crime survey for england and wales (csew) dataset user guide
Contents
1. The response rate for the core survey in 2011 12 was 75 per cent For the children s survey 67 per cent of children within households responding to the main CSEW participated The true response rate taking into account first stage non response to the main CSEW was in the region of 50 per cent For more information on response rates see TNS BMRB 2012 2 2 Fieldwork After obtaining informed consent from the parent guardian and the selected child child respondents were interviewed as in the main CSEW using a combination of computer assisted personal interviewing CAPI and computer assisted self interviewing CASI The latter was used to administer the self completion section of the questionnaire that covered topics respondents might not have been comfortable answering in the presence of an interviewer or their parent or guardian Audio CASI was also available for those respondents who might otherwise have found the self completion module problematic due to reading difficulties The children s questionnaire takes on average around 20 minutes to complete 2 3 Questionnaire The CSEW children s questionnaire consisted of a core set of modules asked of the whole sample a set of modules asked only of different sub samples and a self completion module The questionnaire for children was modelled on the personal victimisation module otherwise known as the victim module see TNS BMRB 2012 of the adult questionnaire with the expec2. because the respondent reported that there was intent to commit an offence these incidents are still considered offences under law The CSEW offence coding system ensures that incidents are recorded as accurately as possible A full list of core and non core offence codes for the CSEW children s extension appears in Appendix 1 For more information on offence coding for the 10 to 15s survey see TNS BMRB 2012 Final offence codes may not necessarily correspond with the screener questions which lead to information collected in the personal victimisation modules for example an incident elicited from the threat screener may turn out to be a case of common assault Children aged 10 to 15 crime statistics are produced from the offence coded data and presented as incidence or prevalence rates based on data about incidents or victims see Chapter 3 2 7 Classifying incidents reported by children aged 10 to 15 Respondents to the CSEW children s survey report a wide variety of incidents with varying degrees of severity As part of the development process to extend the CSEW to those aged 10 to 15 the existing CSEW offence coding system was used to code incidents reported by children Four approaches to counting crime were then published Millard and Flatley 2010 and following a user consultation two approaches were favoured with regard to estimating levels of victimisation among children the Broad measure and Preferred measure a
3. 15 this offence group also includes the offences grouped with the following non core offence codes 146 Theft with aggressive behaviour 147 Attempted theft with aggressive behaviour Theft with violence unspecified crobbeh 146 Theft with aggressive behaviour 147 Attempted theft with aggressive behaviour 20 Other thefts of personal property cothpthef 67 Other theft 73 Other attempted theft For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 166 Theft 167 Attempted theft Theft of personal property unspecified ctheft 166 Theft 167 Attempted theft Bicycle theft cbikthft 64 Theft of a pedal cycle Theft in a dwelling cthindwel 55 Theft in a dwelling Theft from outside a dwelling cthoutdwel 65 Theft from outside a dwelling Theft from in outside a dwelling cthdwel 55 Theft in a dwelling 65 Theft from outside a dwelling All thefts within CSEW household crime cthholdall 55 Theft in a dwelling 64 Theft of a pedal cycle 65 Theft from outside a dwelling Personal acquisitive crime cpersacq 41 Robbery 42 Attempted robbery 43 Snatch theft from the person 44 Other theft from the person 45 Attempted theft from the person 67 Other theft 73 Other attempted theft For the broad measure approac
4. Bicycle theft and Vandalism are designated as household offences for adults on the CSEW respondents reply on behalf of the household However they are presented in Chaplin et al 2011 as offences against children when the property stolen or damaged was reported by the respondent as solely belonging to them This broadens the scope of personal victimisation which may result in some double counting of offences with the adult survey the extent to which this happens will be evaluated in the future Please note that no statistical assessment of the change in levels of victimisation between the 2009 10 2010 11 and 2011 12 CSEW has been made This is due to the developmental nature of the survey and estimates which have been subject to further refinement during the periods presented Each crime type has a specific question relating to intent e g intent to permanently deprive in the case of a theft However it is not possible to apply the simple criterion of excluding cases of a certain crime type with no evidence of intent to robbery as this offence involves both threat or violence and theft 15 Chapter 6 Methodological limitations of the CSEW 10 to 15s survey It should be noted that the figures derived from the Crime Survey for England and Wales CSEW are estimates As with any sample survey the CSEW estimates are subject to sampling error and a range of other methodological limitations 6 1 Non response As in any v
