The Relationship Between Attachment and Grit in Lower Income Adolescents

Each year in the United States, close to 1.3 million students do not finish high school. In fact, reports indicate that only 78.6% of those who enter high school ultimately graduate (Alliance For Excellent Education, 2010; Stillwell & Sable, 2013) and 15% of those between 18 and 24 years of age do not have a high school diploma (U.S. Census Bureau, 2004, 2006). Those from lower income backgrounds as well as minority students are even less likely to successfully complete high school and then continue toward a college degree, often in spite of high achievement during their early school years (Leonhardt, 2013). This disparity in educational completion is linked to the growing economic inequality in this country (Putnam, 2015). Increasingly, researchers are viewing character skills, or noncognitive skills, such as diligence, perseverance, self-control, and grit (Heckman & Rubinstein, 2001; Reeves, Venator, & Howard, 2014; Sroufe, 2005; Tough, 2012) as equally if not more relevant than cognitive skills in understanding this discrepancy. Knowledge of how to engender these skills may help our ability to address the growing economic and educational gaps in our country.

This research aims to further our understanding of one of these noncognitive skills: grit, defined as maintaining persistence and passion in the pursuit of a long-term goal (Duckworth et al., 2007; Duckworth & Gross, 2014). Past research has focused on its definition and its correlates (Eskreis-Winkler, Shulman, Beal, & Duckworth, 2014). However, there is still much to be learned about the factors that influence an individual’s grittiness. These authors suggest that the attachment bond, a construct that is underresearched with respect to grit, may be a critical component of this character skill.

Past research highlights the importance of the attachment relationship to optimal development. Alan Sroufe’s (2005) longitudinal data provides strong support for the notion that a secure attachment relationship can engender better self-regulation, help develop effective coping mechanisms, and improve overall successful functioning, all qualities correlated with grit (Duckworth et al., 2007). Levy and Steele (2011) explored the attachment bond as a likely underpinning of grit and found a positive relationship between the constructs. However, their survey study was conducted on an economically diverse sample in which ages ranged from 18–87, and relied solely on self-report measures. This study used narrative attachment instruments as well as self-report measures and focused solely on lower income high school seniors, a population that has not yet been studied. This research aims to both confirm a connection between attachment and grit and to establish this connection in a lower-income, adolescent population.

Low income high school seniors are seen as an ideal demographic group in which to study the relationship between attachment and grit for the following reasons. First, there is little research examining how adolescents’ levels of grit are affected by living in lower income neighborhoods. There is evidence to support the fact that cognitive development in children may be impacted by the stress of living in poverty. For instance, in a lower-income sample, lower executive function ability as well as IQ at age 3 was associated with higher levels of salivary cortisol, an indicator of a stress response (Blair et al., 2011). Furthermore, Duckworth, Kim, and Tsukayama (2013) found that deleterious life events encountered by school-aged children can negatively affect cognitive abilities such as inhibitory control and attention shifting or flexibility, abilities related to grit. However, the authors are not aware of other studies examining grit in adolescents from lower income areas. This research aims to contribute to the literature on this possible connection.

Secondly, high school seniors are in their late adolescence, a time of great vulnerability. Adolescence, or the period of transition to adulthood, has been called “demographically dense” (Rindfuss, 1991), or a time at which many potentially stressful life stage events often occur. This includes completing school, leaving home, and becoming economically viable. Epidemiological data shows that adolescence, as a phase of life, is associated with profound risks such as alcohol and drug abuse, self-harm, interpersonal violence, and legal difficulties for disadvantaged adolescents (Jensen & Nutt, 2015; Steinberg, 2014). In addition to the many life transitions occurring, adolescents are also undergoing many physiological changes. The brain is not yet fully developed and continues developing into early adulthood (Jensen & Nutt, 2015; Steinberg, 2014). Correlates of brain development, such as maturation of the overall neuroendocrine system, also continue to be experience-dependent during this time. Thus, adolescents may both experience more stress during this life phase and be less robust in their ability to manage stress in pursuit of longer term aspirations and goals. From an attachment perspective, it is a period of evolution where the adolescent moves away from receiving care and toward becoming an adult and being able to provide care (Allen, 2008). During this time of transition, adolescents very much need attachment figures to function as a secure base. Therefore, attachment security or insecurity may feature in how an adolescent persists toward meeting long-term goals.

