Constructing future expectations in adolescence: relation to individual characteristics and ecological assets in family and friends

Abstract Required to manage multiple developmental tasks, multiple systems and processes are involved in making adolescents thinking about and making plans for the future. The current study aims at exploring the manner in which individual (optimism, depressive symptoms, anxiety symptoms, and self-concept) and contextual factors (family and peer influences) contribute to the construction of adolescents’ future expectations. 1509 youth (891 females and 618 males) with different socio-economical backgrounds were questioned using an online platform. Compared with their counterparts, girls, adolescents who were not experiencing severe material deprivation, those with high self-concept, displaying little depressive symptoms, and high level of optimism have more positive future expectations. As most of the available empirical data come from Western countries, these results complement the existing data in the area of emerging adulthood.


Introduction
Adolescence is an important developmental stage when multiple systems are involved in a complex process that influences adolescent's thinking and planning about adulthood and future expectations. Thinking about the future and constructing perceptions of oneself in the future might have a bigger role during adolescence than during other developmental stages of life (Trempala & Malmberg, 2002). However, according to some popular stereotypes, 'adolescence' is also a period when youth are notoriously short-sighted, oriented to the immediate rather than the future, unwilling or unable to plan ahead, and less capable than adults at envisioning the longer term consequences of their decisions and behaviours (Steinberg et al., 2009).

Adolescents' hopeful future expectations
There are different terms used when discussing about 'hopeful future expectations' (McWhirter & McWhirter, 2008;Seginer, 2000Seginer, , 2008, but in the current study, the term will be used and defined as 'the likelihood one expects a specific event will occur in the future' (Oettingen & Mayer, 2002). The topic is important because the way in which adolescents construct their future can have serious and far-reaching effects on their prospective well-being and behaviour. Theory and previous research suggest that positive future expectations can facilitate optimal development and a successful transition into adulthood (Schmid, Phelps, & lerner, 2011). Catalano, Berglund, Ryan, lonczak, and Hawkins (2004) concluded that positive beliefs about the future are associated with long-term plans for higher education, positive thoughts in relation to work, better social and emotional adjustment at school, and self-perceptions of competency. On the other hand, adolescents who anticipate a negative future are more likely to exhibit problem behaviour like engaging in delinquency, substance use, and risky sexual behaviour (Sipsma, Ickovics, lin, & Kershaw, 2012. Starting from these prerequisites, especially in the area of youth programmes, it is essential understanding what combination of factors promotes positive future expectations among youth.

Individual and ecological influences during adolescence
even if 'individual' and 'ecological' attributes are not truly independent of one another, we categorized our independent variables according to the levels of proximity in order to comprehensively address our research objective, namely to identify how different sources of influence (i.e. characteristics attributed to individuals, families and friends) are associated with constructing (positive and negative) future expectations during adolescence. We began by reviewing individual and psychological characteristics that have been previously found to be associated with future expectations. Brown, Teufel, Birch, and Kancherla (2006) found that boys tend to express more worries about the future as compared with girls. Positive expectations towards future were also positively correlated with adolescents' subjective well-being (eryilmaz, 2011). The level of self-confidence is an important resource in conceptualizing the future. Therefore, future expectations are associated with greater confidence (Snyder & lopez, 2005), youth with high self-confidence being more likely to have positive expectations about their future outcomes, (Thompson & Zuroff, 2010).
The development of an individual's future expectations is dynamically linked to his or her relationships with significant others (e.g. family members or friends). Adolescents' perception of warm, supportive and reliable interactions between themselves and their parents provides models of positive behaviours that youth later incorporate into their own self-concepts, which in turn informs their hopes for the future (McCabe & Barnett, 2000). Dubow, Arnett, Smith, and Ippolito (2001) concluded that parental support predicted increases in positive future expectations in a sample of disadvantaged inner-youth. Peer support was also shown to contribute to adolescents' hope (Harter & Whitesell, 1996). Also, positive future expectations are higher when collective efficacy in the community is high and youth are engaged in community activities (Stoddard & Pierce, 2015). Jose, Ryan, and Pryor (2012) found that a higher level of community connectedness was associated with adolescent well-being, including future orientation. However, leventhal and Brooks-Gunn (2000) have previously demonstrated that neighbourhood effects are usually mediated by more proximal factors such as individual-level psychological processes. We may conclude that strong social relationships increase adolescents' perceptions of a positive future (Schmidt, Pierce, & Stoddard, 2016) In turn, without contextual support youth may become disengaged from their future selves. Seginer and lilach (2004) found that a sense of loneliness among youth was associated with lower scores on motivation orientation, measured by value, expectance, and internal control of one's future plans. Therefore, having adequate social support, ranging from one to multiple networks including family, peers and school, is pertinent for maintaining adolescents' psychological well-being and future positive development (Catalano et al., 2004). Since Bronfenbrenner (1979) developed his ecological model, studying ecological influences together with individual characteristics has become a common practice in social sciences. The present study focused on both individual-level (optimism, depressive symptoms, anxiety symptoms, and selfconcept) and contextual factors (family and peer influences) in relation to their hopeful expectations, after controlling for demographic characteristics (gender, residency, and economic status).