5. behaviour 166 Theft 167 Attempted theft 186 Damage to property 187 Attempted damage to personal property All crime calchall 11 Serious wounding 12 Other wounding 13 Common assault 21 Attempted assault 41 Robbery 42 Attempted robbery 23 43 44 45 55 64 65 67 73 805 806 Snatch theft from the person Other theft from the person Attempted theft from the person Theft in a dwelling Theft of a pedal cycle Theft from outside a dwelling Other theft Other attempted theft Arson to personal property 10 15s only Criminal damage to personal property 10 15s only For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 116 117 146 147 166 167 186 187 Aggressive behaviour Attempted aggressive behaviour Theft with aggressive behaviour Attempted theft with aggressive behaviour Theft Attempted theft Damage to property Attempted damage to personal property 24 Bibliography Chaplin R Flatley J and Smith K Eds 2011 Crime in England and Wales 2010 11 Findings from the British Crime Survey and police recorded crime Home Office Statistical Bulletin 10 11 http Awww homeoffice gov uk publications science research statistics research statistics crime research hosb1011 Fitzpatrick A Grant C Bolling K Owen R and Millard B 20
6. cyber security and the experience of crime among children Findings from the 2010 11 British Crime Survey Supplementary Volume 3 to Crime in England and Wales 2010 11 Home Office http Awww homeoffice gov uk publications science research statistics research statistics crime research hosb0612 Statistics Commission 2006 Crime statistics User perspectives Statistics Commission Report No 30 http www statisticsauthority gov uk reports correspondence archive statistics commission archive research index html 25 TNS BMRB 2012 The 2011 12 Crime Survey for England and Wales Technical Report Volume One http www ons gov uk ons guide method method quality specific crime statistics methodology 201 1 12 crime survey for england and wales technical report volume one pdf 26
7. from the existing adult survey and this should be borne in mind when analysing the data 1 The adult survey collects information about sexual offences in a separate self completion section which is not asked of children aged 10 to 15 Where there was more than one offender detailed information was collected if any of the offenders were not pupils at the respondent s school A weapon constitutes any item that was considered to be a weapon by the victim this includes knives sticks stones bottles In addition to questions about experience of crime the survey also gathers information on a number of crime related topics such as perceptions and attitudes to the police anti social behaviour crime prevention and personal security 2 4 Time periods covered Like the adult CSEW child respondents were asked about crime in the 12 months prior to interview Since respondents are interviewed at different times within each month they are asked about experiences of crime in the current month plus the 12 months prior to interview Crimes experienced in the interview month are excluded from the 12 month reference period used for analysis Hence for the 2011 12 children aged 10 to 15 CSEW the reference period includes incidents experienced by respondents between April 2010 and February 2012 The centre point of the period for reporting crime is March 2011 the only month to be included in all respondents reference periods 2 5
8. inclusion of such victims could undermine the ability to measure trends consistently This sort of capping is in line with other surveys of crime and other topics Prevalence rates are not affected by this procedure see TNS BMRB 2012 for information on the measurement of series data 2 6 Offence coding Based on information collected and processed from the personal victimisation modules specially trained coders determine whether what has been reported constitutes a crime and if so what offence code should be assigned to the crime All incidents are allocated a core offence code except the relatively low level incidents recorded on mini victim modules which are allocated non core offence codes Incidents which met the criteria for a mini victim module have a limited amount of information collected which enables classification to a high level crime category Without detailed information it is not possible to assign specific offence codes within the appropriate high level classification according to standard CSEW procedures As a result these cases are designated as unspecified offences Without an offence code it is not possible to tell which detailed crime type the offence would be classified as For example data on whether the stolen item was being carried by the respondent at the time of a theft were not collected so it is not feasible to determine whether this would be a theft from the person or other theft of personal property However
9. repeat victimisation will be reflected in relatively lower prevalence rates compared with incidence rates Repeat victimisation variables for each offence group are included on the non victim form dataset and the variable names include the suffixes _r_brd _r2_brd or _r_prf i 2 prt 7 Note that this example uses unweighted figures Analysis should always be weighted appropriately 12 4 6 Weighting Two types of weighting are used to ensure that the sample of children is representative of the population of those aged 10 to 15 living in households in England and Wales First the raw data are weighted to compensate for unequal probabilities of selection A review of the CSEW by survey methodology experts at ONS and the National Centre for Social Research NatCen recommended that the calibration weighting method be adopted for the CSEW Lynn and Elliot 2000 The weighting is designed to make adjustments for known differentials in response rates between different regions and different age by sex subgroups and also households with different age and sex composition The procedure gives different weights to different household types based on their age sex composition in such a way that the weighted distribution of individuals in the responding households matches the known distribution in the population as a whole and also matches the known distribution of the regional population The weights are generated using an algorithm that m