Duckworth and her colleagues view grit as one of the factors critical to achievement and educational success (Duckworth et al., 2007; Duckworth & Quinn, 2009). According to Duckworth et al., individuals who are gritty maintain passion and perseverance when working toward long-term goals, often in spite of obstacles and setbacks (Duckworth & Gross, 2014). Research has corroborated this definition. Higher levels of grit predicted completion of a cadet summer training course above physical fitness, academic grades, or a leadership score (Duckworth et al., 2007). Gritty high school students were more likely to complete high school; gritty soldiers were more likely to finish three weeks of a difficult selection course; gritty sales people were more likely to exhibit job retention long-term; and gritty males were more likely to stay married (Eskreis-Winkler, Shulman, Beal, & Duckworth, 2014). Furthermore, National Spelling Bee finalists with more grit practiced more and therefore advanced to further rounds than their less gritty counterparts. Teachers who were grittier were more likely to finish the school year and also performed better than less gritty teachers (Robertson-Kraft & Duckworth, 2014). More grit predicted higher GPAs, more extracurricular activities, and less television watching among middle and high school students (Duckworth & Quinn, 2009) and has been related to attaining higher education levels, to fewer career changes, and to undergraduates at a competitive university earning higher GPAs (Duckworth et al., 2007). Thus, better understanding grit’s underlying factors as well as the processes necessary to create it may help to improve adolescents’ chances of success.

John Bowlby (1973), the originator of attachment theory, defined an attachment bond as a relationship that serves a dual function: that of providing a safe haven, or a place of comfort, during times of threat and distress and a “secure base” (Ainsworth et al., 1978) from which an individual springboards to explore the environment when feeling unthreatened (Bowlby, 1973). A child organizes his or her behaviors within this relationship in order to assure himself of safety. A caregiver is either available or unavailable to a child. Based on a caregiver’s responsiveness, the child develops paradigms, or cognitive representations, of the self in relationships and these representations guide future interactions throughout the lifecycle.

Bowlby proposed the attachment bond to be a universal need, such that all human beings are born predisposed to forming this relationship. However, not all people form the same attachment bond. Individual differences exist and these differences are based on the unique transactional experiences occurring within the attachment dyad. Bowlby divided the attachment bond into two types: secure and insecure. Later, Mary Ainsworth, Bowlby’s colleague (1978), expanded the classifications into three: secure, resistant/ambivalent, and avoidant. In the Strange Situation, a laboratory study, infants who are securely attached are able to tolerate brief separations from a caregiver, showing signs of missing the caregiver during separation and actively greeting the caregiver when reunited. Resistant/ambivalent infants are preoccupied with the parent throughout separation and are unable to settle down when the parent returned. Avoidantly attached infants focus on the environment during separation and show little emotion toward the caregiver upon reunion (Ainsworth & Bowlby, 1991). Mary Main introduced a fourth classification, that of disorganized attachment (Main & Solomon, 1990). Here, there was no observable pattern in the child’s reactions to the separation or to the reunion. These four behavior patterns translate into adult attachment styles.

