RESEARCH DISCIPLINE
Criminal Justice Research Project Statistics
Counts, Rates, Repeat Contacts
Court and corrections data analysis, alongside the arrest records that precede it: the unit of analysis fixed, counts turned into rates with a denominator that reflects who was at risk, and repeated contacts modeled so that each person is not counted many times over.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Does Statistical Work on Justice System Records Involve?
Criminal justice research project statistics concerns the records the justice system generates about its own work, such as arrests, bookings, court events and periods of supervision, and the counts and rates built from them. It covers deciding whether the unit is the event, the person or the place, choosing a denominator that reflects who was at risk, and modeling repeated contacts without treating each row as an independent case. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, supports graduate researchers and agency analysts: an in-house PhD-level statistician carries out policing data analysis on arrest and booking files and sentencing data analysis on court files, and is plain about the distance between recorded contacts and the crime behind them. The policy reading is yours to give.
WHEN TO ENGAGE
From Records Request to Agency Query
Questions about the unit of analysis come first, questions about the model come next, and questions from an agency or a reviewer arrive last.
Records Being Requested An agency has agreed to share records, and the variables requested, the time window and the level of aggregation are decided so that the extract can answer the research question.
File With Repeated Rows The extract holds one row per arrest or court event, and it must be decided whether to analyze events, people or places before any count or rate is built.
Rates to Compare Jurisdictions, years or neighborhoods are to be compared, and the population at risk, which may differ from the resident population, is chosen as the denominator.
Agency or Reviewer Query A partner agency or reviewer asks why a rate changed or whether a program reduced contacts, and the figures are examined for shifts in recording or enforcement before an explanation is offered.
WHAT THE SUPPORT COVERS
Turning Case Records Into Research Data
Administrative records are built to process cases, not to answer research questions, so much of the work lies in reshaping the file before any rate is computed.
File Structure
Each record type is mapped to its unit, and identifiers are checked so that one person is recognized across agencies and years and is not counted as several.
Unit of Analysis
Events are collapsed to person-level or place-level files where the question requires it, and the rule for choosing the unit is documented with its consequences.
Rate Denominators
Rates carry a stated denominator, such as residents, licensed drivers or arrests, and the choice is justified because a place with many visitors can show a rate that residents alone cannot explain.
Count Models
Counts that vary more than a simple model allows are analyzed with Negative Binomial Regression, with Poisson Regression as the baseline for comparison.
Repeated Contacts
People with many contacts are modeled as clusters, or their contacts are counted per person over a fixed window, so that heavy repeat contacts are not treated as independent observations.
Reporting With Limits Stated
Results are reported with intervals and the limits of the records stated, following the practice of Criminology & Criminal Justice research, within the broader field of Education & Social Sciences.
COMMON ISSUES
Misreadings of Arrest and Court Records
A record exists because an official acted, not only because something happened, and each reading here turns that fact into a wrong conclusion.
Charges Counted as Offenses Several charges filed for one incident are counted as several offenses, so offense totals rise and fall with charging practice, which differs between prosecutors and over time.
Rates With the Wrong Base A rate per resident is computed for an area with many commuters, tourists or a large jail population, and the area looks unusually high or low for reasons unrelated to risk.
Heavy Users Dominate A small group of people with many contacts supplies a large share of the records, and a model that treats rows as independent understates uncertainty and gives these people outsized influence.
Unequal Time Observed People are compared on contacts without allowing for how long each was observed, so those released earlier or entering the file younger have more opportunity to appear.
No Record, No Offending A person or place with no record is treated as having no offending, when no record can mean no contact, no report or a record held by another agency.
Program Effect From Raw Change Contacts fall in the year after a program and the fall is attributed to it, though the earlier trend, the selection of participants and a return from an unusual year are not examined.
DELIVERABLES
Files That Document the Records Analysis
A records analysis is only as credible as its account of the file, so these materials set out how the file was read and how each rate was built.
Record Layout Map A map of each record type with its identifier, unit and dates, and the rules used to link records across agencies or years.
Person-Level File and Codebook The cleaned file at the chosen unit, with a codebook of offense groupings, date windows and the exclusions applied, each with its reason.
Rate Construction Sheet Each rate with its numerator, its denominator and the population it refers to, so that another analyst can rebuild the figures from the same source.
Count Model Results Model estimates with 95% confidence intervals that allow for repeated contacts, shown with predicted counts in terms an agency can use.
Window and Definition Checks Results under different follow-up windows and offense groupings, showing how far each conclusion depends on the choices that were made.
Records Limitation Note A plain statement of what the records measure and what they miss, with the scripts needed to regenerate every table and figure.
ILLUSTRATIVE EXAMPLE
Bookings After a Diversion Program
An illustration of how repeated contacts in a county jail file are analyzed when people were observed for different lengths of time.
ILLUSTRATIVE EXAMPLE · JAIL BOOKING DATA RESEARCH · UNITED STATES
A graduate researcher in criminal justice wanted to know whether participation in a pretrial diversion program was associated with fewer jail bookings in the following twelve months.
The booking file had one row per booking, so it was restructured to one row per person with a count of bookings and the time each person was observed, and the plan fixed the comparison group and covariates before outcomes were examined. The complication was twofold: a few people accounted for many bookings, and bookings in neighboring counties were not in the file. The count model allowed for extra variation and unequal observation time, a sensitivity analysis set aside the most frequent contacts, and the missing out-of-county bookings were stated as a limit. Because prosecutors chose who entered the program, the result was described as an adjusted association.
Outcome: The researcher received the ratio of booking rates between participants and the comparison group with its 95% confidence interval, the sensitivity analysis, the record layout map, the rate construction sheet 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
Choosing the Package for Records Files
Agency files often arrive as exports from a case management system, and the analysis is set up in the package your department or partner agency already uses.
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
Who Decides What in a Justice Records Project
OnlineSPSS supports the statistical handling of the records, and the legal, policy and theoretical reading of the results is left 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 report, thesis or article remains yours.
Your Research Decisions The research question, the criminological interpretation, the policy implications, 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 Justice System Records
Agency records were never designed for research, and most questions about them come back to what was counted and who could be counted.
What does statistical analysis of justice system records cover?
Court and corrections data analysis covers the records generated by police, courts and corrections: the unit of analysis, counts and rates, models for repeated contacts and a statement of what the records can support. It suits dissertations, agency evaluations and journal articles built on administrative data.
Should I analyze events, people or places?
The unit follows the research question. Events suit questions about case processing, people suit questions about who is reached and what happens afterward, and places suit questions about where. Whichever is chosen, the other levels stay in the structure, since events cluster within people and people within places, which affects the standard errors.
What denominator should be used for a crime rate?
The population that was at risk of the event. Residents are the common choice but can mislead where many commuters, visitors or institutionalized people are present. Licensed drivers, housing units or arrests suit particular questions, and the estimate is always described in terms of the denominator chosen.
Can one agency's records show what happened after release?
Only for contacts that agency recorded. A person may be arrested or sentenced in another jurisdiction, so later outcomes can look rarer than they were. Linking records across agencies, where it is permitted, narrows the gap, and where it is not possible the limitation is stated with the results.
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 a proposal, the record layout or a data dictionary for the extract, 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
Which Records,
and Which Comparison?
Behind every crime rate sits a choice of denominator, and the analysis gives that choice as much scrutiny as the count.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014