Test 1 Flashcards
Define Validity
The strength of our conclusions, inferences, or propositions.
Internal Validity
Are there factors that could affect the outcome of the experiment?
- Concern is flaws within the study itself. Was it carried out as designed?
- Does the IV (and not some other variable) really cause the change in the DV?
External Validity
Refers to our ability to generalize the results of our study to other persons, settings, or times.
Define Variable
A variable is a measure representation of an abstract construct.
-It can differ in magnitude:
- A measurable characteristic that varies
- Age, intelligence, income
Define Independent Variable
The variable being manipulated.
Define Dependent Variable
The variable affected by the change in the independent variable.
Define Control Variables
Other factors that could affect the outcome of the experiment.
- These must be held constant to have a fair test.
Define Control Group
The control group is a selected group that does not receive the intervention or treatment.
Control Variable
Control Variables are factors that could affect the outcome of the experiment so they have to remain constant across all groups.
What is Regression toward the mean?
States that if a variable is extreme the first time you measure it, it will be closer to the average the next time you measure it.
- It can result in wrongly concluding that an effect is due to a treatment when it is due to chance.
- Pick the worst places, treat and then see the affect. Be careful with conclusins
Randomised Control Trials (RCT) - Strengths
Those eligible are randomly assigned into one of two groups.
- The intervention group.
- The control group.
- Remove potential Bias:
*Every subject is as likely as any other to be assigned to the treatment or control group. - Creates balance in the groups on factors in the study, whether those factors are measured or not.
- This allows accurate analysis of the effect of an intervention.
Define Effect Size
Quantifying the difference between the two groups.
Provides a measure of the “size of the effect” from the intervention.
The effect size indicates the magnitude of the observed effect or relationship is due to chance.
What is Statistical Significance?
Tells you how likely a pattern in your data is due to chance.
*Indicates the likelihood that an event is due to chance
* .05 findings has a five percent chance of not being true
Not the only thing:
*Helps you understand how rare the results are
* Does not tell you if the experiment was well conducted or well controlled
Define Paradigm
A paradigm is a model for observation and understanding that shapes both what we see and how we understand it.
Explain Rarity of discoer/ blind spots.
The rarity of discover/ blind spots:
-Can create blind spots
* A failure to see one’s basic assumptions about how things operate.
* The result discoveries are rare because expectations cloud our vision.
Observations about paradigms (Barker).
- Paradigms are common
- Paradigms are useful
- A warning: Sometimes paradigms become the paradigm
- The people who create new paradigms are often outsiders
- Advantages of young/ outsiders
- They don’t “know what can’t be done”
- They do not understand the subtilities of old rules
- They have no investment in old rules
- No peep pressure to keep the old model
- Paradigm pioneers must be courageous
- You can choose to change paradigms
What measuring things get you
Track progress of efforts - judge success
* If you can’t measure it, you can’t improve it
Aline activities with goal
Garner support and credibility
What is the Impact Domain?
How might intended police effects on the environment be measured?
What is the Process Domain?
How might police know if they are doing their works as they should?
What is the Community Assessment Domain?
How might public assessment of police performance be monitored?
What is Organization Health Domain?
How might police department know if their employees are satisfied with their work?
What is the Community Context Domain?
How might police organizations monitor changes in the work environment that impede or promote their ability to achieve organizational goals?
Bratton - Impact on crime with?
Better management
Higher expectations
Better startegies
Public confidence matters - Information (Kelling)
Acts that do not result in arrest or documentation often are not counted but may be very good for police departments and community.
Similarities between public and private organizations
Some things are the same:
Entrusted with assets and responsibilities
Human resources issues
Differences between public and private organizations
- Source of money
- Citizens, clients, politicians…decide that there is a result that can be done by government and it is worth paying for
- Discretion
- Less discretion to define organizational goals and direction
- Authorizing authority - more drivers and more involved than shareholders
- Performance management
- Profit vs. measures of value other than revenue
- difficult to measure and show value in public sector
Goal/ idea of public value
A normative and practical guide for those who are in positions of executive authority in government
Public managers are to deploy publicly owned assets to create public value
* Purposefulness in management - Public value clarifies for managers what they are trying to produce/ are accountable for.
What are the 3 parts of the Strategic Triangle?
Public Value
Operational Capacity
Legitimacy Support
Strategic Triangle - Public Value?
What is the specific improvement expected from a plan or strategy
- Improved lives some how
- Values are fuzzy
Strategic Triangle - Legitimacy & Support?