10. 10 Extending the British Crime Survey to children a report on the methodological and development work TNS BMRB and the Home Office http webarchive nationalarchives gov uk 20110220105210 http www homeoffice gov uk rds pdfs 10 bcschildren pdf Lynn P and Elliot D 2000 The British Crime Survey A review of methodology National Centre for Social Research http webarchive nationalarchives gov uk 20110218135832 http rds homeoffice gov uk rds pdfsO 8 bcs methodology review 2000 pdf Millard B and Flatley J 2010 Experimental statistics on victimisation of children aged 10 to 15 Findings from the British Crime Survey for the year ending December 2009 Home Office Statistical Bulletin 11 10 http webarchive nationalarchives gov uk 20110220105210 rds homeoffice gov uk rds pdfs10 hosb 1110 pdf Pickering K Smith P Bryson C and Farmer C 2008 British Crime Survey options for extending the coverage to children and people living in communal establishments Home Office Research Report 06 http webarchive nationalarchives gov uk 20110218135832 http rds homeoffice gov uk rds pdfsO 8 horr06c pdf Smith A 2006 Crime statistics An independent review Home Office independent report http webarchive nationalarchives gov uk 20110218135832 http rds homeoffice gov uk rds pdfsO 6 crime statistics independent review 06 pdf Smith K Ed Lader D Hoare J and Lau l 2012 Hate crime
11. 44 Other theft from the person 45 Attempted theft from the person Theft 67 Other theft Attempted theft 73 Other attempted theft 801 Arson to a motor vehicle 10 15s only 802 Criminal damage to a motor vehicle 10 15s only Vandalism 803 Arson to the home 10 15s only 804 Criminal damage to the home 10 15s only 18 805 Arson to personal property 10 15s only 806 Criminal damage to personal property 10 15s only Three digit offence codes are used here to differentiate children s offence codes from those of adults as it is not possible to use the same offence codes because children are not asked whether the item damaged was worth more or less than 20 Valid non core offence codes used prior to 2011 12 Code Description 116 Aggressive behaviour 117 Attempted aggressive behaviour 146 Theft with aggressive behaviour 147 Attempted theft with aggressive behaviour 166 Theft 167 Attempted theft 186 Damage to property 187 Attempted damage to property 196 Intimidation coercion 19 Appendix 2 Offence groups Once incidents from the personal victimisation module have been given offence codes they are then classified into groups of incidents The different offence groups are listed below as well as offence codes which comprise each group These groups are used to calculate incidence and prevalence rates The letters in brackets after the group
12. 7 Appendix 1 CSEW 10 to 15s survey offence codes The children s Crime Survey for England and Wales CSEW is designed to collect information about personal victimisation among children aged 10 to 15 The purpose and design of the survey is such that it is not suitable for collecting information on sexual victimisation However children responding to the survey do report small numbers of household and sexually motivated incidents Household victimisation includes those incidents where any or all members of the household may be regarded as victims Household level incidents are coded with existing offence codes but also some new codes specific to the children s questionnaire Incidents with a sexual motivation or including a sexual element are coded with the corresponding offence codes before then being referred to the Home Office on a case by case basis Sexual offences reported by children are not included in the analysis to calculate estimates of incidence and prevalence of victimisation for children Valid core offence codes Category Code Description 11 Serious wounding Assault 12 Other wounding 13 Common assault Attempted assault 21 Attempted assault 31 Rape 32 Serious wounding with a sexual motive Sexual offences 33 Other wounding with a sexual motive 34 Attempted rape 35 Indecent assault 41 Robbery 42 Attempted robbery Personal theft 43 Snatch theft from the person
13. Measures of crime The CSEW extension to children provides information about personal crimes experienced by children aged 10 to 15 Personal crimes relate to all crimes against the individual and only relate to the respondents own experience not that of other people in the household An example of a personal crime would be an assault There are three stages to the questionnaire for measuring experiences of victimisation amongst children aged 10 to 15 First respondents are asked a series of screener questions on the main part of the questionnaire to assess if they have been a victim of crime Screener questions do not ask respondents if they have been a victim of explicit crime types but ask about different experiences such as whether the respondent has had anything stolen in the last 12 months This design ensures that all incidents of personal crime within the scope of the children s survey including relatively minor ones are included Following the screener questions those who have been victimised are asked a number of short questions about the nature of their victimisation in order to identify relatively low level incidents This section of the questionnaire is called the mini victim module and identifies incidents which are minor in nature such as aggressive behaviour resulting in no injury pushing and shoving other thefts of low value items stationery food threats not involving a weapon and damage to belongings stationery T
