RESEARCH DISCIPLINE
Criminology Research Project Analysis
Self-Report Data, Read With Care
For victimization survey analysis and self-report studies of offending: reference periods and sensitive items weighed, skewed frequencies modeled sensibly, and the share who offend kept apart from how often they do in the results.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Does Survey-Based Criminology Analysis Involve?
Criminology research project analysis of survey data starts from what respondents report about their own offending or about being victimized, which is not the same as what police records contain. It covers how reference periods, sensitive questions and nonresponse shape the answers, how scales and frequency counts are built, and how models respect clustering in schools or neighborhoods and repeated waves of the same respondents. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, supports doctoral candidates and research teams in criminology: an in-house PhD-level statistician works through self-report offending survey analysis and victimization data analysis on your file and reports what the answers can support. The theory you test and the conclusions you draw are your own.
WHEN TO ENGAGE
Phases of a Self-Report Study
A survey of offending or victimization is shaped by choices made before anyone answers, and the statistical questions change as the project moves from instrument to model.
Instrument Still Open Items on offending or victimization are being finalized, and the reference period, the response format and the plan for scoring frequency are fixed while changes are still possible.
Data Collected A school, household or online sample has responded, and skipped sensitive items, straight-lining and refusals are examined before scales or outcomes are constructed.
Outcome to Model Whether anyone offended, how often offenders did so, or a victimization count is the outcome, and the model must match its shape and the grouping of respondents in classes or areas.
Panel Waves Compared The same respondents were surveyed more than once, and attrition, changing reference periods and change within persons are handled before development over time is described.
WHAT THE SUPPORT COVERS
Scoring, Modeling and Reporting Self-Reports
Raw answers about offending only become outcomes once they are scored, and the support follows them from the questionnaire to the published estimate.
Item and Scale Scoring
Offense items are combined into any-offending indicators or variety scores, and the reliability and dimensionality of each scale are checked in your sample before it is used.
Prevalence Outcomes
Whether anyone offended or was victimized is modeled with Logistic Regression, with predicted probabilities shown beside the odds ratios.
Frequency Outcomes
Skewed frequencies of offending or victimization, with many zeros and a long tail, are modeled with a count or two-part model and not summarized by a plain average alone.
Clustered Respondents
Pupils within schools and residents within neighborhoods are modeled with random effects, as in Linear Mixed Models (Multilevel Modeling), extended to binary and count outcomes where needed.
Sensitive Items and Nonresponse
Skipped and refused answers on sensitive questions are described by group, with the assumptions made about them stated, since those who skip may differ from those who answer.
Reporting Survey Results
Weights and design features are declared and findings carry intervals, as in published Criminology & Criminal Justice studies, a field that sits under Education & Social Sciences.
COMMON ISSUES
Ways Self-Report Data Mislead
Self-report measures are not a direct count of behavior, so an estimate of offending or of victimization can be bent by any of these problems.
Reference Period Mismatch Respondents are asked about lifetime behavior in one wave and about the last twelve months in another, and a difference between waves is read as a change in behavior.
Long Tails Averaged Mean offenses per respondent are driven by a few who report very high frequencies, so a handful of extreme answers moves a comparison between groups.
Series Events Counted Loosely Repeated similar victimizations that a respondent cannot recall separately may be capped or counted as one under a survey's coding convention, as in NCVS data analysis, so victimization rates depend on a rule that must be stated.
Schools Treated as Independent Pupils in one school resemble each other, so ignoring the school makes intervals too narrow and school-level differences look like individual ones.
Silence Taken as Innocence Skipped sensitive items are coded as no offending, so nonresponse becomes a finding and groups with higher refusal appear less involved.
Panel Dropout Ignored Respondents who offend more may leave a panel more often, so an analysis of those who remain can understate both the level and the change.
DELIVERABLES
Files for a Self-Report Survey Analysis
Survey analyses need a trail from the questionnaire to the table, so the files record how each score and each estimate was derived from the raw answers.
Variable Construction Log How each offending, victimization and covariate variable was derived from the raw items, including reference periods, recodes and the treatment of refused answers.
Scale Properties Report Reliability and structure of each multi-item scale in your sample, with the decision on whether to use a total, a subscale or a variety score.
