Endnotes
1. NCVS data provide only limited coverage of domestic violence victimization, including spouse abuse and child abuse. (See the NCVS sidebar).2. Certain problems are not easily categorized as voluntary or involuntary. For example, illicit drug use may be purely voluntary or it may reflect a condition of dependence or addiction that could be regarded as involuntary or only partially voluntary.
3. Plans are currently under way to collect additional NYS data in 2001 and 2002.
4. Pickpocket and purse-snatching victimization were added in wave 3.
5. In earlier waves, the study distinguished between being beaten up by parents and being beaten up by someone else; in later waves, as respondents moved out of their parents' homes, this distinction was dropped in the wording of the questions but retained in the followup questions to the victimization items (i.e., respondents were asked by whom they were victimized, with parents as one possible category of perpetrator).6. The approach described in note 5 for questions about being beaten up was also used for questions about being attacked with a weapon.
7. The nondomestic violent offending scale of the NYS may include some domestic violence offenses, but as indicated in Mihalic and Elliott (1997), domestic violence is seriously underreported when questions ask about general offending and victimization and do not specifically refer to violence involving a spouse or partner.
8. The NYS collected data on domestic violence by using the Conflict Tactics Scale (Straus and Gelles, 1990), which includes a range of behaviors from nonphysical aggression to life-threatening physical violence. This analysis limits its consideration of domestic violence to the most serious forms, i.e., those roughly comparable to the items in the NYS felony assault and violent victimization scales.
9. See note 8. Because few respondents ages 11–17 were married, the analysis of serious domestic violence victimization uses the more general felony assault measure to control for prior violence.
10. The problem drug use scale was not administered to the full NYS sample until wave 4 (1979); therefore, the analysis uses prevalence of marijuana and polydrug use during adolescence as predictors for problem drug use during adulthood.
11. Beginning with the data for 1983, a subset of the Diagnostic Interview Schedule or DIS (Robins et al., 1981) was administered to adult respondents (age 18 and over), specifically to diagnose depression. In the last wave (1992), this subset was expanded to include anxiety, PTSD, obsessive-compulsive behavior, and schizophrenia. (The prevalence of the last two disorders, however, was too small for analysis.) Researchers did not administer the DIS to adolescents, but rather asked them whether they felt socially isolated or alienated from others and whether they thought that others perceived them as sick, disturbed, or in need of help. The mental health problem scale used with adolescents is described in detail by Elliott and colleagues (1989). In the first wave (1976), researchers also asked parents whether they thought their child was sick, disturbed, or in need of help.
12. In the context of adult victimization, adolescent victimization is both a possible predictor and a prior problem condition. The analysis of adult victimization, therefore, should be regarded as simply an examination of the continuity of victimization from adolescence to adulthood.
13. Both prevalence and frequency measures are available in NYS data for adult and adolescent violent and property victimization, violent offending (felony assault), serious property offending (felony theft), marijuana use, and polydrug use. However, only prevalence measures and ordinal, Likert-type scale scores (as described by Elliott, Huizinga, and Menard [1989]) are available for adult domestic violence victimization and offending and for adult and adolescent problem drug use and mental health problems, and the ordinal scale scores do not necessarily correspond to the frequency scales used for the victimization and offending measures.
14. The impacts of adolescent marijuana use on adult polydrug use and of adolescent polydrug use on adult problem drug use appear to reflect developmental patterns—entry into behaviors that are not just risk factors for, but nearly prerequisites to, similar adult behaviors (Elliott, Huizinga, and Menard, 1989).
15. The measure of adult success used in this analysis obviously is a general, summary measure at a fairly minimal level. One alternative would be to analyze adult success separately for the criteria that are the basis of this measure and for other criteria. It would also be possible to measure adult success as a matter of degree, rather than simply as a yes or no outcome. Such analyses are beyond the scope of the present study, however, and are left to future research.
16. The academic and employment variables were added because Elliott and colleagues (forthcoming) found them to be predictive of adult success.
17. On a technical note, models that treated success as both an ordered and an unordered dependent variable were considered, and the model with the best accuracy of prediction (as measured by RL2) was selected for presentation in table 6. There were minor differences in the statistical significance of some predictors in the different models, but these differences are not discussed in detail here and would make little or no difference in the conclusions reached.
| Previous | Contents | Next |
 |
|
| Short- and Long-Term Consequences of Adolescent Victimization |
|