14. UK Data Archive Study Number 7252 Crime Survey for England and Wales 2011 2012 lt Office for ANS National Statistics Contents Chapter 1 Chapter 2 Chapter 3 Chapter 4 Chapter 5 Chapter 6 Appendix 1 Appendix 2 Bibliography Background to the CSEW extension to children aged 10 to 15 3 Methodology of the CSEW 10 to 15s survey eeeeeeeeeeeeeeeeeeeeeeeeeeeeeerees 4 2 1 Sample CSSION iss see ee E 4 2 2 FICO WON decina a ee Ee 4 2 9 DU CRUIGIIAING sssaaa aaa a a a a a a a a a EAA 4 2 4 Time periods covered sista ciiiniceniicaidonsicnihenteiniceniantheniieniians eaten 6 2 5 Measures Of CME se eS ace aoc onesies epsa ee ey eee Nerada 6 2 6 Offente COdING 2 2 2 2 2h2 dese deena dee eee elected 7 2 7 Classifying incidents reported by children aged 10 to 15 oo eee 7 CSEW 10 to 15s survey dataset ccecceeceeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeneeenene 9 3 1 Datasets available sorrrrenrsnnscnononini reima R R 9 3 2 Variable MAIS S aistdencacaarantdGarecedeesigestssebeactueibuodiaslneadsoadosodseMoondetaUeannengieeds 9 3 3 Derived VAM AS ia cate ta cas etenattnten ha atnten ae esinsad aabtntaenabsnsndaaatotaenabiesnlsaneatans 10 3 4 Don t know and refusal COd CS wreciiecsceescnesectienverseees ves esyeeeeeee vee iene eekeneee 10 Analysis of CSEW 10 to 15s Survey data 00 eee eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 11 4 1 Individual based analysis seas shoes ces ee en cee ence a eee ee 11 4 2 Incident based
15. analysis a er 11 4 3 INeidente ral Senanin aia eaaa eaaa 11 4 4 Prevalence Wale S aysticoersneeiantiiieediguidanisaniiaxditedeniaxdiantnlenienieee 12 4 5 Repeat and multiple victimisation cccceeeceeeeeeeeeeeeeeeeeeeeeeeaeeeeees 12 4 6 Weighting ssns aaaeaii aiaiai aiaei 13 4 7 Statistical SSV MN CM CS a sc eget sreeesendesacd asec ensaseetensaseeece 13 Published data from the CSEW 10 to 15s SUIVeY eeeeeeeeeeeeeeeeeeeeeeees 15 Methodological limitations of the CSEW 10 to 15s survey ee 16 6 1 Non reSpOnSe aieeaa iiaeaa i aiaa EEEa 16 6 2 FRAN ssaa e EE E a a EE OEE 16 6 3 Unwillingness to report oc nice cscs sceeccneeceecenvnceescevetevscnvnteveceyntevsenvnteveteeneeeeeens 16 6 4 Definitions Of CME ccicccssessuasabervesrgcenssavaedsveveeenazeravayseaveeersnevasesseavenenneevate 17 CSEW 10 to 15s survey offence codes sssrsrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrnn 18 Seck KONT o Ee A ree 20 Fe ee ee nT a a a ee eee r 25 Chapter 1 Background to the CSEW extension for children aged 10 to 15 The Crime Survey for England and Wales CSEW formerly known as the British Crime Survey BCS is a face to face victimisation survey in which people resident in households in England and Wales are asked about their experiences of a range of crimes in the 12 months prior to the interview Respondents to the survey are also asked about their attitudes towards different crime related issues such as the police a
16. by children aged 10 to 15 Preliminary results from the first calendar year were published in 2010 Millard and Flatley and following a user consultation these statistics were refined further The results for 2010 11 were published in two reports Chaplin et al 2011 and Smith et al 2012 The questionnaire was refined again for the 2011 12 survey The results for 2011 12 were published by the Office for National Statistics in July 2012 Crime Statistics Period Ending March 2012 The changes to the questions and definitions used should be borne in mind when interpreting the figures Methodological differences also mean that direct comparisons cannot be made between the adult and child data Millard and Flatley 2010 Chapter 2 Methodology of the CSEW 10 to 15s survey 2 1 Sample design The Crime Survey for England and Wales CSEW sample of children was designed to be representative of the population of children aged 10 to 15 resident in households in England and Wales The sample itself was obtained from those households already selected to take part in the main adult survey In each household containing a child in the eligible age range the interviewer randomly selected one child for interview following the completion of the main adult interview Interviews with a total of 3 930 children aged 10 to 15 were conducted during 2011 12 Compared with other similar household surveys the CSEW has managed to maintain a high response rate
17. ce rates prevalence rates only take account of whether a child aged 10 to 15 was a victim of a specific crime once or more during the reference period not the number of times they were victimised Respondents are thus designated as victims or non victims The proportion of respondents who are victims provides the prevalence rate equivalent to the risk of being a victim of crime See Appendix 2 for an explanation of the variables used to calculate prevalence rates Prevalence variables for each of the offence groups are included on the non victim form dataset and the variable names include the suffix _p_brd or _p_prf The number of victims of personal crimes can also be estimated for England and Wales the proportion of children aged 10 to 15 that were victims should be multiplied by the total number of children aged 10 to 15 in England and Wales 4 5 Repeat and multiple victimisation The CSEW children s datasets contain variables which allow users to examine rates of repeat and multiple victimisation Multiple victimisation is the estimated percentage of children aged 10 to 15 who have been a victim of more than one personal crime Repeat victimisation a subset of multiple victimisation is defined as being a victim of the same type of crime more than once in the last 12 months for example vandalism Levels of repeat victimisation account for differences between incidence rates and prevalence rates For instance high levels of
18. edures As a result these cases have been designated as unspecified offences Without an offence code it is not possible to tell which detailed crime type the offence would be classified as For example data on whether the stolen item was being carried by the respondent at the time of a theft were not collected so it is not feasible to determine whether this would be a theft from the person or other theft of personal property However because the respondent reported that there was intent to commit an offence these incidents are still considered offences under law Unspecified offences do not fall within the scope of the Preferred measure because the detailed information above was not collected In 2011 12 this was changed and full information was collected about all incidents of crime with a reduction in the number of incidents asked about to balance data collection needs with respondent burden This means that the data are not directly comparable over the three time periods It should be noted that questions asked of adults about household crimes such as burglary or vehicle related crime were not included in the children aged 10 to 15 survey as these were already captured from the adult interview The small number of household incidents reported by children were given household level offence codes but were not included in any further analysis As such the composition of crime covered in the children aged 10 to 15 survey differs