Practically you need the support of people who can provide legitimacy and resources
This given by “authorizing authority”
Not enough that manager has an idea of public value, others must share this too
- Mayor, council, union, politically powerful, regulatory authorities…
Strategic Triangle - Operational Capacity?
The actual ability to create public value
What is the location and character of the operational capacity required to achieve my goals?
Not just the resources you control:
* Often need others
* Citizens, media, sister organizations…
Strategic Triangle - How it interacts
All 3 matter all the time
Each of the 3 parts of the triangle matter, they are interrelated and influence each other
Goal/Key is to find the fit between the organization and its environment
It is more than satisfying demands, its responding to them in a way that creates value
What is the strategic triangle model used for?
Designed to help public sector leaders develop strong, achievable value propositions
* Helps you diagnose your local environment
Leaders are inclined to limit their thinking to the “what do do” (Public value circle of the triangle)
What is mobilization?
Mobilization is the process by which a legal system acquires it’s cases.
* How the law is set in motion
Why is the mobilization of law important?
It is how the law is set in motion
Define Reactive mobilization
Reactive- a citizen sets the legal process in motion
* Calls for service
Define Proactive mobilization
Proactive - the state initiates legal action upon it’s own authority
How mobilization type impact upon - Legal Intelligence
The knowledge a legal system has about the law violations
* Reactive system has responsibility on the citizen
* So what they can’t see, ignore, mistake, or don’t think is
worth it.
Also, reactive system cases come in one at a time and this tends to hide the impact of macro-level forces of crime
* Can hide relations among and between cases
* Need to collect data and look for patterns
In a reactive system, citizens mobilize during a crime or after
* Prevent is not strong in a reactive system
How mobilization type impact upon - the Avaibility of Law
NOT the access of legal system to cases but now the access of citizens to the law
* The reactive system is an entrepreneurial model of law
* Each citizen will voluntarily and rationally pursue their own I
interest
The proactive system is more of a social welfare model
* What is the legal good for the citizenry being defined and imposed by government
Anti-mobilization norms
* Norms that discourage mobilizing social control against status equals
* These have a larger impact in a reactive system
The Organization of discretion
Legal decision-making allows the agent some margin of freedom or discretion
A reactive system puts a great deal of discretion in the hands of ordinary citizens
* Moral standards of citizens shape case inputs
* Citizens engage unknowingly in selective law enforcement
The Organization of discretion - Possibility of discrimination
Where a system is reactive or proactive does not determine if there is any possible discrimination
* The nature of mobilization shapes where the possible
discrimination may come from
A reactive system can allow average citizens to invoke the law according to their own prejudices
* “Karens,” napping in dorm, sitting in Starbucks, BBQ, Birdwatching
A proactive system can allow state actors to invoke the law according to their own prejudices
Discuss Legal Change
The law is dynamic
One of the things that can change the law is a moral change in the population
* A proactive system is not as easily adaptable to changes in
citizen morality
* A reactive system absorbs every change that comes about in
the population
* Changes in reactive mobilizations change what the police are
sent to
Another way the law can change is by planned action
* A reactive system may be harder to change in planned
fashion - more resistant to centrally directed change
* A proactive system is a more willing instrument of planned
change for it is under the authority of the planners themselves
Reporting in the Black community - Findings/ implications on CFS (Mobilization) Where & how long; violence levels
All neighborhoods: Total calls slight upward trend for all observations for observation period
No shift immediately after event
After publication: 911 calls declined significantly (Net controls)
The effect Diminished over time and returned to normal
Violent crime calls declined significantly after as well
AREAS
Declined lasted longer in Black areas
* Here decline was large & durable (over a year)
Estimated impact on CFS
* Estimated across all call types: 22,200 calls in the year the reporting of the incident. 56% decline occurred in Black areas
* CFS regarding traffic accidents increased: indication no larger force for general decline
Define Cognitive Biases
Cognitive Biases are systematic deviations from rationality in judgement or decision-making
Define Confirmation Bias
Confirmation Bias is the tendency to search for information that supports our beliefs and ignore or distort data contradicting them.
What is Fundamental Attribution Error?
Focus on how we judge the behaviors of others
When we see someone doing something, we tend to think it relates to their personality rather than the situation the person might be in.
The tendency to overestimate the degree to which an individual’s behavior is determined by his/her abiding personal characteristics, attitudes, or beliefs and correspondingly, to minimize the influence of the surrounding situation on that behavior.
What is Bias Blind Spot?
We are less likely to detect bias in ourselves than in others
We have the impression that we see issues and events objectively
What is the Anchoring Effect?
Our tendency to rely too heavily on the first piece of information offered, particularly if that information presented is in numeric form when making decision, estimates, or predictions.
Explain Representativeness Heuristic
This happens when people answer a question of probability or causation. i.e. how likely is that object A belongs to class B? By asking about the extent to which A resembles B
The representativeness heuristic involves judging the likelihood of something by considering how well it matches a particular category or prototype
Define Projection Bias
Misperception of the commonness of one’s beliefs, values, abilities, and behaviors - usually in the direction of overestimating how common they are
We come to see our own values and choices as relatively typical and appropriate and we view alternative responses as unusual and even deviant
Explain Choice Architecture / Behavioral Interventions
The idea is that decisions making and behavior is influenced in the systematic ways by subtle, seemingly insignificant changes in the decision making context
Change to decision making context alters decisions
“Choice Architecture”
* The design of the environment which people make choices
* how information/ environment is presented to you
We can improve decision making by structuring choice architecture
The presentation matters
Opting in vs. opting out
Explain Nudges
Any aspect of the choice architecture that alters people’s behavior in a predictable way without significantly changing their economic incentives
Interventions aimed at influencing behavior without limiting or forcing options
Paternalistic
* Automatic opt-in for employee participation in a company retirement plan
Numdges are being used in CJS
* Crime prevention, text prompts for court
* UK has “behavioral insight team”
Behavioral Nudges - Court
Bahavioral nudges reduce FTA for court
Many people fail to appear
We tend to see this as intentional so try to deter this through sanctions
* Bench Warrants
But can some of this be explained by simple human error- forgot/ not clear
Nudges
* To make sufficiently salient (see it/ remember it)
Behavioral Nudge - Court Redesigned Summons
Two large-scale studies in NYC evaluation interventions to make defendants more aware of court information
Redesigned the summons form that defendants received for low-level offenses
* Old summons listed court information at the bottom of the form below less important information
* New form moves court information to the top of the ticket where people are more likely to see it
* Also clearly states in bold typeface on the front of the form that missing the assigned court date will lead to a warrant
Behavioral Nudge - Text Messages
Send text message to highlight critical court information for defendants in the week leading up to the court date
Assigned to one of four groups:
* Control: No text
* Consequences: Court info/ potential warrant & arrest
* Plan-making: Mark calendar, look up directions, set alarm
* Combination: Mix of consequences & plan making
Behavioral Nudge - Sample-Impact New Form
All 323,922 summonses issued in NYC between 1/1/16 and 7/14/17
* Every summons had a serial number - could match serial numbers to old or new forms
Impact of the new form
* The new forms reduced failure to appear by 6.2 percentage points, or by 13.2% relative to the 47% baseline FTA rate in the estimation bandwidth
Behavioral Nudge - Impact of Texts
Relative to 37.9% failure rate in the control group, receiving any text message reduced FTA by 8 percentage points with represents a 21% relative reduction
Consequences & combination were the most effective
Plan-making also significantly reduced failures
Behavioral Nudge - Overall Impact
These nudges helped avoid at least 30,000 warrants being issued over 3 years and resulted in approximately 20,000 people having their cases fully dismissed instead of having an open warrant
Police Manipulations of Crime Reporting - Results
Email survey retired 1770 NYPD - 48.2% response
Confident major crimes declined by 80%
* 58.2% said no
Personal knowledge of making numbers look better
* changes from no to yes from periods 1 & 2 to period 3.
* Increase in Bloomberg era
Police Manipulations of Crime Reporting - Pressure
The pressure to reduce index crime steadily increased with significant differences for each time category
* 4.46, 5.45 and 6.36
The pressure to downgrade index crime also increased over the time periods
* 3.77, 4.16, and 5.63
1981-994 & 1994-2001 Not statistically different
2002 onward significantly different from previous eras
More pressure in the last era
Police Manipulations of Crime Reporting - Accuracy
Pre-Compstat low mean 5.25
Increased pressure to properly report when first Compstat is introduced 5.75
Does not hold for Kelly/ Bloomberg eras, reverts back to 5.28
Police Manipulations of Crime Reporting - Time to Retirement
More knowledge in recent years (highest pressure)
Personal knowledge of manipulation
* 30.3, 34.5 to 55.5 (Bloomberg era)
Cost of Crime - Explain Accounting Based
Attempts to identify all the costs associated with crime that individuals and society bear and place a dollar value on the costs
* Victim reports, hospital records, workers comp, first responder costs, mental health treatment costs, productivity losses, and incarceration costs
Have to use proxy for tangible costs
* Lawsuit awards - value of pain and suffering
* So less precise
Cost of Crime - Contingent Valuation
Elicit information about individual willingness to pay for crime reduction using survey questions
* What is this non-market good worth to you?
