NEW AND OLD MEASURES OF THE FEAR OF CRIME A MULTILEVEL ASSESSMENT OF MEASURES OF INTENSITY AND FREQUENCY Ian Brunton-Smith: University of Surrey.

Slides:



Advertisements
Similar presentations
Helen Chester University of Manchester. Brief overview of study and findings Focus on issues and recommendations for: Researchers wishing to do similar.
Advertisements

SADC Course in Statistics Assessing data critically Module B1 Session 17.
Sources and effects of bias in investigating links between adverse health outcomes and environmental hazards Frank Dunstan University of Wales College.
Exploring the Goodhart thesis at a local scale: neighbourhood social heterogeneity and perceptions of quality of life in the British Crime Survey John.
1 Bribery, crime and police abuse in Moldova Findings of the Soros Foundation - Moldova Victimisation Survey 2010.
Spatial Aspects of Robbery, Burglary and Vehicle Theft 10-Year Change in Neighborhood Characteristics and Crime.
Their Strengths and Limitations. 1. Practically – available for free 2. More detail as there are more categories of crime than with the British Crime.
Family-level clustering of childhood mortality risk in Kenya
OUTLINE Why are measures of crime important? Crime Rates v. Amounts
Research methods – Deductive / quantitative
TERRITORIAL FUNCTIONING AND VICTIMISATION IN COUNCIL ESTATES IN SHEFFIELD By: Aldrin Abdullah.
Assets, Wealth and Spousal Violence: Insights from Ecuador and Ghana Abena D. Oduro, University of Ghana Carmen Diana Deere, University of Florida Zachary.
Young People’s emotional well-being: The impact of parental employment patterns Dr Linda Cusworth Social Policy Research Unit, University of York International.
The British Crime Survey Face to face interviews with a sample of adults (16+) living in private households in England and Wales Measures crime victimisation.
1 Neighbourhoods matter: spill-over effects in the fear of crime Ian Brunton-Smith Department of Sociology, University of Surrey.
Modelling Crime: A Spatial Microsimulation Approach Charatdao Kongmuang School of Geography University of Leeds Supervisors Dr. Graham Clarke, Dr. Andrew.
The Impact of Crime. How does crime impact on people? The immediate impact – physical harm, loss of / damage to property The ‘aftermath’ (fear of crime)
A model for spatially varying crime rates in English districts: the effects of social capital, fragmentation, deprivation and urbanicity Peter Congdon,
Patterns of Police Reporting Amongst Victims of Partner Abuse: Analysis of the SCJS 2008/09 Sarah MacQueen Scottish Centre for Crime and Justice Research.
LECTURE 3 Introduction to Linear Regression and Correlation Analysis
Small Area Estimates of Fuel Poverty in Scotland Phil Clarke (ONS), Ganka Mueller (Scottish Government)
University of Oxford National data – local knowledge Using administrative data David McLennan & Kate Wilkinson Social Disadvantage Research Centre Department.
Adding Census Geographical Detail into the British Crime Survey for Modelling Crime Charatdao Kongmuang Naresuan University, Thailand Graham Clarke and.
Clustered or Multilevel Data
Today Concepts underlying inferential statistics
Analysis of Clustered and Longitudinal Data
Chapter 12 Inferential Statistics Gay, Mills, and Airasian
David Card, Carlos Dobkin, Nicole Maestas
The new HBS Chisinau, 26 October Outline 1.How the HBS changed 2.Assessment of data quality 3.Data comparability 4.Conclusions.
THE CRIME AND JUSTICE SURVEY Research, Development and Statistics BUILDING A SAFE, JUST AND TOLERANT SOCIETY Tracey Budd.
Multilevel models for predicting personal victimisation in England and Wales Andromachi Tseloni Analysis of crime data ESRC Research Methods Festival 2010.
Chapter 10 Hypothesis Testing
Crime reporting and intervention Psychology of Crime.
Following lives from birth and through the adult years Examining the truth behind the myth of the 'the Monstrous Army on the March' Dylan.
Measurement Error.
Copyright © 2008 Pearson Education Canada Inc Crime Statistics Chapter 2.
Chapter Twelve Census: Population canvass - not really a “sample” Asking the entire population Budget Available: A valid factor – how much can we.
Tobacco Control Research Conference July 2014 Determinants of smoking initiation in South Africa Determinants of smoking initiation in South Africa.
Berna Keskin1 University of Sheffield, Department of Town and Regional Planning Alternative Approaches to Modelling Housing Market Segmentation: Evidence.
Scot Exec Course Nov/Dec 04 Survey design overview Gillian Raab Professor of Applied Statistics Napier University.
Additional analysis of poverty in Scotland 2013/14 Communities Analytical Services July 2015.
GEOG3025 Census and administrative data 1: Sources and methods.
Highway accident severities and the mixed logit model: An exploratory analysis John Milton, Venky Shankar, Fred Mannering.
2007 CAS Predictive Modeling Seminar Estimating Loss Costs at the Address Level Glenn Meyers ISO Innovative Analytics.
Centre for Housing Research, University of St Andrews The Effect of Neighbourhood Housing Tenure Mix on Labour Market Outcomes: A Longitudinal Perspective.
Chapter 16 Data Analysis: Testing for Associations.
General Register Office for S C O T L A N D information about Scotland's people Household Estimates and Projections Esther Roughsedge General Register.
Spatial Statistics in Ecology: Point Pattern Analysis Lecture Two.
Multivariate Data Analysis Chapter 1 - Introduction.
Chapter 6: 1 Sampling. Introduction Sampling - the process of selecting observations Often not possible to collect information from all persons or other.
Victims, Communities & Society
Extending the crime survey to include fraud and cyber Joseph Traynor.
CROSS-COUNTRY INCOME MOBILITY COMPARISONS UNDER PANEL ATTRITION: THE RELEVANCE OF WEIGHTING SCHEMES Luis Ayala (IEF, URJC) Carolina Navarro (UNED) Mercedes.
An ecological analysis of crime and antisocial behaviour in English Output Areas, 2011/12 Regression modelling of spatially hierarchical count data.
Experience and expression in the fear of crime Stephen Farrall, University of Keele, Emily Gray, University of Keele & Jonathan Jackson, London School.
Reliability performance on language tests is also affected by factors other than communicative language ability. (1) test method facets They are systematic.
Statistics Canada Citizenship and Immigration Canada Methodological issues.
By R. Gambacorta and A. Neri Bank of Italy - Statistical Analysis Directorate Wealth and its returns: economic inequality in Italy, The Bank.
Can we detect ‘Thatcher’s Children’ in data on attitudes to crime and disorder: A longitudinal analysis of age, period and cohort effects. Emily Gray*,
NCRM is funded by the Economic and Social Research Council 1 Interviewers, nonresponse bias and measurement error Patrick Sturgis University of Southampton.
Restaurant Smoking Policies and Reported Exposure to ETS The case of Massachusetts Tandiwe Njobe National Conference on Tobacco or Health November 2002.
University of Warwick, Department of Sociology, 2012/13 SO 201: SSAASS (Surveys and Statistics) (Richard Lampard) Survey Design: Some Implications for.
Copyright © 2013 Pearson Education, Inc. Publishing as Prentice Hall Statistics for Business and Economics 8 th Edition Chapter 9 Hypothesis Testing: Single.
Educational Research Inferential Statistics Chapter th Chapter 12- 8th Gay and Airasian.
Denise Kendrick University of Nottingham.  Inequality or inequity?  Differences in injury risk ◦ Child factors ◦ Family factors ◦ Social factors ◦ Environmental.
Partner violence among young adults in the Philippines: The role of intergenerational transmission and gender Jessica A. Fehringer Michelle J. Hindin Department.
Kobe Boussauw – 15/12/2011 – Spatial Planning in Flanders: political challenges and social opportunities – Leuven Spatial proximity and distance travelled:
When is a crime not a crime
Revised Questionnaire EU Survey Eurostat, June, Prof Jan Van Dijk
Chapter 2 The Incidence of Crime
Presentation transcript:

NEW AND OLD MEASURES OF THE FEAR OF CRIME A MULTILEVEL ASSESSMENT OF MEASURES OF INTENSITY AND FREQUENCY Ian Brunton-Smith: University of Surrey

Outline Introduction –Measures of fear –Neighbourhood influence –Multilevel modelling Area structure –Local focus within a national framework –Initial results- area differences? Expanded models –Incorporating neighbourhood data –Preliminary findings Summary

Current research The influence of local area differences on fear of crime –British Crime Survey individual level data Inclusion of area based measures to characterise neighbourhood difference –Geodemographic data from administrative sources –Local neighbourhood focus within a national survey context Examine the frequency, as well as intensity of personal fear –New questions- alternative look at fear of crime (Farrall, Jackson and Gray)

INTENSITY “How worried are you about being ‘mugged or robbed’?” Very worried, fairly worried, not very worried, or not at all worried? Use of ‘how’ is leading- suggestive that all people experience worry No specific timeframe Difficulty of summarising emotions Fail to distinguish specific events from generalised anxieties Measuring fear of crime: Introducing the frequency of fear FREQUENCY Q.1 “During the last 12 months, have you ever felt worried about being ‘mugged or robbed’?” Yes, no Q.2 “If yes, how many times have you felt like this in the past 12 months?” Relate to actual experiences (not opinions/attitudes) Refer to a specific time frame Use of a filter question reduces overestimation Distinguish between low level emotional states and intense but transitory reactions to specific events (Farrall et al, 2006)

Differences in the prevalence of fear of crime Clear differences in overall estimates of fear by crime type Reduced estimates of fearful population with frequency measures Intensity (Very worried)Frequency (Once a month or more) Mugging 34.0% (10.5%)13.5% (3.3%) Burglary 45.3% (12.2%)30.0% (7.5%) BCS 2003/04

Local area focus? “Increasing awareness of the importance that local area characteristics can have on individual outcomes” (Ellen and Turner, 1997) –Policy interventions –Understanding social processes Growing focus on the impact of local area influence on fear of crime –Recorded crime rates... –Incivilities Recent availability of Geodemographic identifiers on British Crime Survey Increasing use of multilevel analysis techniques –Incorporate context in individual level analysis

Multilevel modelling Individuals often subject to the influences of groupings (pupils within schools, individuals within neighbourhoods) –Individuals within an area will be more alike, on average, than they will be alike those of another area (dependency) Estimate neighbourhood based models within individual level analysis –Relative contribution of individual and area to unexplained variation in dependent variable Area variables included at correct level of influence –Avoiding problems of aggregation or disaggregation

Data BCS 2003/04: Sub-sample of respondents (n=4,000) asked both question types 2 crime types used, matching ‘worry’ questions Truncated to more closely resemble intensity measures INTENSITYFREQUENCY How worried are you about...During the last 12 months, have you ever felt worried about... If yes, how many times have you felt like this in the past 12 months?...being mugged or robbed?...having your home broken into and something stolen?...having you’re your home broken into and something stolen? No worry (not at all/not very) No, never worried Fairly worried Yes, up to once a month (1-12 reported episodes) Very worried Yes, more than once a month (13+ episodes)

Ordered multilevel regression modelling Proportional odds Underlying Linear threshold model- –Conservative estimate of area variance Reduced model assumptions –Improved estimates and standard errors –Particularly important for area component (typically underestimated when normality assumption violated) Logit link function –Odds of moving into a higher fear category, controlling for the influence of the other covariates

Local area structure MSOA –Clusters of census output areas –Account for population size, area proximity and social homogeneity –Average population 2,500 households –Spatially and statistically stable –6,781 in England Sample nesting –2,046 MSOA –Max 8 respondents per MSOA –Limits model complexity Individual MSOA

Results I: Existence of area differences Individual level controls –Gender; Age, Education; SEC; Ethnicity; Length of time in area... –Direct victimisation experience Significant impact of local area differences on fear of mugging –Both intensity and frequency No significant area component to burglary –No further examination required Intensity of fear Frequency fearful events Variance (S.E)Percentage of total variance Variance (S.E)Percentage of total variance Mugging.244(.078)6.9%.240(.130)6.8% Burglary.048(.059)1.4%.050(.070)1.5% BCS 2003/04

Method Area level measures –Neighbourhood data from Census and National Statistics –Range of neighbourhood characteristics identified Disadvantage; household type; housing details; land usage; population structure –Multicollinearity problem Factorial ecology “Components (factors) represent concise descriptions of patterns of associations of attributes across observations” (Rees, 1971: 221) –Identify key contextual attributes –PCA at neighbourhood level, based on data for all MSOA –4 factors extracted- 80% of total variance

Socio-Economic Disadvantage – % income support; % unemployed; % lone parents; % managerial; % car owners; % terraced housing; % local authority owned Urbanisation and overcrowding – Pop density; % flats; % people per room; % agriculture; % green-space; % domestic property Transition Communities – % in-migration; % out-migration; % vacant properties; % single-person non-pensioners Age Structure – % younger residents (<16); % older residents (65+); % owner occupied Method II: PCA Extracted Factors

Method III: Additional Area Characteristics Ethnic Diversity (Taylor and Hudson, 1972) –Herfindahl concentration formula = –Si is the share of group i out of a total of n groups –Probability that two randomly selected individuals from the same area will be of different ethnic origin Interviewer rating of area –Measure of low level area disorder –Housing quality; extent of litter; vandalism and graffiti Crime Index (from IMD) –Incidence of recorded crime for 4 major crime types –Burglary (4 offence types) –Theft (5 offence types) –Criminal Damage (10 offence types) –Violence (14 offence types)

Results II: Mugging and robbery Neighbourhood level of socio-economic disadvantage, and low level disorder associated with intensity of fear, but not frequency Recorded crime associated with frequency of fearful episodes, but not intensity of fear Ethnic fractionalisation, urbanisation and extent of transition associated with both intensity and frequency Intensity measureFrequency measure Odds RatioPrecision Odds RatioPrecision Socio-Economic Disadvantage1.11* Urbanisation and overcrowding1.23* * 2.79 Transition Community.90* * 2.29 Age Structure Ethnic Diversity2.09* * 3.38 Low-level disorder1.14* Recorded crime * 2.61 Remaining Area Variance4.7%*3.5% BCS 2003/04 * Sig p(<.05)

Summary Incorporating local neighbourhood information can provide new clues about what does and doesn’t influence fear of crimes Neighbourhood people live in makes an important contribution to fear of mugging, but has no significant influence on levels fear burglary –Extent of area variation consistent when considering frequency and intensity of fear Clear differences in the neighbourhood influences on mugging when looking at the frequency of fear compared to intensity –People’s intensity of fear more influenced by external factors Low level neighbourhood disorder, and increasing socio-economic disadvantage No significant association with recorded levels of crime –The frequency of recalled fearful incidents more directly affected by crime Positive relationship with recorded crime figures