The field of attachment has different conceptualizations and evaluations of individual adolescent and adult attachment. In 1988, Hazan and Shaver (1987) initiated the study of adult romantic attachment. Social psychologists, building on this research, have concluded that adult attachment consists of two dimensions: Anxiety and Avoidance (Shaver & Mikulincer, 2002). Those with attachment anxiety fear abandonment or rejection, feel excessively upset when a partner is unavailable, have an overwhelming need for approval, and use hyperactivating strategies to negotiate their distress. Those with attachment avoidance fear intimacy and relying on others, are overly independent and self-reliant, and utilize deactivating affect-regulation strategies to deal with distress (Shaver & Mikulincer, 2009). Bartholomew and Horowitz (1991) furthered the study of adolescent and adult attachment by postulating a four category classification of attachment styles existing along the two-dimensional space defined by attachment-related anxiety and avoidance. These four types can be conceptualized as a quadrant, where those with high attachment anxiety (termed resistant/ambivalent in the Strange Situation) or avoidance (termed avoidant in the Strange Situation) are understood as being insecurely attached, those with low levels of attachment anxiety and avoidance are assumed to be securely attached, while those with both high levels of anxiety and avoidance are seen as fearfully attached (similar to what is termed disorganized in the Strange Situation) (Shaver & Mikulincer, 2008). This line of thinking has led to the development of many self-report instruments measuring adolescent and adult attachment patterns.

Recently, another thrust in adolescent and adult attachment research examines the secure base script. As stated above, an attachment relationship provides a secure base from which to explore the environment and a safe haven to return to in times of distress. Lately, researchers have hypothesized that an individual’s heuristic that is developed within the attachment bond is organized or represented as a secure base script. This theory is based on Bretherton’s (1987) work suggesting that experiences are encoded as cognitive structures, or summaries of commonalities in events that occur in particular situations. These summaries result in a script, or a map that guides expectations and behavior in the future (Waters & Waters, 2006). With respect to attachment, these secure base scripts are thought to be the cognitive building blocks of attachment representations (Coppola, Vaughn, Cassibba, & Constantini, 2006). The scripts are assumed to be organized around the following three central affect-regulating strategies: the recognition and manifestation of distress, engaging and resolution-seeking, and soliciting support (Shaver & Mikulincer, 2002). Thus, when a stressor is encountered, presumably the secure base script is activated and the individual will expect script-consistent behavior in close relationships. Also, the individual will use the script to organize his or her reactions to the stressor (Waters & Waters, 2006). Hence, secure individuals feel confident that either they can rely on attachment figures for support or that they can solve their own distress. Insecurely attached individuals will be uncertain of the attachment figure’s availability and will doubt his or her ability to resolve the issue at hand. Research supports the fact that early caregiving experiences manifest in individual differences in secure base script knowledge and that the script assessment can measure this abstraction (Steele et al., 2014). Secure base scripts are linked with attachment in adolescents and adults (Dykas, Woodhouse, Cassidy, & Waters, 2006; Waters & Rodrigues-Doolabh, 2001) and measures such as the Adolescent Script Assessment are based on this understanding of attachment patterns.

As stated previously, grit is defined as maintaining passion and perseverance when working toward long-term goals. Meeting these goals often entails competently responding to setbacks, adversity, and failures (Duckworth et al., 2007). Presumably such stamina requires effectively self-regulating the stress and negative emotions that may ensue from the encountered difficulties. According to Eisenberg and colleagues (2014), self-regulation is defined as the modulation of emotional, behavioral, and attentional responses in the service of a desired goal. In fact, grit and self-regulatory capacities are inextricably linked. Grit has been found to be highly correlated with self-control (r = .63, p < .001) (Duckworth et al., 2007; Duckworth & Seligman, 2017), a construct involving the capacity to control, inhibit or override thoughts, emotions, impulses and behaviors in order to maximize future gains MacCann & Roberts, 2010; Tangney, Baumeister, & Boone, 2004) and to correlate with aspects of self-regulation (Rojas et al., 2012; Wolters Y Hussain, 2015), a facet of self-control.

As stated above, this research was based on the supposition that attachment and grit are related constructs. According to attachment theory, children learn to self-regulate through caregiver-child interaction (Diamond & Fagundes, 2008; Kopp, 1982). Indeed, research has confirmed a strong link between dyadic regulation within the attachment relationship in infancy and later self-regulation functioning in childhood into adulthood (Sroufe et al., 2005). From a theoretical perspective, the attachment system can be viewed as a psychobiological regulator of stress whereby the child utilizes a “safe haven” to achieve homeostasis and from there, a “secure base” to launch from and continue with exploration. Bowlby defined three interlocking behavioral systems; the attachment system, the exploratory system, and the fear system. The activation of the fear system disables the exploration system and enables the attachment system. A frightened child seeks proximity to an attachment figure (attachment system) and temporarily curtails exploration until he or she feels secure once more. Activation of the attachment system usually instigates behaviors to try to either reestablish contact with the caregiver or to ensure the attachment figure’s availability (Kobak et al., 2006). The availability, or the cognitive appraisal of a caregiver’s availability, is crucial in a child’s feeling securely attached (Bowlby, 1973; Kobak, 1999). It is also essential in a child’s developing ultimate autonomy. According to Ainsworth, the exploration system allows the learning process to occur (Ainsworth et al., 1978). A child learns self-regulation through repeated interactions with an available caregiver who encourages exploration. We hypothesize that with the development of self-regulation comes the capacity to persist and persevere through adversity and to work toward meeting long-term goals, or to be gritty. To summarize, this research, though exploratory in nature, is intended to add to the growing body of research examining how to potentially enable change and improve passion, persistence, and perseverance in low-income adolescents when pursuing long-term goals. Understanding the connection between attachment and grit in a disadvantaged population of high school seniors could perhaps contribute to efforts addressing the socioeconomic gaps in school achievement and completion.

The following were hypothesized:

1. There is an association between attachment pattern (operationalized in this study as Attachment-related Anxiety, Attachment-related Avoidance, and ASA score) and Total Grit Score, where those who are less anxious and avoidant and those who are more secure have more grit.

2. Both factors that comprise Total Grit, Grit Consistency and Grit Perseverance, will correlate with attachment pattern (Attachment-related Anxiety, Attachment-related Avoidance, and ASA score), where those who are less anxious and avoidant, and those who are more secure have more consistency of interest and more perseverance.

1. High school in Northeast Philadelphia: This high school has Title 1 status, meaning that as per “Title 1, Part A of the Elementary and Secondary Education act, as amended (ESEA) provides financial assistance is provided to local educational agencies (LEAs) and schools with high numbers or high percentages of children from low-income families to help ensure that all children meet challenging state academic standards (http://paschoolperformance.org/Profile/6787).” One hundred percent of enrolled students are economically disadvantaged. Of the 1,210 students enrolled, .17% are American Indian/Alaskan Native, 1.24% Asian, 52.64% Black or African American, 33.64% Hispanic, and 9.5% are White. Females make up 42.64% of the student body and males make up 57.36%. For the school year of 2013–2014, the dropout rate from this high school was 9.5%, while the graduation rate was 60.10%. The high school’s academic performance score is a 38%, based on a range of scores from 0 to 100. This is a number based on calculations of data elements of standardized test scores. The calculations are a weighted average of test scores across 17 factors. The details can be found on the following website: http://paschoolperformance.org/Profile/6787.

2. High school in Camden, New Jersey. This high school is a Christian-based, private school in Camden, New Jersey, whose mission is to “equip Camden children and young adults with the skills necessary for academic achievement, life management, spiritual growth, and Christian leadership.” Entrance is based on grades and an entrance exam that measures math, reading, and vocabulary comprehension. Those who are more than three levels below grade level are not accepted. The average tuition is $170 per month. Tuition is scale-based and depends on parental income as well as number of people in a household. The average reported income for families of students in this high school was $28,600 and the median income was $23,100. In 2013–2014, there were 43 students attending the high school, 44.44% males and 55.56% females. Of the 8 seniors enrolled, 100% graduated. Fifty percent of students are Black/African American individuals and 50% are Hispanic. As this is a private school, they do not have an academic performance score.

3. Charter school in Northeast Philadelphia. This charter school was founded in 1999 by a coalition of educators and community residents committed to creating a high performing middle and high school for the students of the Kensington neighborhood. Of the student body, 77% are Hispanic, 20% are Black, 2% are White, and 1% are two or more races. Fifty-two percent of the student body is female, while 48% is male.

Participants were 42 high-school seniors from three different school programs, all serving lower socioeconomic neighborhoods. Of the 42 participants, 28 (66.67%) were male and 14 (33.33%) were female. The participants’ ages ranged from 18 to 20, with 25 (59.52%) 18-year-olds, 14 (33.33%) 19-year-olds, and three (7.14%) 20-year-olds. The mean age was 18.48. The sample included 12 (28.57%) Hispanic individuals, 23 (54.76%) Black/African Americans, 2 (4.76%) White/Caucasians, and 5 (11.90%) undeclared. Twenty-one (50%) of the participants lived in a two-parent household, 18 (42.86%) lived in a single-parent household, 2 (4.76%) lived without parents, and 1 (2.38%) had other living arrangements.

In terms of high schools attended, 9 (21.43%) students attended a high school in Northeast Philadelphia and 27 (64.29%) were enrolled in a program within that high school designed for those who had dropped out of school and then returned. This is an online, teacher-facilitated program primarily designed to help undercredited students work toward high school graduation. Two (4.76%) students attended a school in Camden, NJ, and 4 (9.52%) students attended a charter school in Northeast Philadelphia. The participation rates were very different. The school in Northeast Philadelphia had 1,210 students enrolled, as compared to the school in Camden, which had a total of 8 seniors. This explains the difference in participation rates between these two schools. Those from the Charter School participated only toward the end of data collection. Thus, the researchers did not have time to allow for more participants at this site. Twenty (47.62%) participants were unsure of their grade point average (GPA). There was a large range of grade point averages for the 22 (52.38%) subjects who reported their GPA, with the lowest average being .4 and the highest being 3.4. The GPA average of those who reported their grades was 2.60. Only three participants (7.14%) had SAT scores to report, two of 1440 and one of 1190, all out of 2400.

To recruit subjects, flyers were posted at each site on multiple bulletin boards in visible locations. The researcher also spoke to several classes about the research study and passed around a sign-up sheet with different time slots available for those who expressed interest in participating. Interested participants who saw the flyers that had been posted either signed up in a time-slot on a sheet next to the flyer or contacted the researcher in person or by phone and asked to participate in the study. Those who heard about the study through word of mouth personally approached the researcher and asked to participate in this study.

Data was gathered between March and June of 2014. Each subject completed the Adolescent Script Assessment (ASA) as well as self-report measures, including another attachment measure and a measure of grit. A researcher sat with each subject as he or she completed the questionnaires and in most cases, due to varying literacy levels, read the questionnaires to the subjects and then recorded their answers. The stories told during the ASA were audio-recorded. Because there was a researcher present for each participant, there were no questions that were left blank. It took approximately 1 to 1½ hours for each participant to fill in self-report forms and to tell stories. Each subject was given a $40 Amazon gift card for participation in the research.

Two different attachment measures were used: the ASA and the ECR-S:

1. ASA

   • a. Administration: The assessment is designed to elicit attachment-related stories from adolescents. Subjects were presented with five laminated cards, each with a different story title. On each card were twelve prompt words. Subjects were asked to construct stories using these words. Girls were asked to complete stories entitled “The Party” and “Acne,” using their mother as a character in these stories, and “Studying for an Exam” and “The Tennis Match,” using their father as a character for these stories. Boys were asked to complete the same stories with the same instructions, except “Acne” was replaced with “The Haircut.” Each of the four stories was recorded using an iPod for later verbatim transcription.

   • b. Scoring: Each ASA was transcribed and then coded. Two coders were used per ASA to ensure interrater reliability. Both coders had successfully completed reliability training for the ASA. Scores were assigned for each story. Once the coders had scores for each story, they averaged the secure base script knowledge scores across all four stories to obtain a single composite score.