19. h to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 166 Theft 167 Attempted theft All personal thefts cthpers 43 Snatch theft from the person 44 Other theft from the person 21 45 Attempted theft from the person 67 Other theft 73 Other attempted theft For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 166 Theft 167 Attempted theft All thefts ctheftall 43 Snatch theft from the person 44 Other theft from the person 45 Attempted theft from the person 55 Theft in a dwelling 64 Theft of a pedal cycle 65 Theft from outside a dwelling 67 Other theft 73 Other attempted theft For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 166 Theft 167 Attempted theft Wounding cwounding 11 Serious wounding 12 Other wounding Assault ccomasinj 13 Common assault 21 Attempted assault Robbery crobbery 41 Robbery 42 Attempted robbery All violence cviolnos a Serious wounding 12 Other wounding 13 Common assault 21 Attempted assault 32 Serious wounding with sexual motive 33 Other wound
20. he mini victim module was incorporated into the design of the children s questionnaire following questionnaire testing in order to reduce respondent burden one of the recommendations of the methodological advice on the feasibility of extending the survey had been that the survey not be longer than 20 minutes on average Only basic information is collected about these types of relatively low level incidents Full details of experiences of personal crime are then recorded in the personal victimisation modules of the questionnaire Most incidents reported are one off single occurrences but in a minority of cases respondents may have been victimised a number of times in succession In these cases respondents are asked whether they consider these incidents to be a series that is the same thing done under the same circumstances and probably by the same people Where incidents are determined to be in a series the number of incidents is recorded but with only one personal victimisation module being completed based on the most recent incident CSEW estimates only include the first five incidents in this series of victimisations in the count of crime The restriction to the first five incidents in a series has been applied since the CSEW began in order to ensure that estimates are not affected by a very small number of respondents who report an extremely high number of incidents and which are highly variable between survey years The
21. ing with sexual motive 41 Robbery 42 Attempted robbery For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 116 Aggressive behaviour 22 117 Attempted aggressive behaviour Aggressive behaviour unspecified caggbeh 116 Aggressive behaviour 117 Attempted aggressive behaviour Damage to personal property ccrimdam 805 Arson to personal property 10 15s only 806 Criminal damage to personal property 10 15s only Damage to personal property unspecified cdamage 186 Damage to property 187 Attempted damage to personal property All personal not including sex offences calchallp 11 Serious wounding 12 Other wounding 13 Common assault 21 Attempted assault 41 Robbery 42 Attempted robbery 43 Snatch theft from the person 44 Other theft from the person 45 Attempted theft from the person 67 Other theft 73 Other attempted theft 805 Arson to personal property 10 15s only 806 Criminal damage to personal property 10 15s only For the broad measure approach to classifying incidents reported by children aged 10 to 15 this offence group also includes the offences grouped with the following non core offence codes 116 Aggressive behaviour 117 Attempted aggressive behaviour 146 Theft with aggressive behaviour 147 Attempted theft with aggressive
22. inimises the differences between the weights implied by sampling and the final weights subject to the weighted data meeting the population controls They are based on calibration of population figures provided by the ONS Labour Force Survey LFS Calibration weights were applied from the 1996 CSEW onwards using CALMAR a SAS based macro since 2006 07 the CSEW has used g Calib within a new SPSS based data processing system the weights produced by g Calib are the same as those from CALMAR For the children s survey an additional adjustment was made for possible non response bias through weighting The weighting strategy used information about the e sampled address such as type of area e household and adult respondent collected during the main interview and e the age and sex of the selected child again collected during the main interview CHAID Chi Square Automatic Interaction Detector analysis and logistic regression were used to test which variables were most strongly associated with responses and to develop a non response model Once weighting classes were derived non response weights were calculated as the inverse of the probability of response within each class The child response weights were then multiplied by the weights for the household to take into account probability of household selection to give an overall weight for non response at the child and household levels Calibration weights were then generated in the sa
