Goal conflict, ambivalence and psychological distress: Concurrent and longitudinal relationships

Conflict between goals (inter-goal conflict) and conflicting feelings about attaining particular goals (ambivalence) are believed to be associated with depressive and anxious symptoms, but have rarely been investigated together. Kelly et al. (2011, Personality and Individual Differences, 50, 531-534) reported that inter-goal conflict interacted with ambivalence to predict concurrent depressive symptoms in undergraduates, with ambivalence being more strongly associated with depressive symptoms for persons reporting less inter-goal conflict. We sought to replicate and extend this finding in a larger sample, using separate measures of inter-goal conflict and facilitation, and a longitudinal follow-up. Undergraduates (N=210) rated their goal strivings for ambivalence, inter-goal conflict and facilitation, and completed measures of depressive and anxious symptoms that were repeated after one month. Inter-goal conflict (but not facilitation) and ambivalence were both uniquely positively associated with depressive and anxious symptoms concurrently, but did not predict symptom change. Inter-goal conflict and ambivalence did not interact to predict concurrent symptoms, but inter-goal conflict was associated with greater reductions in anxious symptoms for people reporting low ambivalence. Findings suggest that different forms of motivational conflict across the goal hierarchy are associated with symptoms, but do not exacerbate


Introduction
Making progress on personal goals imbues life with meaning and contributes to well-being (Brunstein, 1993;Klinger, 1977;Klug & Maier, 2015), so it is unsurprising that goal conflict has long been considered to be associated with psychological distress (Higginson, Mansell, & Wood, 2011). This article examines how two different forms of conflict (inter-goal conflict and goal ambivalence) contribute to anxious and depressive symptoms.
A person experiences inter-goal conflict when one of their goals makes it more difficult to pursue their other goals (Emmons, 1986;Riediger & Freund, 2004). For example, a person's goal to 'spend more time with my family' may conflict with their goal to "get promoted at work". Conversely, a person may experience inter-goal facilitation if one of their goals makes it easier to pursue their other goals (e.g., "spend more time with family" may facilitate the goal to "deepen my relationships"). Inter-goal conflict is associated with negative affect and lower life satisfaction (Emmons, 1986) and more psychiatric symptoms among undergraduates (Perring, Oatley, & Smith, 1988) and adolescents (Dickson & Moberly, 2010). However, some studies using undergraduate samples have not found associations between inter-goal conflict and depressive (Emmons & King, 1988, Study 2;King, Richards, & Stemmerich, 1998;Segerstrom & Solberg Nes, 2006) or anxious symptoms (Emmons & King, 1988, Study 2). In community samples, no significant correlations emerged between inter-goal conflict and depressive symptoms (Wallenius, 2000) or negative affect (Kehr, 2003;Romero, Villar, Luengo, & Gómez-Fraguela, 2009). Equivocal results may reflect the use of bipolar measures that conflate inter-goal facilitation and conflict. Riediger and Freund (2004) found that unipolar measures of inter-goal conflict and facilitation loaded on distinct factors, with only inter-goal conflict being significantly associated with negative affect at the between-and within-person level. Boudreaux and Ozer (2013) found that inter-goal conflict, but not inter-goal facilitation, was positively correlated with anxiety and negative affect in undergraduates; the correlation with depressive symptoms was not significant. In their meta-analysis, Gray, Ozer, and Rosenthal (2017) revealed that goal conflict was positively associated with psychological distress (weighted effect size: r = 0.34), with studies using unipolar scales yielding larger effect sizes.
Inter-goal conflict may be less distressing if it represents competition among goals for a shared limited resource (e.g., time or money) rather than inherently incompatible outcomes (Riediger & Freund, 2004;Segerstrom & Solberg Nes, 2006). However, conflicted motives about attaining specific goals, i.e., ambivalence (Bleuler, 1911;Sincoff, 1990), may illustrate more profound motivational conflict that is more strongly associated with psychological symptoms. Goal ambivalence has been conceptualised as an approach-avoidance conflict about the pursuit of a particular goal (Emmons, King, & Sheldon, 1993) that is generated by conflict between relevant goals at a higher level in the goal hierarchy (Kelly, Mansell, & Wood, 2015). For example, a student may feel ambivalent about an essay-writing goal because it is relevant to a higher-level goal conflict between excelling academically and maintaining interpersonal relationships. Higher-level goal conflict may be more irresolvable because such goals are self-defining (Powers, 1973). Goal ambivalence has indeed been found to be associated with anxious and depressive symptoms among undergraduates (Emmons, 1986;Emmons & King, 1988;King et al., 1998;but see Romero et al., 2009, for null results). Other research has examined the association between psychological symptoms and ambivalence about goals relevant to particular life stages. For pregnant women, ambivalence about childbirth was associated with concurrent depressive symptoms and increasing symptoms post-partum (Koletzko, La Marca-Ghaemmaghami, & Brandstätter, 2015). In another sample, daily fluctuation in ambivalence about having the child was associated with negative affect. In another study, ambivalence about attaining a degree was associated with lower life satisfaction both concurrently and longitudinally (Koletzko, Herrmann, & Brandstätter, 2015).
Inter-goal conflict and ambivalence may overlap because people will often feel ambivalent about conflicting goals (Emmons & King, 1988). Indeed, modest positive correlations have been reported between goal ambivalence and inter-goal conflict at the within-person level (Emmons, 1986;King et al., 1998), if not at the between-person level. Few studies have examined whether inter-goal conflict and ambivalence have independent or interactive associations with symptoms (Kelly et al., 2015). Although Emmons (1986) found that ambivalence but not inter-goal conflict explained unique variance in psychological symptoms, this study was underpowered.
Kelly et al. (2011) reported that goal ambivalence was positively associated with concurrent depressive and anxious symptoms, whereas inter-goal conflict did not predict significant additional variance. Moreover, these forms of conflict interacted such that ambivalence was more strongly associated with depressive symptoms for participants reporting less inter-goal conflict. The authors speculated that ambivalence may be more distressing if it is not attributable to the pursuit of conflicting lower-level goals, suggesting that the ambivalence is generated by higher-level goal conflict. A person who strives to run marathons and learn guitar may report inter-goal conflict due to limited leisure time, but may experience no ambivalence if these pursuits are consistent with higher-level goals (Kelly et al., 2015). Conversely, a person who strives to care for the vulnerable and provide childcare may report no inter-goal conflict, but may experience ambivalence if these pursuits conflict with a higher-order goal of being independent. A combination of low inter-goal conflict and high ambivalence may indicate a distressing lack of integration across levels of the goal hierarchy. However, Kelly et al.'s (2011) result requires replication, and it is unclear whether the relationship between ambivalence and depressive symptoms is moderated by lower levels of inter-goal facilitation and/or higher levels of inter-goal conflict.
To further illuminate the unique and interactive relationship between inter-goal conflict, ambivalence and psychological distress, we extended Kelly et al.'s (2011) research using a larger sample and distinct measures of inter-goal conflict and facilitation (Riediger & Freund, 2004). We also examined whether inter-goal conflict, goal ambivalence and their interaction would predict symptom change over one month, consistent with the notion that inter-goal conflict actively contributes to psychological distress. Boudreaux and Ozer (2013) found that intergoal conflict predicted increases in depressive and anxious symptoms over five weeks in undergraduates. Similarly, Koletzko, La Marca-Ghaemmaghami, and Brandstätter (2015) found that ambivalence about having a child in women was associated with worsening depressive symptoms after birth.
Based on the notion that conflict is deleterious at all levels of the goal hierarchy (Powers, 1973), we hypothesised that inter-goal conflict and goal ambivalence would each predict unique variance in anxious and depressive symptoms. Inter-goal facilitation was included as a covariate, but was not expected to be associated with anxious or depressive symptoms (Riediger & Freund, 2004). We sought to replicate Kelly et al.'s (2011) interaction between ambivalence and inter-goal conflict, such that anxious and depressive symptoms would be highest for individuals reporting high level of goal ambivalence and low levels of inter-goal conflict. Prospectively, we expected that higher levels of ambivalence and inter-goal conflict would each predict increases in anxious and depressive symptoms. More tentatively, we predicted that the interaction between inter-goal conflict and ambivalence would explain additional variance in symptom change.

