocos088.htm.

Posted 2006-02-20.

The first of two quizzes will be held on Monday, 20 March 2006. That is the second class after spring break.  

You are allowed to prepare for the exam in study groups or to work solo. Study groups work well if you are careful to let only reliable people into the group.

Below is a panel of twelve questions. Six of the questions will appear on the exam exactly as I list them in the panel, and there will be two other questions that are variants of questions from the panel. That makes an exam of eight questions, which is, of course, impossible o complete if you have not written your answers in advance for the twelve panel questions.

Let's say that you are part of a study group of four students. You can assign three questions to each member of the group or you can work as a group on all twelve panel questions. So long as you swap answers among the group, I will not call this plagiarism. Then you can word process four copies of each answer (one answer to a sheet, please, and give each member of the group a copy of each answer (which they should immediately label with their name).

Then each of the four of you (or however many are in the group) will come to the exam with answers to all twelve panel questions. There will be no study group collaboration during the exam. 

When I give you the exam, each of you will turn in the answers to the six questions that are identical on the exam and the panel, and then each of you can answer the two "variants."

This system works fine if everyone shows up and does what they agreed to. But it is a disaster for a team  that enlists an unreliable member.

Okay, here is your panel of twelve questions.

(1)  It is widely asserted that black people receive inferior treatment than white people in the health care system.

Suppose you want to formulate a research study. How would you make operational "inferior treatment"? Hint: one way would be to count the number of tests administered per patient. But what other measures (nominate three) might you use of quality of care?

(2) The Republican party wants to take a sample of political attitudes of a wide range of people in Deerfield Beach, Florida. They can afford to sample 1000 persons. They approach you as a consultant to see if you want to conduct a survey of 1000 residents of Century Village retirement community in Deerfield Beach.
 
What objections you would have to selecting randomly only in Century Village..

(3) A county government ended up defendingf itself in federal court because it administered an entrance exam for police academy to applicants for the fire academy. What is wrong with this? .

(4) You are conducting a face-to-face interview of a research subject in her office. The course textbook by Trochim advises you to "Be alert for any sign that the respondent is uncomfortable."

What would be three signs that the respondent is uncomfortable, and how would you accommodate to the signs? The answer is not in the text: use your common sense or ask a friend or colleague.

(5)  You have been invited by student government to sit in on two sessions of a course, to provide evaluation of how students "react to the professor personally." There are 300 students in the course, which is conducted in a huge lecture hall, so assume that your attendance will be unobtrusive, unless you draw attention to yourself.

What steps will you take so that the professor is not aware of your presence? What steps will you take to make yourself unobtrusive to the other students?

Remember that you are not there to evaluate the professor's teaching effectiveness, but rather to provide student government some measure or measures of the professor's rapport with students. What would you observe, and what would you report?  

(6)  Universities may make changes in their programs, to subvert  measurement of their performance by the legislature.

Select three from the following list and tell me how a university might noodle its way around the indicators by making changes that do not necessarily improve the performance that was intended to be measured.

(a) The retention rate of students.
(b) The proportion of lower division instructional courses taught by fulltime faculty with Ph. D. degrees.
(c) The minimum number of hours per semester required to be spent by faculty in student advisement.
(d) The proportion of graduate and undergraduate students participating in sponsored research programs. 
(e) Placement data on graduates.
(f) The proportional changes in the participation and graduation rates of students from groups historically underrepresented in higher education.
(g) The proportion of graduate students who received undergraduate degrees (A) at the institution, (B) within the state, (C) within the United States, and (D) from other nations.
(g) The number of full-time students who have transferred from a community college.
(i) Demonstrable evidence of improvements in student knowledge, capacities, and skills between entrance and graduation, where this evidence exists.
(j) Results of surveys of students regarding student attitudes and experiences, where the surveys exist."