CRM/LAW C167 Lecture Notes - Lecture 1: Incipit, Level Of Measurement, Anthology
CRIM 167: CRIME MEASUREMENT
WEEK 1: Measurement Principles
Why do we measure crime?
• Comparative research
o New York v. Los Angeles
o 1980s vs. 1990s
• Etiological Research (etiology refers to
causality)
o Causes of delinquency
o Offender rehabilitation
• Public safety
o Evidence based policy
Why is crime so hard to measure?
• Crimes are rare events
o In 2000, Irvine had:
▪ 50 robberies
▪ 143,072 residents
▪ covered 65 sq. mi.
• Crime are heterogeneous events
o Personal vs. property
o Commercial vs. individual
o Instrumental vs. irrational
• Crimes are arbitrarily defined events
o White color crime
o Vice crime
o Class “warfare”
• Crimes are “squishy” events
o Time, place, seriousness
• Victims don’t report crimes
o Embarrassment
o Time and effort
• Reports are unreliable and/or biased
o Victims
▪ Lie, forget, cover up, make
mistakes
• Offenders don’t co-operate
o Offenders
▪ Exaggerate
▪ Minimize their culpability
▪ Stand moot
How do we measure crime?
• Crimes are known to the police
o Uniform crime reports (UCR)
o 911 calls for service
o arrests
• Crimes not known to the police
o National crime survey
o 911 calls for service
o offender self-reports
• Other archives…
Principle of Observation
Two Dimensions of Observation
• Observations can be:
o Natural or experimental
o Direct or indirect
Direct vs. Indirect
• Archeologists reconstruct the culture of an
ancient civilization from graves and garbage
dumps
• Climatologists reconstruct the historical
atmosphere from glaciers cores and tree rings
• Pathologists reconstruct the course of
neurological disease from scares in post-
mortem brain tissues
• Paleontologists reconstruct the decent of
mammals from dinosaurs by reading the fossil
• The measure the physical dimensions of
women in an African village, an anthologist
(Francis Galton) sets up a surveyor’s transit on
a hilltop outside the village. Women who walk
the past line-of-sights are measured without
their knowledge—or permission
• To study the radio listening habits of
customers, a Chicago auto dealer has its
mechanics record the preset stations of the
radios in autos brought in for service
• To observe clandestine liquor consumption
during prohibition, a sociologist counts the
empty bottles in trash hauled away from
homes.
Principles of Measurement
Definition
• By measurement, we refer to the general
process through which numbers are assigned to
objects in such a fashion that is also understood
just what kinds of mathematical operations can
legitimately be used, given the nature of the
physical operations that have been used to
justify or rationalize this assignment of
numbers to objects (Blalock, 1982)
• Incipit numerare, incipit errare
(Morgenstern 1950)
find more resources at oneclass.com
find more resources at oneclass.com
Modern theories of measurement
• Hempel equates measurement with 3 every day
activities:
o Sorting one pile of “same” and
“different” objects in 2 piles
o Comparing 2 objects to see which is
“greater” and which is “lesser”
o Counting “how many” objects there are
in each of 2 piles
“Sorting” as Measurement
• At its most basic level, measurement consists
of sorting N objects into k categories in which a
way that objects in the same category are
“similar” and objects in different categories are
“different”
• N = 142 people are assigned to either of k=2
sexes, M or F
• N = 122 tests are assigned to one of k=5
grades, A, B, C, D, and F
• N = 65 objects are assigned to k=65 weight
categories
Nominal Scale
• A scale is nominal if it has _____ mutually
exclusive, exhaustive categories
o Sex = {Male, Female}
o Race = {Black, White, Other}
o Religion = {Buddhist, Christian, Jew,
Muslim, Other}
• Arithmetic is not possible with a nominal scale.
The sole purpose of a nominal scale is to
“compare and contrast.” For that reason,
nominal scales with many categories are not
very useful
“Ranking” as Measurement
• At a slightly level, measurement consists of
ranking N objects that the jth object is never
larger than the kth object. If the jth and kth
objects are the same size, they are “tied”
in rank. Otherwise, the kth object
outranks the jth object
o N people are ranked by
socioeconomic status
o N substances are ranked by
“hardness”
o N questionnaire responses are
ranked by social desirability
Ordinal Scale
• An ordinal scale is a nominal scale whose K> 2
categories are ordered
• Grade = {A, B, C, D, F}
• Year = {Freshman, Soph., Junior, Senior}
• Rank = {Private, Sergeant, Lieutenant, etc.}
• Ordering the K>2 categories of an ordered
nominal—i.e., ordinal—scale amounts to
sorting the categories to reflect he ordered
relationship.
“Counting” as Measurement
• At the highest level, measurement consists of
counting the number of units of some property
that each of N object possesses
• The Nth person weighs 58.6 kg
• The volume of the Nth object is 164.21cc
• The income of the Nth household is $21,098,54
Interval Scale
• An interval scale is an ordinal scale whose K>2
categories are categories are separated by
exactly one unit. For the ith object
• Weight = 0 kg. < Weighti < 1000 kg.
• Income = $0 < Incomei < $1,000,000
• IQ = 0 < IQi < 250
find more resources at oneclass.com
find more resources at oneclass.com
WEEK 2: Crime Known to the Police
Theories of Measurement
• Classical theories of measurement were
designed to explain why individual cases of the
same thing “look different.” They were
ontological theories
• Rationalist theories of measurement were
exemplified by Plato’s theory of ideals
• Empiricist theories of measurement were
exemplified by Aristotle’s theory of essences
Classic Theories of Measurement: Rationalist
• Plato’s Theory of ideas
Classic Theories of Measurement: Empiricist
• Aristotle’s Theory of Essences
Examples from the Disciplines
Two Types of Measurement Error
Random Errors
Constant Errors
• Noise
• Good luck
• Bad luck
• Biases
• Method errors
• Idols
Random errors tend to
cancel out in the long run
Constant errors tend to
add up in the long run
UCR History
• 1927: IACP Committee on UCR
o Murder, Rape, Robbery, Aggravated
Assault, Burglary, Larceny/Theft, Auto
Theft
o Prevalent, Serious, Distinguishable,
Reported
• 1958: FBI Crime Index
• 1978: Arson
• 1985: Blueprint for the Future (NIBRS)
• 1990. 1994: Special Populations
• 2004: Personal Property Indices
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Why do we measure crime: comparative research, new york v. los angeles, 1980s vs. 1990s, etiological research (etiology refers to causality, causes of delinquency, offender rehabilitation, public safety, evidence based policy. How do we measure crime: crimes are known to the police, uniform crime reports (ucr, 911 calls for service, arrests, crimes not known to the police, national crime survey, 911 calls for service, offender self-reports, other archives . Two dimensions of observation: observations can be, natural or experimental, direct or indirect. Incipit numerare, incipit errare (morgenstern 1950: at a slightly level, measurement consists of ranking n objects that the jth object is never larger than the kth object. If the jth and kth objects are the same size, they are tied in rank. Otherwise, the kth object outranks the jth object: n people are ranked by socioeconomic status, n substances are ranked by. Hardness : n questionnaire responses are ranked by social desirability.