23. l based analysis is carried out on the non victim form to produce incidence and prevalence rates of crime and also when the intention is to make statements about the characteristics attitudes or experiences of children in the sample Analysis of attitudinal questions is individual based as is analysis of victims of personal crimes such as assault All individual based analysis should be weighted by cindivwgt 4 2 Incident based analysis Incident based analysis is carried out on the victim form when the intention is to make statements about characteristics of incidents of crime such as the timing location or perceived seriousness of offences All incident based analysis should be weighted by cweighti Although incidents occurring outside of England and Wales are given a valid offence code only those incidents which occurred within England and Wales should be retained This should be done by selecting cases based on responses to the variable cwherhap 4 3 Incidence rates For personal crime incidence rates give the number of crimes experienced per 1 000 children aged 10 to 15 from 2011 12 onwards per 10 000 children in 2009 10 and 2010 11 To calculate an incident rate the number of incidents experienced by respondents is aggregated together for each crime type This is either one incident per victim module or where one of these victim modules contains a series it is the number of incidents in the series cap
24. me way as they are for the main survey The individual design weights are calibrated to produce the individual weight cindivwgt which is used for individual based analysis attitudinal questions and estimates of personal crime rates and the incident weight cweighti which is used for incident based analysis 4 7 Statistical significance CSEW estimates are based on a representative sample of children aged 10 to 15 resident in households in England and Wales A sample as used in the CSEW children s survey is a small scale representation of the population from which it is drawn Any sample survey may produce estimates that differ from the figures that would have been obtained if the whole population had been interviewed The size of this difference depends on the 13 sample size the size and variability of the estimate and the design of the survey The number of cases upon which analysis is based is important as it influences the precision standard error of the estimates CSEW estimates where the unweighted base is less than 50 cases are not generally published Because of this variation changes in estimates between survey years or between population subgroups may occur by chance In other words the change may simply be due which children were randomly selected for interview It is possible to measure whether this is likely to be the case using standard statistical tests and conclude whether differences are likely to be due
25. n are at a higher risk of victimisation than adults but that they are also frequently the victim of low level incidents It may well be that given the number of incidents that children are likely to be victims of that they may not be able to recall all of these incidents when asked 6 3 Unwillingness to report Respondents may be unwilling to disclose victimisation experiences in a face to face interview setting The circumstances of the interview may have a greater influence on the children s survey where the child s parent or guardian has to be present in the home in order that the survey can be 16 completed such is the mode of administration of the CSEW children s survey that estimates of sexual abuse cannot be obtained using the household survey design Self completion modules are used to collect sensitive information which alleviates this problem to some extent 6 4 Definitions of crime Incidents which are legally offences may not be reported to the survey if the respondent does not view them as such In an attempt to overcome this problem the CSEW screener questions ask whether the respondent has experienced certain types of events They do not refer to crimes use legal terminology or refer to specific offences The definitional problem is particularly relevant to minor incidents and some forms of violence Moreover different social groups may have different perceptions of what does and does not constitute an incident 1
26. nalysis However it is important that anyone undertaking analysis understands the structure of the data Listed below are some general points about the data and how Office for National Statistics ONS analysts use it 3 1 Datasets available End user License analysts can download the following CSEW children s datasets from the UK Data Service e Child non victim form and e Child victim form The victim form dataset contains information collected from the full and mini victim modules long and shortened victimisation modules The non victim form dataset contains information collected from the rest of the interview excluding information from victimisation modules Each case on the non victim form refers to an individual respondent whereas each case on the victim form refers to an individual incident reported by a respondent Each individual respondent has a unique case identifier rowlabel consisting of an eight digit number This identifier is the same on each data file on which information is held about the respondent and allows files to be combined by matching on this variable The children s dataset includes key demographic information about the household from the adult survey But it is also possible to match the CSEW children s dataset with the adult respondent in the same household from the CSEW adult survey Data from the self completion modules and some low level geographic variables have been removed from these data
27. name are the root of the variable name for that group This is followed by _i to give the variable name for the incidence rate for that group and _p to give the variable name for the prevalence rate for that group In addition because there are two different approaches to classifying incidents reported by children these classifications are differentiated by the suffix _brd broad measure and _prf preferred measure For example the name of the variable for the offence groups from which to calculate incidence or prevalence rates for wounding is cwounding The variable from which to calculate rates of incidence for wounding is cwounding_i However there are two different approaches to calculating incidence rates for this offence group each of which can be obtained by using the variables cwounding_i_brd broad measure and cwounding_i_prf preferred measure Non core offence codes are only used within the 2009 10 and 2010 11 datasets Offence groups for 10 to 15 year olds Snatch theft from the person csnatch 43 Snatch theft from the person Stealth theft from the person cstealth 44 Other theft from the person 45 Attempted theft from the person Theft from the person ctheftper 43 Snatch theft from the person 44 Other theft from the person 45 Attempted theft from the person For the broad measure approach to classifying incidents reported by children aged 10 to
