CJ 160

Questions to accompany your reading of the Research Methods in Criminal Justice pages

If there is just an item on a line, it refers to a term you want to be able to define. Page sections are listed in the order assigned

DATE OF LAST UPDATE: 9/23/05

38-46 (ignore section on different causal logics 42-43)

theorizing
macro vs. micro vs. bridging theories: how are they different?
concepts or constructs
predictor concepts
outcome concepts
features (or attributes)
propositions
operational definitions (or - what is the process of operationalization?)
How do concepts get turned into variables in your head?
hypotheses (what are they and how are they different from propositions?)
variables (what are they? Before measurement? After measurement?)
How are "true scores" on a variable different from "observed scores"?

100-111

measurement error - key idea - what is it? where does it come from? does it mean we are doomed?
How are "true scores" on a variable different from "observed scores"?
How are attributes different from objects?
What do we measure when we measure (p. 104)
What does this mean: Observed score = true score + measurement error Can you see how this would apply to your scores on an exam? Can you see how it would apply when officers conduct field sobriety tests? Can you see how it would apply when stopped drivers blow into a breathalyzer?

Variables can be classified based on the level of measurement present in the variable. You may have covered levels of measurement in a statistics course. The basic idea is that "not all variables are created equal" - some give you more information than others, some give you different types of information than others. The main point is that you have an understanding of the measurement properties of typically used criminal justice indicators, which we will talk about.

What does it mean to say a nominal variable has categories that are exclusive and exhaustive?

the ordering property

77-94

What is the difference between categorical and continuous data? Can you give an example of each?
When you are looking at a bar chart, can you spot ways it might be incorrectly set up? I.e., do you know what makes for "good form" in a bar chart?
What kind of information do histograms convey?
What is "good form" for a line chart?
SKIP SECTION ON SCATTERPLOTS 85-90
What do frequency distributions tell you?
With tables, where does the independent variable go? The dependent variable?
What do the column percentages do?
P 92: YOU REALLY want to understand how to decode tables using column percentages and the how the "WHEREAS...." suggested phrasing for decoding a table works.
 

111-118

There is an equation (oh no!) on p. 112 - see below

We will be talking in class about how this applies to two specific instances: grades in a course like this one, and an index we are going to put together showing your confidence in the criminal justice system. More specifically, in each instance you want to be able to think about:

* what contributes to measurement error?
* what does specificity mean?
* what does communality mean?

Also, in each example, can you understand in practical terms how "items are noisy" and how the index permits MORE discrimination? In other words, can you understand what the purposes are of indexes?

On RATES:
How is a rate different from a count? Can you see the LOGIC of how a rate standardizes? We will be spending a lot of time with an in-class example.

There are two types of rates: incidence based, and prevalence based - you want to be clear about how they are different

With the example on p. 117, can you state in your own words what the problems are with the differential item weighting used in the Philadelphia Suburban Crime Safety Index?

183-206 - SAMPLING

MOST IMPORTANT: you want to understand the major differences between probability and non-probability samples - both in terms of how they are carried out, and in terms of what valid INFERENCES may or may not be drawn from the two types of samples.

What is a population?

What is a census?

What is a population parameter?

(Will NOT ask you about confidence intervals)

Normal curve:

 will have an M&M in-class example. Since 161 spends some time on this, our only purpose will be to demonstrate that when you take means of samples, those means start to form a normal curve. So you want to understand what is a sample mean.

KEY practical point to understand: we care about the normal curve because it tells us EXACTLY how closely our sample results will be to the REAL population value.

SAMPLING ERROR - key idea - you want to know WHERE does it come from? is it ALWAYS bad to have it? HOW can we insure that sampling error is UNBIASED rather than BIASED?

What are limits of non-probability samples - this is key

On VARIETIES of sampling types, you want to know HOW you do a simple random sample, WHAT is the SAMPLING FRAME, and based on an extensive in-class example (the suitcase example), how does a MULTISTAGE CLUSTER sample work.

What is a stratum?

What is a cluster?

235-263

This chapter on surveys is pretty straightforward. The main things you want to understand are:

  1. What are the advantages and disadvantages of each different type of survey MODE (in person, telephone, mail)
  2. What makes for BAD survey questions and response categories; what makes for GOOD ones?
  3. What is RDD? How does it work? WHY do people use it? What are its advantages?
  4. Who are contact persons? Who are designated respondents?
  5. It would be GOOD if you knew some of the main features of the National Crime Survey (now called the National Crime Victimization Survey.
  6. What do we know about the reliability and validity of surveys?

131-171 (note 172-178 cut off)

On reliability, you want to understand what BASIC QUESTIONS are being asked when we talk about reliability OVER ITEMS (internal consistence), OVER RATERS (inter-rater reliability) and OVER TIME (test-retest reliability).

On validity:

On external validity:

266-291

In what ways can we classify quasi-experimental and experimental designs? (267)

What is an experimental or treatment group?

What is a control group?

What is a treatment?

What does it mean to say a study's results have internal validity?

Why are PLAUSIBLE THREATS TO INTERNAL VALIDITY something we worry about?

You will need to understand how some simple quasi-experimental designs are set up: