How were the 'ALICE' schools selected?
The sample of 'ALICE' schools was chosen to be broadly representative of Scottish secondary school pupils, accessible to the Glasgow-based survey team and not already overburdened with requests to take part in research studies. The schools sample was therefore drawn from two education departments representing smaller Scottish central belt urban areas outwith Glasgow, Edinburgh and Stirling.
Sample size calculations for the European 'Smoking in Movies' project suggested that around 13,800 adolescents (2,300 from each of the six participating coutires) were needed. This sample size would ensure that each national sample would be large enough to show an effect of seeing smoking in films on smoking uptake in early adolescence (when many young people begin smoking), accounting for estimated losses between the baseline and follow-up surveys.
The sample of 'ALICE' schools from the two selected education departments took into account information on:
- The total number of pupils in every Scottish secondary school, and the proportion of secondary school pupils in each of the school years S1-S3 (available from the most recently available [2007-8] Scottish Government 'Pupils in Scotland' statistics). This allowed us to calculate what total school roll would be needed to obtain a sample of 2,300.
- Data on school roll, 5th year staying on rates, absences and free school meals for schools in the two selected education departments (available via Scottish Schools Online).
The final sample comprised seven schools with a larger total roll than we were likely to require, on the basis that not all might agree to take part. In fact, all the schools did agree, so the final pupil sample was somewhat larger than the 2,300 required for the European smoking in the movies study
Within the seven selected schools, all S1, S2 and S3 pupils were invited to participate in early 2010. Around a year later, the same pupils (now in S2, S3 and S4) were invited to take part in the follow-up survey.
How were data collected?
Pupils completed both the baseline and follow-up questionnaires during school lesson periods. In all schools, questionnaires were completed in as near as possible exam conditions. In some schools, whole year groups were surveyed together in the school hall or gym, in others the surveys took place on a class-by-class basis, depending on the preference of each participating school.
All surveys were staffed by a team leader and several survey assistants (at least one for every ten pupils). All survey staff had enhanced disclosure. Most teachers left the classroom during the survey or worked quietly in the corner. No teacher had access to the completed questionnaire of any pupil.
At the beginning of the session, the team leader introduced the study and talked pupils through completion of consent forms and questionnaire front sheets. These allowed pupils to create individual codes which made it possible for information provided at baseline to be linked with information from the same pupil at one-year follow-up (see below for further details). Survey assistants helped pupils with questionnaire completion as required, ensured good discipline, checked finished questionnaires and sealed them into envelopes.
What data were collected?
The 2010 (baseline) questionnaire included the unique lists of 50 films, so allowing us to estimate the total number of occurrences of smoking and drinking which each respondent had seen. The follow-up (2011) questionnaire focused on the context in which adolescents had seen 20 popular recent films with questions about number of times seen, who with and where.
The questionnaires also included: demographic details; items on TV and film watching and associated parental restrictions; self-descriptions (such as 'I believe in following rules'); smoking; drinking; health and well-being, life at home and in school, consumerism, internet use and friendships.
How many pupils took part?
At the time of the baseline (2010) surveys, there were 3,189 S1-S3 pupils in the seven 'ALICE' schools, 2,937 (92%) of whom completed a questionnaire. Only 11 parents and 15 pupils refused consent, representing less than 1% of the total possible sample. Almost all non-responders were pupils who were absent on the day of the survey.
In 2011, 2,783 pupils completed a questionnaire. Once again, the main reason for not taking part was absence on the day of the survey; only 16 parents and 13 pupils refused consent.
If the questionnaires were anonymous, how were pupils followed up?
Both the baseline and follow-up questionnaires included a detachable front sheet which asked pupils to write in their name, date of birth, and their mother's and father's first names. The pupils themselves then transfered pre-assigned letters or digits (e.g.1st and 3rd letter of their last name etc) from these answers onto the main part of the questionnaire and removed the top sheet with its identifying information.
This method of anonymisation of questionnaire has been successfully used in studies in schools in a number of European countries (Galanti et al 2007).