I really want criminal justice involvement, including events aside from incarceration lengths, criminal charges, and length of time on probation or parole. I want to include arrests and contact without arrest.
I am doing a time series and I want to compare low and high criminal justice involvement (arrests, contact without arrest, convictions, incarceration lengths, amount of incarceration, and time on probation/parole). Thus far I only have found studies who talk about "more or less criminal justice involvement" but they don't describe how they calculated this, if it is in fact a value. What I think is actually happening, is that they are talking about beta estimates from regressions, and not a value. Criminal history scores are used in the US in our sentencing guidelines. So someone that committed an armed robbery and a lengthy criminal history should get more time than someone who committed the same crime and only had been arrested once before. However, trying to calculate the score can be timely, and I am going to have to do it for 150 participants, and I wanted to do quota sampling for this population, but I can't really do it before I do their interview. That is why I want to avoid using the criminal history score that I mentioned in my post if there is another way that I just haven't come across. Or another idea.
Two well-validated measures of "official" criminal history, i.e., resulting in charges, were developed as part of the Violence Risk Appraisal Guide and the Sex Offender Risk Appraisal Guide.
These are the Cormier-Lang measures, one for nonviolent offenses and the other for violent offenses. Scores reflect both number and severity of offenses. You can find articles that use these measures in a Google Scholar search; the scoring instructions can be found in the 2006 book by Quinsey et al., published by the American Psychological Association.
The Cormier-Lang measures were not designed to include self-reported offenses or offenses that did not result in criminal charges, so you would have to see if you are willing to modify them to capture the "non-official" criminal histories as well.
I forgot to add that the Cormier-Lang was built on the Canadian Criminal Code, which differs in crimes and distinctions from state and federal laws in the US. But it can be adapted by scoring the equivalent, e.g., battery is an assault charge in Canada.
There have been American studies using the Cormier-Lang.