Participants
Two hundred and ten undergraduate students (169 women, 41 men; M = 20.0 years, SD = 2.5, range = 18-35) were recruited from the University of Exeter campus via online advertisements. Participants were remunerated with course credit or £15.

Materials and procedure
Participants attended an initial 1 h session in which they provided informed consent, before completing a personal strivings assessment, inter-goal conflict and facilitation matrices, and depressive and anxious symptom scales.

Personal goal strivings (Emmons, 1986)
Participants first read instructions asking them to list at least ten personal goals, defined as "things that you typically or characteristically are trying to do", by completing the stem: "I typically try to…" Examples were provided (e.g., "Convince others that I am intelligent") and participants were told that they should list goals that identified them as individuals, rather than goals that other people thought they should have. Participants who generated more than ten goals were asked to choose the ten that represented them most accurately. Allowing for minor wording changes, Emmons (1986) found that 82% of goals were consistent over one year.

Goal ambivalence (Emmons, 1986)
Participants rated their ambivalence about each of their goals on a 6-point scale from 0 (none at all) to 5 (extreme) in response to the following question: "Sometimes even though we successfully reach a goal, we are unhappy (e.g., if you're "trying to become more intimate with someone" and you succeed, you might also feel concern about being tied down). How much unhappiness do you or will you feel when you are successful in this striving?" Mean ambivalence scores across goals were calculated for each participant (α = 0.79). Goal ambivalence has previously shown a one-year stability correlation of 0.65 (Emmons & King, 1988). (Riediger & Freund, 2004) Participants next completed two 10 × 10 matrices to rate inter-goal conflict and facilitation respectively. In each matrix, each of the participant's ten goals was listed in both rows and columns. In the conflict matrix, participants rated the extent to which pursuing each of their goals in the rows "makes it more difficult to pursue" each of the other