IJE Advance Access published online on March 24, 2008
International Journal of Epidemiology, doi:10.1093/ije/dyn054
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Letter to the Editor |
Author's Response
1Institute for Therapy and Health Research, Düsternbrooker Weg 2, 24105 Kiel, Germany.
2Dartmouth Medical School, Lebanon, NH 03756, USA.
3Dartmouth Medical School, Director, Cancer Control Research Program Norris Cotton Cancer Center, Lebanon, NH 03756, USA.
Corresponding author. Institute for Therapy and Health Research (IFT-Nord), Düsternbrooker Weg 2, 24105 Kiel, Germany. E-mail: hanewinkel{at}ift-nord.de
We would like to thank Ms Naila Dinani, Ms Nicola R Wood and Dr Iain J Robbé for their interest in our article on alcohol use in motion pictures and teen alcohol use.1 They ask important questions and prompt us to provide some further information on our study.
Dinani et al. voiced concerns about the sampling bias at the levels of school and student recruitment, specifically that 36% of schools did not participate and that, among schools that did, 12.7% of the parents did not give permission for their sons or daughters to take part in the survey. School non-participation relates more to the views of the school administrators and school board members than the characteristics of the pupils in our experience. The state Schleswig-Holstein, where the study took place, has a complex school system with four different kinds of schools; as mentioned in the article, school type is a marker of socioeconomic status. The recruited school sample did not differ from the distribution of schools within Schleswig-Holstein as a whole; this suggests that the selection process did not bias the sample from a socioeconomic standpoint.
According to German law, active parental consent was required for study participation. We achieved active parental consent for 87.3% of the parents, which compares very well with other school-based survey studies.2,3 The key question raised by Dinani et al. is whether children whose parents did not give permission were different from those with permission with respect to their exposure and response to movie alcohol use, and to what extent this may have altered our findings. Since we were not able to gather data on students without parent consent, we are unable to answer the question directly. Although, Dinani et al. do not propose how students might differ, we suggest the following hypothesis: willingness of a parent to give consent for a survey is related to parent characteristics that also have a strong main effect on alcohol use. We assessed parenting style, the student's perception of mother's warmth/responsiveness and supervision/demandingness; neither of these were strongly linked with alcohol use. Furthermore, we found no evidence that parenting style affected the students’ response to movie alcohol use (no interaction effect). This makes us less concerned that the exclusion of these youth would nullify our findings.
Dinani et al. also ask why we include movies that were released when our subjects were pre-schoolers. If adolescents were viewing their movies only at the box office, this sampling strategy would clearly be flawed. However, our data indicate that only a small proportion of movies are viewed at the box office; instead, most are viewed on television or through DVD, where access is without regard to theatre release date. To illustrate this point, we aggregated responses of adolescents to movie title queries by year of theatrical release (shown in Table 1). Recent releases were seen by about half of the adolescents, but movies released as far back as 10 years prior to the survey were still seen by about a quarter of adolescents. Adolescents reported viewing 44.5% of the titles for 2001; this viewing rate dropped for movies released after 2001 because we sampled movies from the top 100 box office hits for those most recent years, including titles not popular among adolescents. Table 1 provides our justification for including older box office hits in the parent sample of movies.
|
We have trouble understanding why Dinani et al. call for us to account for whether movie alcohol use was depicted in a positive or negative manner, or in what context alcohol was consumed. These questions are very interesting but do not detract from our main effects finding and are probably better answered by an experimental study where movie exposure is presented in a controlled fashion.
Although, one would always like to have verified self-reports, biochemical validation is not possible for adolescent alcohol use. Considerable methodological research has verified the validity of adolescents' reports of substance use and other problem behaviours. This research has consistently shown that when confidentiality of responding is assured, adolescents' reports of tobacco and alcohol use have high validity; validity is indicated by convergence with biochemical measures and collateral reports.4–7 We assured confidentiality for the adolescents in this study. Similarly, findings from adolescent reports on predictor variables such as self-control and academic competence have been corroborated consistently by collateral reports, such as teacher ratings.8,9 Lintonen et al.10 recently studied reliability of adolescents' self-reported drinking and perceived drunkenness in surveys. The authors concluded that the adolescent drinking self-reports are reliable and valid both on a population and individual level, and that a set of closed questions, comparable with the questions used in this study, may capture the amount drunk even better than an open question. As Patrick et al.6 note, there may be situations where one should be suspicious of self-reports of drug use (e.g. in a treatment setting where detected use could result in negative consequences for the patient), but in epidemiologic research where confidentiality is assured the literature consistently supports the validity of adolescents' self-reports.
We agree with Dinani et al. that a causal relationship between the exposure to movie alcohol use and alcohol use initiation cannot be drawn from a cross-sectional study, and think we were appropriately cautious in our discussion. To investigate whether seeing alcohol use in movies precedes alcohol initiation we have followed-up our sample and will publish these data in due course. Notably, another study has shown that that baseline exposure to movie alcohol use predicts taking up alcohol use at follow-up when never drinkers are followed longitudinally, and that the association was very similar to the association seen in cross-sectional studies.11
References
1 Hanewinkel R, Tanski SE, Sargent JD. Exposure to alcohol use in motion pictures and teen drinking in Germany. Int J Epidemiol (2007) 36:1068-a–77.
2 Ji PY, Pokorny SB, Jason LA. Factors influencing middle and high schools' active parental consent return rates. Eval Rev (2004) 28:578–91.
3 McMorris BJ, Clements J, Evans-Whipp T, et al. A comparison of methods to obtain active parental consent for an international student survey. Eval Rev (2004) 28:64–83.
4 Murray DM, O’Connell CM, Schmid LA, Perry CL. The validity of smoking self-reports by adolescents: a reexamination of the bogus pipeline procedure. Addict Behav (1987) 12:7–15.[CrossRef][Web of Science][Medline]
5 Murray DM, Perry CL. The measurement of substance use among adolescents: when is the ‘bogus pipeline’ method needed? Addict Behav (1987) 12:225–33.[CrossRef][Web of Science][Medline]
6 Patrick DL, Cheadle A, Thompson DC, Diehr P, Koepsell T, Kinne S. The validity of self-reported smoking: a review and meta-analysis. Am J Public Health (1994) 84:1086–93.
7 Wills TA, Cleary SD. The validity of self-reports of smoking: analyses by race/ethnicity in a school sample of urban adolescents. Am J Public Health (1997) 87:56–61.
8 Wills TA, DuHamel K, Vaccaro D. Activity and mood temperament as predictors of adolescent substance use: test of a self-regulation mediational model. J Pers Soc Psychol (1995) 68:901–16.[CrossRef][Web of Science][Medline]
9 Wills TA, Cleary S, Filer M, Shinar O, Mariani J, Spera K. Temperament related to early-onset substance use: test of a developmental model. Prev Sci (2001) 2:145–63.[CrossRef][Medline]
10 Lintonen T, Ahlstrom S, Metso L. The reliability of self-reported drinking in adolescence. Alcohol Alcohol (2004) 39:362–68.
11 Sargent JD, Wills TA, Stoolmiller M, Gibson J, Gibbons FX. Alcohol use in motion pictures and its relation with early-onset teen drinking. J Stud Alcohol (2006) 67:54–65.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||