Skip to main content

Table 2 Description of outcome and explanatory variables

From: Poly-substance use and sexual risk behaviours: a cross-sectional comparison of adolescents in mainstream and alternative education settings

Variable

Questions asked

Variable creation details

Primary Outcomes

 Substance Use and Poly-substance

Pupils in both MES and AES were asked three questions about substances: “how often in the last twelve months have you got drunk/ used cannabis, hash, weed or grass /smoked tobacco (cigarettes)?” Answers were given on 6 point scale for alcohol and cannabis use (1 = never, 6 = more than once a week) and 7 point scale for tobacco use (1 = never, 7 = every day). For each substance, responses dichotomised into used once a week or more vs. less often.

A binary variable was created for each substance by dichotomizing responses into “used once a week or more” vs. “less often”.

A binary variable for poly-substance use was created by dichotomizing use of all three substances (alcohol, tobacco and cannabis) into “once a week or more” vs. “less often”. The decision to define poly-substance use as using all three substances was based upon the DSM-IV diagnostic criteria that polysubstance dependence relies on individuals being reliant upon three or more substances [63]

 Has had penetrative sex activity

Pupils in both MES and AES were asked “have you ever had penetrative sex? Response options were “yes” and “no”.

A binary variable was created. Pupils with missing data who had provided answers to subsequent questions about the circumstances surrounding were recoded to “yes” and included in the analysis. Otherwise, missing data was excluded from the analysis.

 Protected against both pregnancy and STIs, either through abstinence or by using condoms or condoms plus another contraceptive at sexual debut or beyond.

Pupils in both MES and AES were asked two questions about being protected against pregnancies and STIs. 1) “when you first had penetrative sex, did you or your partner use protection against pregnancy or sexually transmitted infections?” Responses included: “No”, “Penis pulled out before ejaculation or cumming”, “condom was used”, “I/my partner was on the pill”, “I/my partner used emergency contraception (the morning after pill)”, “other (please write in)” and “don’t know”. 2) “thinking carefully about all the times you have had penetrative sex ever… how often did you or your partner use a condom?” Responses included: “never”, “not very often”, “about half the time”, “most of the time”, “always” and “don’t know”.

A binary variable was created for protection against pregnancy and STIs, either by always using a condom or through abstinence. Pupils who stated that they had used a condom at sexual debut (question 1) and also reported that they had “always” used a condom when having penetrative sex (question 2) were coded as having always been protected against pregnancy and STIs. As were pupils who had answered “no” to the question “have you ever had penetrative sex?” Pupils with missing data for these questions were excluded from the analysis.

Secondary Outcomes

 Sexual debut was planned

Pupils in both MES and AES were asked which of the following statements best described the circumstances surrounding their sexual debut: “I planned it by myself without my partner”, “we planned it together”, “it was completely unexpected” and “I can’t remember”.

Pupils who answered “I planned it by myself without my partner” or “we planned it together” were coded as having had a planned sexual debut. Those who answered “it just happened on the spur of the moment”, “I expected it to happen soon but was not sure when” or “it was completely unexpected” were coded as having had an unplanned sexual debut. Pupils who answered “I can’t remember” or had missing data were coded as “can’t remember/missing”.

 Age at sexual debut

Pupils in both MES and AES were asked “when you first had penetrative sex, how old were you?”.

Responses were recoded as “under 13 years of age” vs. “13 years of age or over” vs “missing”. Age 13 was chosen as in the UK, sexual activity under this age is a reportable offence.

 Partner age at sexual debut

Pupils in both MES and AES were asked “how old was the person you had penetrative sex with?” Age differences ranged from −2 years (younger) to 42 years (older).

Responses were recoded as “three plus years older” vs. “1–2 years older, the same age or 1–2 years younger” vs. “can’t remember/missing”.

 Felt pressured at sexual debut

Pupils in MES were asked which of the following statements best described the circumstances surrounding their sexual debut: “he/she put a lot of pressure on me”, “he/she put pressure on me”, “he/she put a bit of pressure on me”, “there was no pressure either way”, “I put a bit of pressure on him/her”, “I put pressure on him/her”, “I put a lot of pressure on him/her”. Pupils in AES were provided with a condensed response set of “he/she put pressure on me”, “there was no pressure either way” and “I put pressure on him/her”.