     Narratives were coded using the method developed by Waters and Rodrigues-Doolabh (2001). They were scored on a 7-point scale. A prototypic secure base script uses the prompt words to describe a rich interplay between main characters and a sequence of events where exploration is encouraged and distress is effectively resolved. The highest score (7) is given to passages with extensive secure base script organization and a high degree of elaboration. The lowest score (1) is assigned to passages where prompt words are used in a haphazard way, with no secure base content (Waters & Rodrigues-Doolabh, 2001). In this sample, scores ranged from 1.06 to 5.06.

   • c. Psychometric Properties: Because this measure was recently developed, psychometric properties are still being ascertained. The ASA for adults has been shown to have adequate reliability and both construct and convergent validity have been established (Waters & Waters, 2006; Steele et al., 2014) (

2. Experiences in Close Relationships Scale -Short Form (ECR-S)

   • a. Description: The ECR-S is a 12-item, self-report questionnaire adapted from the ECR (Brennan, Clark, & Shaver, 1998) by Wei and colleagues (2007). The scale measures two relatively orthogonal continuous dimensions of attachment labeled Anxiety (6 items) and Avoidance (6 items). Respondents evaluate statements such as “It helps to turn to my romantic partner in times of need” on a scale from 1 (“strongly disagree”) to 7 (“strongly agree”). Higher scores on either subscale indicate higher levels of either attachment avoidance or anxiety. In this sample, scores for attachment-related anxiety ranged from 1.00 to 6.67, while scores for attachment-related avoidance ranged from 1.17 to 6.17.

   • b. Psychometric properties: The ECR-S is based on the ECR, a 36-item questionnaire, whose psychometric properties have been well established (Brennan et al., 1998; Shaver & Mikulincer, 2002). The ECR-S has been shown to be a reliable and valid measure, possessing stable factor structure, adequate internal consistency, test-retest reliability, and construct validity (Wei et al., 2007).

3. Grit-S

   • a. Description: The Grit-S is an 8-item, self-report scale developed by Angela Duckworth and colleagues (Duckworth & Quinn, 2009) to measure grit, defined as perseverance, tenacity, and passion for goals. A maximum score is 5, indicating high grit and the lowest score on the scale is 1, or not at all gritty. In this sample, scores for grit ranged from 2.25 to 4.50.

     The Grit-S is confirmed as a two-factor model where “consistency of interest” and “perseverance of effort” both load onto grit (Duckworth & Quinn, 2009; Duckworth, 2016). Scores for each factor were calculated. In this sample, scores for “consistency of interest” ranged from 1.00 to 4.50 and scores for “perseverance of effort” ranged from 2.75 to 5.00.

   • b. Psychometric properties: The Grit-S has been shown to have internal consistency, test-retest stability, consensual validity with informant-report versions, and predictive validity (Duckworth & Quinn, 2009).

The means, standard deviations, and Pearson correlations (evaluated with two-tailed significance tests) among the main variables of the study are shown in Table 1.

As can be seen in Table 1, Attachment-related Anxiety correlated negatively with Total Grit (r(40) = –.582, p < .01), as did Attachment-related Avoidance (r(40) = –.326, p < .05). There was not a significant relationship between Secure Base Script Knowledge, as evidenced by the ASA scores, and Total Grit.

Because Attachment-related Anxiety and Attachment-related Avoidance were slightly, although not significantly correlated with each other, an ordinary least squares regression analysis was performed, where the two attachment variables were regressed onto Total Grit. (As there was not a significant relationship between Total Grit and Secure Base Script Knowledge, the latter variable was not included in the regression model.) The results are shown in Table 2.