* Average willingness to pay for this thing
Respondents are typically asked whether they would support a hypothetical referendum in which they find a program providing specific benefits in exchange for tax increases of a given amount
A survey might ask individuals how much they are willing to pay for a reduction in crime
* 30% reduction in gun crime. What are you wiling to pay to
have this?
* Then infer cost of crime based on these resonses
Contingent Valuation - Strengths
Can capture the value of hard to measure things
Shows public preferences
Shows differences among population segments
Gives an indication of value/ cost before crime/ event happens
Contingent Valuation - Concerns
Hypothetically willing to pay is not the same as actually paying
* They may be willing to pay beyond their means
Surveys often have a low response rate - representative sample
Some may have poorly defined preferences
Cost of Crime - Hedonic Valuation
The idea that the price of a good is related to a number of characteristics - some of which are intrinsic to the good and some of which are extrinsic to it
Hedonic methodology shows that price of a good is determined not only by its physical qualities but also by its hedonic characteristics such as it’s location, amenities, or environment
Hedonic Valuation - Example
Take a house that is close to a scenic landscape. The proximity of this house to a scenic view increases it value relative to a similar house in another less favorable location. This in turn allows for an indirect determination of the value of the environment amenity itself by comparing the different prices attached to the two houses.
Compare the price of otherwise similar homes in neighborhoods with different exposure to crime
You can then infer the cost of crime exposure by the price differences
hedonic Valuation - Advantages
Uses actual consumer behavior in the market place to develop estimates, which avoids any of the biases introduced by contingent valuation
Captures tangible and intangible costs in the mind of the consumer
* Not hypothetical
Cost in Anticipation of crime - Tangible/ Intangible
Tangible
Expenditures to reduce the likelihood of victimization (i.e. purchasing security alarms)
Government crime prevention programs
Intangible
Avoidance behavior (i.e. avoiding people & places)
Fear of crime
Costs as direct consequence of crime - Tangible/ Intangible
Tangible
Lost wages & productivity because of victimization
Cost to recover or repair property
Intangible
Pain, suffering, lost quality of life because of victimization
Second generation costs (i.e. increased likelihood of same victims becoming offenders)
Costs in response to crime - Tangible/ Intangible
Tangible
Criminal justice costs like police investigations & incarceration
Cost to defend accused offenders in court
Intangible
Psychological cost to offender’s family & community
Overdeterrence costs (i.e. accusing innocent individuals, restricting legitimate activity in a community)
What are the 3 general types of costs of crime
Cost in anticipation of crime
*Security measures, crime prevention programs, insurance administration costs
Costs as direct consequence of crime
* Victims lost wages, employer lost productivity, victim medical/ mental health care, lost property and/ or insurance costs, funeral costs
Costs in response to crime
* Police investigation, prosecution, court, victims services, incarceration
Kansas City Patrol Experiment - Reported Crime
One statistically significant difference among the 51 comparisons drawn between reactive, control, and proactive
* Other sex crimes - molestation & exhibition - not all that is treatable by patrol
Kansas City Patrol Experiment - Victimization
No statistically significant difference in crime in any of the 69 comparisons made between reactive, control, and proactive beats
Kansas City Patrol Experiment - Arrest Patterns
No statistically difference in 27 comparisions
Kansas City Patrol Experiment - Rates of Crime
Five of the 48 statistically significant but no consistent pattern in the five - some in control, some proactive, and one in reactive
Kansas City Patrol Experiment - Citizen Fear of Crime
Citzin fear of crime was not significantly affected by the change in level of RPP
* 5 out 60 - 4 reactive and 1 in proactive
Kansas City Patrol Experiment - Proactive Measures
Measures taken by citizens against victimization were not significantly affected by variations in the level of routine policing
Kansas City Patrol Experiment - Business Proactive Measures
Did not show significant differences due to changes in the level of routine patrol
* 21 Comarisions
* Telephone survey not the best - don’t like revealing security measures over the phone
Kansas City Patrol Experiment - Attitudes Toward the Police
Citizens attitudes toward the police were not significantly affected by changes in the patrol level
111 comparisons with 16 instances of significant
* 5 reactive, 10 control beats, and 1 proactive
Kansas City Patrol Experiment - Business Attitudes Toward the Police
Not significantly affecteed
48 comparisons, non significant
* Poice-citizen encounters: citizen attitudes not significantly different
* Police officer attitudes not significantly different by ara of assignment
Kansas City Patrol Experiment - Response Time
The amount of response time taken by police in answering calls for service was not significantly affected by the variations in the level of routine patrol
42 comparisons, one significant finding: number of officers at the scene: More in reactive beats