As we were interested in the proportion of pupils who had felt pressured into having sex, were responses were recoded into “I felt pressured at sexual debut” vs. “I did not feel pressured at sexual debut” vs. “missing”. MES pupils who answered “he/she put a lot of pressure on me” and “he/she put a bit of pressure on me” and AES pupils who answered “he/she put pressure on me” were coded as having experienced pressure. All other responses were coded as not having experienced pressure.

 Sexual debut was under the influence of drugs or alcohol

Pupils in both MES and AES were asked “were you drunk or stoned when you first had penetrative sex?” Responses included “yes” and “no”.

Responses were recoded as “yes” vs. “no” vs. “missing”.

 Discussed using condoms with partner at sexual debut

Pupils in both MES and AES were asked “did you talk about protecting yourself from pregnancy and sexually transmitted infections with the person you had penetrative sex with before having penetrative sex for the first time?” Responses included “yes”, “no” and “can’t remember”

Responses were recoded as “yes” vs. “no” vs. “can’t remember/missing”.

 Has had more than the median number of partners

Pupils in both MES and AES were asked “how many people have you ever had penetrative sex with?”

Responses were recoded into “1–2” and “3+” partners, based upon a median response of 2 partners in a positively (right-tailed) skewed distribution. Missing data was coded as a category and included in analyses.

Explanatory variables: pupil characteristics

 Pupil sex

Pupils in both MES and AES were asked “are you a male or a female?”

Responses were left as a binary variable comparing “male” vs. “female”.

 Pupil age

Pupils were asked “what month and year were you born?”

Pupil’s age was calculated in years by subtracting the month and year of birth supplied from the month and year questionnaire data was completed. Comparing the mean ages of pupils in AES and MES settings using independent samples t-tests highlighted that AES pupils were significantly older than MES pupils (AES mean age 15.9 vs. MES mean age 15.5; t(214) = 6.017, p = 0.000). As the likelihood of engaging in sexual activity increases with age, and there was the potential for a dose related response for secondary sexual health outcomes such as the number of sexual partners, age was dichotomized above and below the mean age of MES pupils to control for any increased levels of risk associated with age.

Explanatory variables: socio-economic circumstances

 Family composition

MES pupils were asked “when you are at home, which adults do you normally stay with?” AES pupils were asked “over the past two years, which adults have you stayed with most of the time?” Possible responses included “mother”, “father”, “step-mother”, “step-father”, “grandmother”, “grandfather”, “I live in a care or foster home” and “other”. Pupils could tick all that applied.

Responses were recoded into “lives with both biological parents”, “lives in a single parent household”, “lives in a reconstituted (step) family” and “lives in a care or foster home”. Missing data was coded as a category and included in the analysis.

 Maternal and paternal employment status

Pupils were asked if their mother/father was “in full time paid work”, “in part time paid work”, in “full time housework”, “unemployed”, “a student”, “sick or disabled” or “retired”. Young people could also state that they were “not sure” and that they did not have a “mother/female guardian” or “father/male guardian”.

Responses were recoded to “works full time”, “work part time”, “is unemployed” (e.g. in housework, unemployed, student, sick/disabled or retired) and “no guardian/missing”.

Explanatory variables: family influences

 Frequency of parent-child arguments

Pupils in MES and AES were asked “how often do you have serious disagreements or arguments with your parents/guardians/carers about things, for instance drinking, your friends, homework, tidiness or what you wear?” Responses included “every day”, “most days”, “weekly”, “less often” and “never”.

Responses were recoded into “argues with parents at least once a week” vs. “argues with parents less often”. Missing data was coded as a category and included in the analysis.

 Parent-child connectedness

Pupils in MES and AES were asked how strongly they agreed with the following statements using a 5 point scale (1 = strongly agree, 5 = strongly disagree), my parents: “sense when I’m upset about something”, “try to control everything I do”, “encourage me to talk about my difficulties” and “treat me like a baby”.

A mean score (alpha = 0.592) was created for each participant, and recoded into high and low connectedness based upon mean scores above /below the median in a negatively (left-tailed) skewed distribution. Missing data was coded as a category and included in the analysis.