Attachment-related Anxiety had significant unique effects on Total Grit, while the effect that Attachment-related Avoidance had on Total Grit was not significant. The model was statistically significant, F(2,39) = 12.97, p < .001, and the two attachment variables together accounted for 39.9% of the variance in Total Grit (R = .632). For Attachment-related Anxiety, β = –.547, p < .001; for Attachment-related Avoidance, β = –.248, p = .055. Thus, Attachment-related Anxiety seems to have a negative impact on Total Grit.

As can be seen in Table 1, Attachment-related Anxiety correlated negatively with Grit Consistency (r(40) = –.512, p < .01) and with Grit Perseverance (r(40) = –.463, p < .01). There were no significant relationships between Attachment-related Avoidance and Grit Consistency or Attachment-related Avoidance and Grit Perseverance. Furthermore, there were no significant relationships between Secure Base Script Knowledge and Grit Consistency or with Grit Perseverance.

Note about the attachment measures: As can be seen in Table 1, Attachment-related Anxiety and the ASA were not significantly correlated with each other, but Attachment-related Avoidance and the ASA were significantly negatively correlated with one another (r(40) = –.317, p < .05), meaning there was an inverse relationship between Attachment-related avoidance, as measured by the ECR-S, and Secure Base Attachment, as measured by the ASA.

This study provides further evidence for a relationship between attachment patterns and grit. More specifically, the research supports this connection in a lower income high school senior sample. Results indicate that those who are insecurely attached are likely to be “less gritty”. In this study, attachment anxiety and attachment avoidance were both significantly negatively correlated with overall grit. Conversely, those who were more securely attached were more gritty and those who were more fearfully attached were less gritty. This endorses the idea that attachment bonds and the attachment patterns that develop within these relationships can affect an individual’s ability to persevere over the long haul, often despite obstacles or setbacks that may be encountered.

The relationship between attachment patterns and grit, as shown by regression analyses, provides further support for a connection between attachment patterns and an individual’s level of grit. Because this was a cross-sectional and not a longitudinal study, a potential causal relationship between attachment and grit was not testable. However, the results indicate that this is an avenue worth pursuing in the future.

As far as grit’s two factors, Grit Consistency and Grit Perseverance were inversely related to Attachment-related anxiety. Thus, results support the notion that attachment anxiety may impact both factors of grit: passion and perseverance in adolescents. This exploratory study not only provides evidence for a connection between how much grit a disadvantaged adolescent possesses and that individual’s attachment patterns, but suggests that attachment patterns may be related to an adolescent’s consistency of interest as well as his or her perseverance when working toward meeting long-term goals. Of note, two different attachment measures were used and different results were obtained with each attachment measure. Furthermore, Attachment-related Anxiety was not correlated with the ASA, while Attachment-related Avoidance was negatively related. The ASA focuses on the secure base relationship and the kinds of expectations an individual has from his or her caregivers. With the script approach, those who tell narratives with supportive caregivers are thought to be secure and those who tell narratives with unsupportive caregivers are thought to be insecure. This is in contrast to the ECR-S which is a self-report measure that assesses an individual’s own understanding of their characteristic feelings and behavioral tendencies in close relationships, mostly concerning romantic relationships. Thus, the measures are examining different aspects of attachment. This data showed a correlation between Attachment-related Anxiety and Attachment-related Avoidance and conversely a positive relationship with Secure Base Script Knowledge (as measured by the ASA) and Total Grit, but not between Secure Base Script Knowledge, as measured by the ASA, and Total Grit. Perhaps grit and the ECR-S are measuring similar dimensions of self-regulatory capacities and self-perception in relationships, while the ASA is measuring an individual’s perceived security in caregiving relationships, which in and of itself may not translate to grit.