Sample and participant selection
This study is part of a larger longitudinal investigation on transition of adolescents to young adulthood in Romania (OoA) (www.viitoradult.ro/en/). This article is based on data available from the second wave. Participants included 1509 students (43% response rate compared to the first wave), 59% girls, 61.9% with residency in urban areas and 15.6% living in severe material deprivation conditions (Table 1).

Procedure
using the information from participating in the first wave, during november 2014 and februray 2015, students were contacted via email, phone or facebook account and asked to log-in again on the online platform and fill the questionnaire. Participants were reminded again about the confidentiality of their responses and were informed about the raffle ticket upon filling the questionnaire as an incentive. using their log-in information, four tablets and ten memory sticks were given as prizes to winning students.

Measures
Seven continuous independent variables (optimism, depressive and anxiety symptoms, general self-concept, perceived parent and friend support, family connectedness) and one continuous dependent variable (hopeful future expectations, divided into two subsets: positive and negative outcomes) were employed:

Hopeful future expectations
Items related to hopeful future expectations evaluated participants' opinion about specific future events. Situations later in life. Participants were asked the following question: 'Think about a person of same sex, gender and social status as yourself, how are your chances for the following?' Items included graduating from college, being healthy, having a job that one desires, having a happy family life, but also negative expectations like being convicted, having financial difficulties, and having health problems. There was a total of 12 items, with responses ranging from 1 = very little chances to 5 = very high chances. Higher scores indicate higher expectations of the likelihood that certain future outcomes will occur. In order to reduce the degrees of freedom in the models estimated for the present study, the 12 items were divided into two groups of six items each, serving as output variable of the positive and negative hopeful future expectations. In the OoA data-set, the Cronbach's alphas for the hopeful future scale were .72 for positive and .78 for negative outcomes. Table 2 summarizes the descriptive results for this measure and for each of the following scales.

Optimism
Optimism was measured with a short version of the lOT-R scale (life Orientation Test Revised: Scheier, Carver, & Bridges, 1994). The life Orientation Test measures dispositional optimism, which is defined as generalized positive outcome expectancies. It provides a self-report measure of individual differences in global optimism. Optimism scale included items like 'In uncertain times, I expect the best' , 'I'm always optimistic about my future' , 'I rarely count on good things happening to me' , and 'Overall, I expect more good things to happen to me than bad' . Participants were asked to indicate the extent of their agreement to the 10 items on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). Higher scores indicated more optimistic attitudes. for the current study, the Cronbach's alpha for this measure was .67.

Depressive symptoms
The Beck Depression Inventory is a widespread self-report rating tool measuring characteristic attitudes and symptoms of depression. for the current study we used the short version because it also has good psychometric properties, identify the multiple aspects of depression, and proved to assess symptomatology in general population of adolescents and young adults (Storch, Roberti, & Roth, 2004).
Participants were asked to think at the last two weeks and to rate how true statements were for them, on a scale from 1 (not at all) to 4 (very much). Items on the Depressive Symptoms scale included 'I lost interest in others' , 'I am too tired to do anything' , 'I see my future as hopeless and nothing will change' , and 'I blame myself all the time' . Higher scores indicated more frequent depressive symptoms. Internal consistency in the current sample was .81.

Anxiety symptoms
Anxiety symptoms were measured using HADS (Hospital Anxiety Scale: Zigmond & Snaith, 1983). The items included 'I feel tense or wound up' , 'I can sit at ease and feel relaxed' , and 'Worrying thoughts go through my mind' . Participants were asked to think at the past two weeks and rate how true statements were for them, on a scale from 1 (not at all) to 4 (almost every day). Higher scores indicated more frequent anxiety symptoms. The internal consistency of the Anxiety Symptoms scale in the current sample was .88.

General self-concept
General self-concept refers to a person's beliefs about themselves, including their self-worth and self confidence. The scale is comprised of 5 items ('I take a positive attitude toward myself' , 'On the whole, I am satisfied with myself' , 'I am able to do things as well as most other people' , 'I feel that I have a number of good qualities' , and 'I have faith in myself'), each rated on a 3-point likert scale (1 = not at all; 3 = a lot). By summing up the responses of the 5 items, a total score was generated, with higher scores indicating higher levels of self-confidence. Cronbach's alpha for this measure in the current study was .81.

Perceived parent support
Adolescents' perceived sense of support from their parents described the degree to which youth report that adults in their home provide them with loving support and encouragement. It was measured using a 5-item parent supportiveness subscale from the School Success Profile-Romania Survey (Hărăguș,

Family connectedness
family connectedness scale measured using a 4 item scale, the degree in which youth report that the people in their home feel a sense of emotional closeness and bonding with one another, and solve problems together (e.g. 'When you have personal problems or decisions to make, how often do you talk with your parents' , 'How often they listen to your arguments'). Participants were asked to rate their response on a 4-point likert scale (1 = never -4 = always) and by summing up the values, higher scores indicated higher connectedness with family members. The Cronbach's alpha for the current measure in this study was .76.

Perceived friend support
Adolescents' perceived sense of support from their peers measured the degree to which youth perceives their friends as trustworthy and supportive and as responsive to their needs and feelings (e.g. 'I can trust my friends' , 'I feel close to my friends'). The combined 4-item scale used a three-point likert response format for each item (1 = not at all -3 = a lot) and was derived from the School Success Profile-Romania Survey (Hărăguș et al., 2010). By summing the values, a general score was generated, higher values indicating greater peer-group support. In this study, the Cronbach's alpha reliability coefficient for the perceived peer support scale was .92.

Demographics
Participants were asked on the survey to indicate their gender, residency, and to rate the affordability of a selection of items (goods or services) that are considered to be necessary or desirable for people to have an 'acceptable' standard of living. Based on this eu definition of 'material deprivation' (eurostat, 2012) a measure of 'severe material deprivation' was generated in which the threshold has been raised from three to four out of the same 9 items.

Preliminary analysis
Prior to all analysis, descriptive statistics was run for the interest variable in order to offer a picture of the phenomenon further discussed. Significant differences were found only for positive hopeful future expectations for all the grouping variables. Girls, youth from urban areas and adolescents experiencing no severe material deprivation have more positive expectations, compared to their counterparts. Respondents have quite optimistic thinking about their future: 71% agree or totally agree that overall, they expect more good things to happen than bad and 687% agree or totally agree that the future will be better. Adolescents residing in urban areas are more optimistic compared to their rural fellows [t(1343 = -3.91, p < .001]. Interestingly, youth with severe material deprivation display significantly more optimistic views about their future [t(1372) = 4.25, p < .001]. Students did not display high levels of depressive symptoms. The highest prevalent symptom was that in the last 2 weeks they felt tired in doing almost everything (22.5%). One out of ten also declared they lost most of the interest in other people, they felt guilty most of the time, they cannot decide on anything or woke up several hours earlier and could not get back to sleep. There are no significant differences in depressive symptoms for gender, residence and severe material deprivation (p >.05). Anxiety symptoms had a higher prevalence during the last 2 weeks: 27.7% declared that worrying thoughts went through their minds at least for half of the days, 22% did not feel restless and had to be on the move, and 20.8% cannot sit at ease and feel relaxed for almost half of the time. The only significant difference was noticed for gender [t(1413)=-6.01, p < .001], girls (M = 1.92) displaying more anxiety feelings compared to boys (M = 1.71). In general, adolescents had a positive image about themselves. More than 50% of them displayed high levels of self-worth and self-confidence, but with no significant differences among compared groups.
Parents are still an important source of emotional support, more than 60% declaring they feel supported and encouraged by their parents. The highest level of agreement is for counting on their help whenever needed. Significant differences were notices only for experiencing severe material deprivation [t(1361) = 2.54, p < .01], youth with no severe material deprivation benefiting from more support from their parents. More than one-third of adolescents have a good sense of family connectedness: 43.9% always listed to their arguments, and 1:3 discuss with them when important decisions are to be made. Gender [t(1439 = -4.91, p < .001)] and economic experience [t(310.35) = 2.46, p < .01] are important grouping variables, girls and adolescents experiencing no severe material deprivation declaring significant higher levels of family connectedness. Peer support is less than parent support, but still at high levels. 56.8% declared much closeness to their peers, 55.4% can count on their help at any time and 53.6% can discuss with them their problems. Significant differences in the level of peer support was noticed for severe material deprivation [t(313.05) = 446, p < .001] and residency [t(1073.19)=-4.81, p < .001]. As expected, youth residing in urban areas and those with no material deprivation have trustful relationships with their peers.

Correlational analysis
Correlational analyses were also conducted to determine the order in which to enter the variables of interest. following the ecological theory proposed by Bronfenbrenner (1979), experiences that are most salient to the individual should have the greatest impact on any outcome (e.g. everyday experiences vs. less frequent experiences). However, analyses were also run to see how well the present data fit this theory. Bivariate correlations are presented in Table 3. As would be expected, individual-level variables (general self-concept, depressive symptoms, and optimism) were most highly correlated with hopeful future expectations followed by the ecological-level variables. Based on these analyses and the literature, variables were entered into the model from most salient to least salient influence, specifically individual-level characteristics were entered first, followed by perceived family and friend qualities.
Overall, the entire model accounted for a total of 16.9% of the variance in positive and 6.8% of negative hopeful future expectations. Girls, adolescents who were not experiencing severe material deprivation, those with high self-concept, displaying little depressive symptoms, and high level of optimism have more positive future expectations. Contrarily, boys, adolescents with low self-concept and optimism, displaying high levels of depressive symptoms, with little family connectedness, and with high level of friend support have more negative expectations.

PRATT-index
A final step was to calculate the PRATT-index to assess the relative importance of each of the predictor variables in the full hierarchical regression model. The PRATT-index calculates the percentage of the total variance in the model that is explained by each variable based on a calculation of its beta weight, its correlation with the outcome variable, and the total R 2 in the model (liu, Zumbo, & Wu, 2014): each variable receives a score from 0 to 1, with all variables taken together accounting for 100% of the variance explained in the model. Including the PRATT-index adds a level of analysis as it indicates not only which variables are statistically significant but which contribute meaningfully in the model. Thomas (1992) suggested that as a general rule, if d <1/(2*p) (where p is the number of predictors), namely half the average importance, then the predictor can be regarded as unimportant. In the present study, predictors with a PRATT-index score smaller than .05 [d <1/(2*10)] explained relatively little of the variance in the model. PRATT-index scores for the current model are presented in Tables 4 and 5. four variables were identified as the most important in explaining adolescents' positive hopeful future expectations: self-concept and optimism (each accounting for 30% of the 16.9% total variance explained by the model), depressive symptoms (15%), and family connectedness (6%). Similarly, six variables were also identified as the most important in explaining adolescents' negative hopeful future expectations: depressive symptoms (30%), optimism (22%), self-concept (19%), perceived peer support (13%), family connectedness (9%), and anxiety symptoms (7%).

Conclusions
The topic of transition to adulthood is relatively new on the Romanian social science research agenda. However, from an applied perspective, the role of these results is important for understanding the dynamics of adolescents' aspirations, by informing practitioners with specific directions for facilitating the smooth transition to adulthood. Taking into consideration the correlates and effects of future expectations among youth, it is important to understand what promotes positive future expectations. The current study aimed to examine some individual characteristics and ecological assets that predict future expectations and worries about transitioning into adulthood in a sample of Romanian youth. employing two hierarchical regression models, the following conclusions emerged: (1) individual characteristics, such as gender and experiencing material deprivation are related with conceptualizing the future expectations. Consistent with previous research (Brown et al., 2006), girls have more positive future expectations compared to boys; (2) psychological attributes such as displaying self-confidence and optimism in youth are related with more positive expectations about the future. This means that adolescents with a strong and positive image about themselves are more likely to construct positive futures (McWhirter & McWhirter, 2008); (3) overall, family and peer-group support both showed little contribution in shaping adolescents' future expectations. Previous research suggests that adolescents' d = * r xy R 2 connectedness to both family and school predicted more positive perceptions of future orientation (Crespo, Jose, Kielpikowski, & Pryor, 2013). Our study has several strengths, including the use of a quite large sample to explore the future expectations of adolescents. Research was also theoretically grounded, using the ecological model to account for multiple domains of influence. However, accompanying the study's contributions are several inherent limitations: (1) using data on students enrolled in education, quite limits the generalizability of the results. Though our sample provides unique insight into this sample of youth, future research should investigate these relationships in other samples of youth (e.g. vulnerable children); (2) all measures were self-reported, which means that some data might be contaminated with social desirability. Additional to this, using electronic methods in gathering data required access to a computer and internet facilities and basic computer literacy.

Disclosure statement
no potential conflict of interest was reported by the authors.