 Family time

Pupils in MES and AES were asked about the amount of time they spent “eating a meal”, “going for a walk or playing sport” and “going places” with their parents, and “doing other things as a family group”. Responses were based on a 4 point scale (AES 1 = more than once a week, 4 = never; MES 1 = everyday, 4 = never).

Items on “eating a meal” and “going for walks” together were combined to form a mean score (alpha = 0.509) for each participant. Other items were excluded due to reliability testing showing low internal consistency (alpha < 0.2) when all items were combined into a scale. The combined family time score was then recoded to “frequent family time” vs. “less frequent family time” based upon mean scores above /below the median in a positively (right-tailed) skewed distribution. Missing data was coded as a category and included in the analysis.

 Spending money (proxy variable for young people”s autonomy)

Pupils in both MES and AES were asked “each week, how much money do you get to spend on things you want?” using a seven point scale (1 = nothing, 2 = less than £7, 3 = £7–12 increasing through £5 increments to 7 = £30 or more).

Responses were recoded into “less than £25 a week” and “£25 a week or more”. Missing data was coded as a category and included in the analysis.

 Parental monitoring

Pupils in both MES and AES were asked two questions about parental monitoring: “do you have to be back by a certain time” and “does anybody stay up until you get home”. Responses were based on a 4 point scale, with 1 = always and 4 = never.

Items on “do you have to be back by a certain time” and “does anybody stay up until you get home” were combined to form a mean score (alpha = 0.591) for each participant. Scores were recoded into high and low parental morning based upon mean scores above/below the median in a positively (right-tailed) skewed distribution. Missing data was coded as a category and included in the analysis.

Explanatory variables: educational factors

 Intended school leaving age (Educational engagement)

Pupils were asked when they intended to leave school. Pupils in MES were asked if they intended to leave school as soon as legally possible (“at the end of S4” or “Christmas S5”), intended to remain in school until completion of upper school exams (“the end of S5” or “the end of S6”). As pupils in AES were likely to be older they were asked if they wanted to leave “at age 16/as soon as possible” or “at an older age”. A “don’t know” category was provided in both questionnaires.

Responses of “at the end of S4 or Christmas S5” among MES pupils and “at age 16/as soon as possible” were recoded as leaving school “as soon as possible”. Response of “the end of S5”, “the end of S6” for MES pupils and “at an older age” were recoded as “remain in school”. Missing answers and “don’t know” were combined as a category and included in the analysis.

 Future expectations of: employment, education and training by age 19

Pupils in both AES and MES used a series of 3- (AES, 1 = likely, 3 = unlikely) and 5-point scales (MES, 1 = very likely, 5 = very unlikely) to report the perceived likelihood of their being: “in a secure job/apprenticeship”, “in a training scheme”, “at college or university”, or “unemployed or in casual work” in the next three years.

Responses of “likely” and “very likely” were recoded into “will likely be in education, employment and training” whilst responses of “unlikely” and “very unlikely” were coded as “will not be in education, employment or training”. Missing answers and “unsure” categories were combined and included in the analysis.

 Future expectations of: parenthood by age 19

Pupils in both AES and MES used a series of 3 (AES, 1 = likely, 3 = unlikely) and 5-point scales (MES, 1 = very likely, 5 = very unlikely) to report the perceived likelihood of having “a child/children” in the next three years.

Responses of “likely” and “very likely” were recoded into “will likely be a parent” whilst responses of “unlikely” and “very unlikely” were coded as “will not be a parent”. Missing answers and “unsure” categories were combined and included in the analysis.

Explanatory variables: peer influences

 Peer engagement with education

Pupils in both AES and MES were asked what proportion of their peers had “left school”, using a 5 point scale (1 = none, 5 = all).

Responses were recoded into “more than half” vs. “less than half”. Missing data was coded as a category and included in the analysis.

 Peer smoking behaviour

Pupils in both AES and MES were asked what proportion of their peers

Responses were recoded into “more than half” vs. “less than half”. Missing data was coded as a category and included in the analysis.

 Peer sexual behaviour

Pupils in both AES and MES were asked what proportion of their peers do “you think have had penetrative sex”, using a 5 point scale (1 = none, 5 = all).

Responses were recoded into “more than half” vs. “less than half”. Missing data was coded as a category and included in the analysis.