The fact that Attachment-related Anxiety and Attachment-related Avoidance yield differing results in this study may be because the ECR-S is measuring different self-regulatory capacities. As stated, Attachment-related Anxiety is highly significantly correlated with overall Grit, as well as with the two Grit subscales: Grit Consistency and Grit Perseverance. Attachment-related Avoidance, however, correlates significantly with Total Grit, but not with either of the subscales. To reiterate, individuals who are gritty maintain a consistent interest over a long period of time and are able to persevere toward meeting long-term goals related to this interest, often in spite of obstacles and setbacks. It is likely that those with grit are able to utilize effective regulatory strategies to better negotiate the affect dysregulation that may occur with encountered setbacks. As stated above, those who are anxiously attached tend to utilize hyperactivating strategies and those who are avoidantly attached typically employ deactivating strategies when feeling dysregulated. Hyperactivating strategies consist of heightening emotion and increasing efforts to elicit attention, while deactivating strategies are developed in order to minimize emotion and minimize the attachment relationship (Cassidy, 1994). The results of this study suggest that hyperactivating strategies may interfere with maintaining a persistent interest and persevering when pursuing long-term goals, while deactivating strategies may actually be more in line with what is required to maintain consistent interest and perseverance over the long haul. Perhaps feeling flooded with emotion is more detrimental to pursuing long-term goals as compared to the shutting down of emotion, which may help with being able to persist. This is an area for future research. Attachment-related Anxiety’s not being related to the ASA, in contrast to Attachment-related Avoidance can possibly be explained by story-telling styles. Perhaps those who are avoidantly attached tell stories that are less developed and have less secure base knowledge exhibited in their stories, while those who are anxiously attached, as measured by the ECR-S, may not exhibit their attachment insecurity in relation to their caregivers, but only in relation to their feelings with regards to romantic relationships. While these hypotheses may explain the different findings, an alternative explanation is that the stories told by this particular sample were not always adequate and the coders expressed difficulty with coding because of the quality of some of the stories. Several subjects in this sample were not proficient readers and were inexperienced with telling stories and both of these factors may have contributed to insufficient stories. The ASA is a new measure that may need further development with low-income samples.

There were several limitations to this study. Firstly, the sample size is small and therefore null findings may be due to low power and not disconfirmations of hypotheses. Secondly, the sample is fairly homogenous, given that the sample was only composed of low-income, high school seniors. The results, though significant, are therefore not generalizable. Thirdly, the sample may have been inherently biased when measuring grit. The 27 students that were enrolled in the program for returning students had already demonstrated grit as they were tenaciously working toward their degree instead of discontinuing high school. Many of these students had dropped out and then returned to high school and many were in danger of dropping out, but were instead persevering toward graduation.

In terms of the measurements used, self-report measures can be problematic due to subject reporting biases. In this sample in particular, many were illiterate and had the questions read to them. This could have further biased the subjects’ responses as they may have had response bias by seeking to please the interviewer. The coefficient reliability of the ECR-10 for this sample was adequate, but far below what is usual. This is potentially due to a misinterpretation of a double negative in one of the questions. The fact that most in the sample may have misinterpreted double negatives casts doubt on other double negative questions that are prominent in other measures such as the Grit-S. As for the ASA, many subjects had difficulty generating stories and whether or not that was related to their attachment patterns or to their ability to tell stories is unclear. This could have influenced the results that were obtained.

We are hopeful that a more in depth understanding of the underlying psychological processes of grit can make a contribution to the development of more interventions to improve successful outcomes in a disadvantaged population. The results of this investigation confirm a link between grit and attachment in a disadvantaged high school senior population, where those who are insecurely attached are likely to be less gritty which suggests that an individual’s attachment style may be one possible mechanism that can affect an individual’s level of grittiness. As studies show that attachment patterns can change and that nurturing attachment relationships can be cultivated (Bakermans-Kranenburg, Van IJzendoorn, & Juffer, 2003; Bernard, Dozier, Bick, & Gordon, 2014; Lieberman & Van Horn, 2009), there is the possibility that influencing attachment patterns may effect change in levels of grit. This study provides a springboard for future research to explore the benefit of relational interventions for adolescents from disadvantaged backgrounds when pursuing long-term goals.

Author Note: I have no financial interest or benefit that will arise from the direct application of this research.