Variations in patrol had no significant effect upon traffic accidents
Kansas City Patrol Experiment - Summary Findings
Decreasing or increasing routine preventive patrol within the range tested in this experiment had no effect on crime, citizen fear of crime, community attitudes toward the police on deliver of police services, police response time or traffic accidents
One other attempt has been made to replicate the Kansas City study, in Albuquerque, NM which yielded similar results
Kansas City Patrol Experiment - Conclusions
Not that routine patrol is of no value
* Citizens, business persons, and criminals all assumed status quo
Does show there is slack in the system
* Perhaps need some: but flexible and adding 2-3 times as much does not get you much more
* Dosage: High/ low relative to geography: Chemo vs. radiation
Patrol Staffing San Diego: 1 or 2 Officer Units - Findings
Overall performance for type and frequency of calls for service and officer initiated activities were about the same
Two officer units produced more traffic ciatations
One officer were more likely to result in arrest and formal report
No difference in those arrested by being held to answer by type of unit
Fewer complaints by citizens for one officer units
Two officer unit faster on calls but not enough to make up the cost of having two officers
One officer unit produced equivalent officer initiated activity
One officer less resisting arrest, equal on assaults and injuries
Same in auto accidents and miles driven (risk exposure to accident)
Officer have preference of two person units
Criminal Investigation Process: Clearance Rates - Findings
Not a great measure of effectiveness
* Variation in practice across departments in practice and definitions makes comparisons impossible
These definitions/ rules change over time within departments too
Administrative discretion high-subject to manipulation
Most important factors in solving cases:
* The single most important determinant of whether or not a case will be solved is the information the victim or witnesses supplies to the responding officer
* Role of responding officer and witnesses in solving cases
Types of Responses Police Problems Receive - Emotional Response
Programs which “intuitively seem like they should work,” “common sense”
Types of Responses Police Problems Receive - Political Response
Programs implemented as a political mandate, usually in response to a high profile problem
Types of Responses Police Problems Receive - Affective Response
Programs based on tested research with respoect to either “cause and effect” or “correlations”
Response Time & Citizen Evaluation of Police
Satisfaction with the responding officer
* Citizen satisfaction with response time was the best predictor of citizen satisfaction with responding officer
Attitudes toward the police in general
* Satisfaction with the responding officer was best predictor of general attitude toward the police
* Age and race were also associated with general attitudes
Response Time Relationship to Arrest for Part 1 Index Crimes
Police response time had no effect upon on-scene arrest for 70 to 85% of part 1 index crimes because most (62.3%) were discovered after completed
Response Time Relationship to Arrest for Property Crimes
“Discovered crime” - mobilization 10-15 minutes
Citizens “involved” in a crime-time to mobilize
* 4 to 7 minutes - burglary, larceny, or auto theft
Why do citizens take so long to mobilize?
Apathy - don’t want to get involved, not worth it
No use to police - no evidence or police would not think it’s important
Contact security (alternative)
Telephone another person
Observing the scene - more data to calculate seriousness
Injury
Citizens mobilization findings
Decision-making delays by citizens appear more important than access to phone
Victims are in stress, trying to make sense of what is happening or just happened
Million Dollar Murray
Chronic Homelessness
A small number of chronics contribute greatly to the overall problem
* Small number = much of the cost
* Consume vast amounts of resources/ create much of the cost of the problem
Normal Distribution
Bulk of the problem situated in the middle
The normal distribution is the state in which the variability of sample data points is similarly distributed before and after the mean value
More sample data is present around the mean value
Power Law Distribution
AKA Long Tail Distribution
The activity is not in the middle but at one extreme
There are a small number of people who are Hyper High Performers, a swath of people who are good performers and a smaller number of people who are low performers
It essentially accounts for a much wider variation in performance among the sample
Substantial amounts of an issue/ problem may be due to a few hard cases
The hard cases are hard
Political and moral impediments to solutions to problems with a power law distribution - Economic vs moral notions
Economic vs moral notions
Efficiency vs. fairness
* Treat all people in need the same vs impact the problem with differential treatment
Political and moral impediments to solutions to problems with a power law distribution - Political Views
Political Views
Special treatment for people who do not deserve special treatment
Emphasis on efficiency over fairness suggest the cold number-crunching of cost benefits anaylsis
Problems with Prevention - Why is it not liked?
Boring (not action)
Lacks satisfaction (nonevent)
Methodical and structured (limited discretion)
