Speaking Notes

PADM 5502

September 6, 2007

Dr. Neubauer

WHERE WE ARE: [THESE INSTRUCTIONS WERE AMENDED 9/17/09]

• The first installment of the research prototype is due Sept. 27. It should include an acceptable topic, a cover page, a first draft of an introduction, a first draft of a literature review (based on three peer-reviewed articles), and a research model. It can also include a section titled “Hypotheses and Rationales,” although that is not required as part of the first installment.
• Our midterm exam will be October 8. It will include questions related to all the things we are discussing in class. I cannot map the midterm to specific book chapters.

REVIEW OF MAJOR TERMS AND CONCEPTS SO FAR

• survey research
• independent variable, dependent variable
• causation, and spurious relationships
• research diagram
• (compound) dependent concept
• hypothesis, and rationale
• how to write a hypothesis as taught in class
• variable
• variable using Likert scale pattern of response
• IRB – human subjects review
• common motivations for research
• peer-reviewed journal
• literature review
• cumulative nature of scientific exploration
• why nothing is even "proven" in the social sciences
• APA style
• population, sampling frame, sample
• operationalization of a variable
• question with face validity
• reasons for turning the "polarity" of some questions
• primary and secondary research
• inferential statistics
• comparison of groups
• sample of convenience
• random sample from population
• common challenges when trying to actually interview a random sample
• sampling bias
• anonymous, confidential
• informed consent
• use of deception in the research design
• risk to participants
• usual demographic variables (age, gender, race, income, education)
• the difference between variables (such as gender) and values (such as female and male)
• Likert questions

AGAIN, A BIG PICTURE

When "doing" INFERENTIAL STATISTICS you are trying to understand things about a POPULATION based upon a SAMPLE of its members.

Something of interest in the population is called a PARAMENTER. Unless we do a CENSUS we will never exactly know the values of parameters in a population. A STATISTIC is a value derived from a SAMPLE of the POPULATION which is intended to be an estimate of the related parameter in the population.

You should not claim to be doing INFERENTIAL STATISTICS unless the sample is a SIMPLE RANDOM SAMPLE, which means that every member of the population has an EQUAL KNOWN CHANCE of being included in the sample. Simple random samples are really hard to get. To do one you must have a SAMPLING FRAME which is a list of every member of the population. You must then select from the sampling frame RANDOMLY. You must find all the people so selected. They must each agree (voluntarily) to participate.

A SAMPLE OF CONVENIENCE is not a random sample. You cannot really INFER from the STATISTICS derived from a sample of convenience to estimate PARAMETERS in the population.

CHAPTER 4 MATIERIAL

You have a variable in your RESEARCH MODEL. (Actually, you have many variables.)

In SURVEY RESEARCH you must write one or more QUESTIONS intended to OPERATIONALIZE each variable in your research model.

There are three (technically four) LEVELS OF MEASUREMENT.

Level of measurement is important because it determines the PRECISION of the data and what kinds of statistical analyses are appropriate.

The levels of measurement:

• nominal
• ordinal
• interval
• ratio (interval level of measurement with a natural zero)

NOMINAL VARIABLES

The values are simply names with no natural/numeric order.

Example: Religious affiliation --

Muslim

Jewish

Christian

Buddhist

Other

Example: Make of vehicle

Buick

Ford

Toyota

Other

ORDINAL VARIABLES

The values of this kind of variable has a natural order but NO STANDARD UNIT OF MEASUREMENT.

14) I like potato salad.

[ ] [ ] [ ] [ ] [ ]

Strongly Disagree Undecided Agree Strongly

Disagree Agree

18) My vehicle is . . .

[ ] a subcompact

[ ] compact

[ ] midsize

[ ] full size

[ ] a monster truck

INTERVAL LEVEL VARIABLES

Here, the answer (value) involves use of a standard unit of measure.

9) How tall are you in inches? ______________ inches

10) How many children do you have? _______________

11) What is your family's annual income before taxes? $___________________

-------------

Interval level date is the most PRECISE and you can do "better" statistics on it.

If you have interval level data you can "down grade it" to ordinal if needed.

If you have ordinal data you cannot "upgrade" it to interval. If you collect family income in ranges you cannot than figure out the exact values. If you have exact values you can create ranges.

THREE MAJOR MEASURES OF CENTRAL TENDENCY -- mean, median, mode

	Mean	Median	Mode
Nominal	no	no	yes
Ordinal	no	yes	yes
Interval	yes	yes	yes

Regarding the measures of DISPERSION, the same applies. Interval data is "better" than the others.