Prevalence and Frequency Tables Weighted prevalence and frequency by group with 95% confidence intervals, with the reference period stated beside each figure.
Clustered Model Results Estimates from models that allow for respondents grouped in schools or areas, with predicted probabilities or counts in plain terms.
Nonresponse Summary A description of who skipped which sensitive items and a check of how far the results change under other assumptions about them.
Methods Documentation Note A note recording the instrument, the sampling and the analysis, with the scripts that regenerate each table, so the work can be described accurately.
ILLUSTRATIVE EXAMPLE
Victimization and Later Delinquency in Schools
An illustration of how a two-wave school survey is analyzed when respondents are grouped in schools and some leave between waves.
ILLUSTRATIVE EXAMPLE · SCHOOL CLIMATE AND DELINQUENCY SURVEY · UNITED STATES
A doctoral candidate in criminology wanted to know whether adolescents who reported being victimized by peers were more likely to report delinquent acts a year later, using a two-wave survey across a number of schools.
The plan fixed the outcome definitions, the reference period and the covariates before wave two was examined, and treated any-delinquency and frequency of delinquency as separate outcomes. The complication was attrition: adolescents who reported more delinquency at wave one were less likely to complete wave two. Completers and non-completers were compared on wave one measures, a sensitivity analysis explored how the estimate changed under other assumptions about those lost, and the models allowed for pupils within schools. The association was described as such, since victimization and delinquency may share causes that the survey did not measure.
Outcome: The candidate received the adjusted odds ratio for later delinquency with its 95% confidence interval, predicted probabilities, the attrition comparison, the scale report and the scripts.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Software for Survey Files on Offending
The package follows the survey file and your department's preferences, and the syntax or scripts are supplied so that each table can be regenerated.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
RESEARCH RESPONSIBILITY
Your Theory, Our Statistical Treatment
OnlineSPSS carries out the scoring and modeling agreed in the scope, and the criminological argument built on the results belongs to the researcher.
Agreed Statistical Support OnlineSPSS provides the agreed methodological, analytical and statistical reporting support, defined in a scope that is confirmed before any work begins.
No Intellectual Contribution OnlineSPSS does not claim any intellectual contribution to your research. Authorship of the thesis, report or article remains yours.
Your Research Decisions The research question, the theoretical framing, the substantive interpretation, the discussion and the conclusions remain yours, with overall responsibility.
Confidential Handling Research data and project details are handled confidentially, project information is not published, and an NDA is available where required.
FREQUENTLY ASKED QUESTIONS
Questions About Analyzing Offending and Victimization Surveys
Self-report surveys depend on who is asked and what they are willing to say, which is where these answers begin.
What does the analysis of a self-report criminology survey cover?
Victimization survey analysis at OnlineSPSS covers survey-based studies of victimization and self-reported offending: scale construction, prevalence and frequency outcomes, models for clustered or repeated respondents and the reporting of results. It suits dissertations, evaluations and articles, and the theory and conclusions remain with the researcher.
Can a school-based survey describe offending among all young people?
Only among those enrolled and present on the survey day. Young people who are often absent, excluded or no longer in school have been found to report more offending when reached by other means, so a school sample can understate offending in the age group as a whole. The population the sample represents is stated, and household or justice-system samples are used for comparison where they exist.
Should offending be analyzed as a count or as yes or no?
Both can be informative. Whether a respondent offended at all and how often are different questions, and the factors linked to each may differ. A binary model answers the first, a count or two-part model the second, and reporting both prevents a few high-frequency reporters from hiding the pattern among everyone else.
What if many respondents skip the sensitive questions?
Skipping is itself informative. The share who skip is reported by group, those who answered are compared with those who did not on other variables, and results are checked under alternative assumptions, such as imputing the skipped items under a stated model, so that a reader sees how far the conclusion depends on them.
How do you protect my research data and confidentiality?
Your data and project details are handled confidentially, and project information is not published. Researcher names and institutions are not used publicly without permission. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, such as the questionnaire, a codebook or the survey dataset, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Self-Reports Deserve
a Second Reading
Self-reported offending reflects the wording of each question as well as the behavior, which is why the instrument is examined before any response is modeled.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014