28. nd criminal justice system and perceptions of crime and anti social behaviour The CSEW is a large nationally representative sample survey of the population resident in households in England and Wales For the crime types and population it covers the CSEW provides a better reflection of the extent of crime than police recorded statistics because it includes crimes that are not reported to or recorded by the police Following recommendations in two related reviews of crime statistics Smith 2006 Statistics Commission 2006 the CSEW was extended to children aged 10 to 15 from January 2009 Before extending the survey the Home Office who held responsibility for the CSEW prior to April 2012 commissioned methodological advice on the feasibility of extending the survey to both children and those living in group residences Pickering et al 2008 and subsequently undertook a National Statistics consultation following the two reviews on plans to extend the survey to children Following the consultation an extensive period of development and testing work was undertaken throughout 2008 before live data collection started in January 2009 Fitzpatrick et al 2010 The primary aim of extending the CSEW to children was to provide estimates of the levels of crime experienced by children and their risk of personal victimisation The survey also provides a rich source of data to assist in understanding the nature and circumstances of crimes experienced
29. ode frames up to 7 and 99 for code frames up to 97 In most Home Office ONS analysis refusal codes are excluded Don t know codes are also usually excluded unless there is interest in these responses for example in the case of attitudinal questions After the survey went live in January 2009 and during the testing of the existing CSEW Offence Coding system with the children s data some minor changes were made to some items in the questionnaire questionnaire routing changes for example For some cases where this resulted in it not being possible to obtain valid data for some questionnaire items due to these changes responses for these items have been coded as 1 10 Chapter 4 Analysis of CSEW 10 to 15s survey data There are two main types of analysis that can be carried out using Crime Survey for England and Wales CSEW children s victimisation data individual based analysis and incident based analysis Each of these types of analysis needs to take into account the appropriate weighting of the data It should be noted that methodological differences between the adult and child survey mean that direct comparisons cannot be made between estimates obtained from each data source However while the questions asked and levels of detail collected in the children s survey differ from the adult survey there is a similar approach to the coding of incidents to offences in law 4 1 Individual based analysis Individua
30. oluntary survey the 10 15s CSEW is subject to non response error Non response has implications for the measurement of crime if non respondents have different experiences of victimisation to respondents As well as using the same calibration weighting procedure to account for differing rates of non response between people of different sex ages and regions amongst adult respondents see the Weighting section in Chapter 4 the CSEW children s survey also incorporated additional adjustment for non response by using a model based on case level response propensities derived from auxiliary data available about each sampled address in the children s survey through for example the age and sex of the selected child respondent To account for non response all CSEW analysis should be appropriately weighted 6 2 Recall The CSEW asks respondents to recall their experiences of crime in the previous 12 months The CSEW measure of crime is thus dependent on respondents ability to accurately remember their experiences in the reference period The accuracy of CSEW estimates could be affected by recall if a respondent simply forgets a relevant incident reports an incident that occurred outside the reference period as having happened within the reference period or fails to report an incident that occurred within the reference period because they thought it happened outside the reference period Furthermore previous research suggests that childre
31. ped at five The number of incidents for each respondent is then multiplied by 1 000 or 10 000 in 2009 10 and 2010 11 and added to the non victim form A mean of this number produces the incidence rate for a particular offence type See Appendix 2 for an explanation of the variables used to calculate incidence rates Incidence variables for each of the offence groups are included on the non victim form dataset and the variable names include the suffix _i_brd or _i_prf For the children aged 10 to 15 survey respondents completed a maximum of up to four full or six mini victim modules 11 Example producing rates per 1 000 children aged 10 to 15 On a dataset of 4 000 respondents if 100 respondents have each reported 3 incidents of assault the total number of assaults is 300 When this is multiplied by 1 000 the mean value across the whole dataset is 75 300 000 4 000 This means that there are 75 assaults per 1 000 children aged 10 to 15 The overall number of incidents can be estimated for England and Wales based on the incident rate and using estimates of the populations of children aged 10 to 15 in England and Wales For the financial year April 2011 to March 2012 the population estimate of 3 753 871 children aged 10 to 15 was used to provide estimates of the number of incidents 4 4 Prevalence rates Prevalence rates give the proportion of the population who were victims of an offence once or more Unlike inciden
