On the Edge of Adulthood: Young people's school and out-of-school experiences at 16

13. Intersections of relationships and experiences

Are young people who give low ratings to their level of communication with their parents more likely to have friends with risky behaviour, or undertake risky behaviour themselves? What are some of the other experiences and attitudes that are related to the variables that we have seen associated with lower school participation, engagement, and achievement? In this chapter we look at the intersections of different patterns of friendships, relationships with parents, experiences, values, and interests.

As in earlier chapters, we use the correlations between factors that we could put onto a 1–10 scale; one-way or single-factor ANOVA comparisons of average scores on the scales for categorical variables, and some models that use variables that showed moderate to strong linkages in the correlations and ANOVA results to shed some light on key factors that may be related to differences in kinds of friendship and behaviour.

Intersections between family relationships, friendships, and experiences

We start with the correlations among the factors related to young people’s views of their family relationships (see the Appendix for a description of the factors). These include similar factors from two years earlier, showing a reasonable degree of continuity in the young people’s views, particularly in relation to both habits of communication, and what some young people experienced as pressure. Young people who see their family as supportive are also highly likely to see themselves as part of their family, and to report a high level of communication with their parents. There is less correlation between these three aspects of family life and young people’s reports of feeling pressured by their families: indicating that it is possible to feel supported, included, and able to talk with your parents, but still feel pressured by them. There is even less correlation with two key aspects of the young people’s behaviour two years earlier: young people who tried out risky behaviour at age 14 (not shown in the table as all correlations were below the 0.2 cut-off), or who experienced rejection then were no less, or more, likely to be closer to their families than others.

However, there is more correlation between current levels of risky behaviour and reports of parental pressure; as there is with experiences of rejection.

Young people who reported friendships that extended them—and involved communication—were more likely to have good family communication levels also, suggesting that communication skills in one sphere are of use in the other—and what is learnt in one sphere about communication may transfer to the other. This interpretation is consistent with another pattern evident in the table: solid friendships showed similar levels of correlation with aspects of family inclusion and support, but not with the level of family communication.

– indicates -0.2 < r < 0.2; all correlations stronger than ± 0.4 in absolute value are in bold face.

When we looked at the relationship between young people’s views of their relationship with their parents, and our categorical variables, we found both differences and a lack of differences that are of interest. These are summarised below (the full analysis is contained in the technical report (Hodgen, 2008)):

• Attendance, gender, ethnicity, and maternal qualification are unrelated to differences in scores on these measures of the quality of young people’s family experiences. Thus, for example, males were just as likely as females to experience communicative families, and family pressure; and the students whose school attendance was poor were just as likely as those whose school attendance was excellent to experience a supportive family, or one that pressured them.
• Students whose families had had low incomes at age 5 had lower scores for the family communicates well and supportive family measures (at the indicative level).
• Students with standing out values had lower scores for the positive family measures, and higher scores on the family pressure measure. Similarly, students in the electronic games/no interests cluster at age 14 had lower scores for the positive family measures. Students whose parents thought they had not enjoyed school over the past four phases of the study were likely to have lower scores on the positive family measures.
• Family pressure levels were higher for students in “vocational” or “contextual” subject clusters, and in terms of patterns over ages 8 to 14, those who had not enjoyed school, for those who had been involved in bullying in at least one of the four previous study phases, and for those who had never enjoyed reading; and at an indicative level, for young people whose families were in a difficult financial situation at age 14.
• Family inclusion was also related to subject cluster (highest levels among those in the “traditional science” cluster; lowest in the “vocational” cluster).
• Scores on the family inclusion and supportive family measures were higher for those with more Level 1 NCEA credits.
• Family communicates well scores were lower for those who had never enjoyed reading, or who were heavy TV watchers, and higher for those with higher age-14 motivation levels.

Intersections between friendships and experiences

The next table looks at the correlations between friendships and experiences over the past year, and back to age-14 levels. There is a moderately strong likelihood that those who had risky behaviour and friends with risky behaviour two years earlier continued along this path at 16; there is less correlation over the two years for the positive kinds of friendship. When we look at current correlations, there is a moderately strong likelihood that those who find their friendships extending also find them solid.

There is very little correlation between these positive kinds of friendship and having friends with risky behaviour. This suggests how risky behaviour shared with friends can become entrenched; and how such behaviour may indicate a desire to impress others rather than share feelings. The correlation between risky behaviour and praise and achievement may surprise, but this item did include items such as being included in a group you really wanted to be in, and supporting a friend in trouble. The risky behaviour pattern also shows a moderate correlation with adverse events, and some correlation with rejection (and rejection and adverse events are also moderately correlated (0.38).

– indicates -0.2 < r < 0.2; all correlations stronger than ± 0.4 in absolute value are in bold face.

Risky behaviour and friends

Young people with higher levels of risky behaviour and friends with risky behaviour were more likely to be in the “vocational” or “contextual” subject clusters at school, to have “standing out” values, to be in the sports or electronic games/no interest clusters at age 14, to have not enjoyed reading, and been heavy TV watchers, had some involvement in bullying in at least one of the four previous phases of the study, and be less likely to have had high motivation levels at age 14. Gender was unrelated to patterns of friendship and events. Young people whose mothers had no qualification or a trades level qualification, and Mäori/Pacific young people were more likely to have higher levels of risky behaviour and friends with risky behaviour, as were those whose family income had been low at age 5, low at 14, and whose family was in a difficult financial situation at age 14.

Results of multivariate models

We fitted some multivariate models to gain additional insight into the stability of the patterns of friendship and family relations we were seeing at age 16. To what extent did age-14 patterns contribute to them, and which were the age-14 aspects of friendships, family relationships, and experiences that showed a continuing contribution? We also included social characteristics in these models: but on the whole they did not show significant contributions to age-16 patterns, indicating that these patterns are not very different for young people in different social groups, once other characteristics of their lives have been taken into account.

Because of the strong correlation level between each of the three positive aspects of family relationships, usually only one of these family factors appeared in the model. The patterns reported in Table 106 show that the greatest continuity between age 14 and age 16 is in the area of risky behaviour, and having friends with risky behaviour. Continuity is least with regard to experiencing adverse events (which include accidents and illness).

The comparatively low level of continuity for experiencing solid friendships is likely to be because the age-16 factor did not include all of the age-14 items, some of which formed the age-16 factor extending friendships.

Some “virtuous” cycles can be seen, e.g. in relationship to extending friendships; and some “vicious” cycles, e.g. in relationship to family pressure, and most strongly in relationship to risky behaviour and having friends with risky behaviour.

The praise and achievement pattern is interesting: it shows perhaps two different sets of experiences and how those link together, for those who support friends in trouble or are praised for achievement in a positive set of relationships and experiences, and those who provide such support or receive such praise in a negative, or risky set of relationships and experiences. One example of how the same thing can interconnect either positively or negatively is that negative scores for age-16 risky behaviour were more likely for those in the sports-interest cluster (which makes sense in terms of some of the contexts in which team sports occur), as were lower levels of enjoyment of reading, yet this cluster was just as likely to experience praise and achievement, which can also support more positive family relationships.

Thus when we are thinking of “at-risk” young people, we need to be mindful that they do not inhabit a distinct and separate niche. Thinking about those who play sports, for example, our analysis does not suggest that a warning light goes on for every young person who does play sports; but that we pay attention to what the specifics of what happens around our young people’s involvement in sports—whether it provides opportunities for risky behaviour, or contrarily, for extending friendships.