32. pproaches Broad measure formerly All in law The Broad measure formerly known as the All in law approach is the widest possible count but will include minor offences between children and family members that would not normally be treated as criminal matters Preferred measure formerly Norms based The Preferred measure formerly known as the Norms based approach is a more focused method which takes into account factors identified as important in determining the severity of an incident but will still include incidents of a serious nature even if they took place at school These factors were identified from the findings of qualitative research with children that informed the development of the children s survey Unspecified offences do not fall within the scope of the Preferred measure because detailed information on the full victim form was not collected Development of the Preferred measure approach to producing a total count of crime from the incidents reported by children is described in detail in Chapter 8 of Fitzpatrick et al 2010 The remaining two approaches All in law outside school and Victim perceived may still be derived by users based on existing variables on the dataset Chapter 3 CSEW 10 to 15s survey dataset The Crime Survey for England and Wales CSEW children aged 10 to 15 dataset is large and complex but provides a rich source of data for a
33. sets Analysts who need to access these data for their research can request access under the terms of the ONS Approved Researcher method through the UK Data Service These requests will be sent to the ONS for approval The datasets available under this procedure are e Child self completion dataset and e Low level geographic variables 3 2 Variable names CSEW variable names can be found in the questionnaire alongside the question to which they refer This includes responses to questions about bullying street gangs carrying knives truancy exclusion from school cannabis and alcohol use 3 3 Derived variables In addition to the questions directly asked of respondents the CSEW data files also include derived variables There are two sets of derived variables those produced by the survey contractor can be found in TNS BMRB 2012 and those produced by Home Office ONS analysts The derived variables produced by Home Office ONS analysts have been included on the children s datasets so researchers can replicate the analysis produced in various publications 3 4 Don t know and refusal codes Respondents are not usually explicitly given the options don t know or refusal However for every question respondents may say they do not know or refuse to answer and these are valid responses The code for refusal is 8 for code frames up to 7 and 98 for code frames up to 97 The code for don t know is 9 for c
34. tation of being able to code incidents to offences on a broadly comparable basis Like the main survey the children aged 10 to 15 survey excludes crimes often termed as victimless for example possession of drugs and as a survey that asks people whether they have experienced victimisation homicides cannot be included The children aged 10 to 15 survey was not designed to collect information on sexual offences the small numbers of sexual offences reported by respondents were referred to the Home Office Development and testing work showed it was necessary to adapt the existing questions asked of adults to make them suitable for children In addition to reduce respondent burden and to reflect that some incidents reported by children may be considered relatively minor in 2009 10 and 2010 11 detailed information about an incident was not collected if e the incident happened at school and e the offender was a pupil at the victim s school and e the offender did not use a weapon and e the victim was not physically hurt in any way This was to reduce respondent burden and to reflect that some incidents reported by children may be considered relatively minor Incidents which met these criteria had a limited amount of information collected to enable classification to a high level crime category and so it was not possible to assign specific offence codes within the appropriate high level classification according to standard CSEW proc
35. to chance or represent a real difference Analysts should be familiar with how to carry out such tests before conducting analysis on the CSEW children s dataset However please note that no statistical assessment of the change in levels of victimisation between the 2009 10 2010 11 and 2011 12 CSEW has currently been made This is due to the developmental nature of the survey and estimates which have been subject to further refinement during the periods presented 14 Chapter 5 Published data from the CSEW 10 to 15s survey Due to refinements as part of the development of the survey and presentation of estimates there have been some amendments to the presentation of figures from the 2009 10 and 2010 11 Crime Surveys for England and Wales CSEW Chaplin et al 2011 compared with the first experimental statistics Millard and Flatley 2010 For an incident to be a crime in law there must be an element of intent apparent For example for an incident to be considered theft there must be intent to permanently deprive someone of their property The information captured around intent has improved as the survey has developed due to increased focus through questions and interviewer and coder training Published figures from the 2009 10 and 2010 11 CSEW exclude incidents where the respondent perceived no intent from the perpetrator to hurt steal or damage Offences of Theft from a dwelling Theft from outside a dwelling
Download Pdf Manuals
Related Search