Associations Between Attachment and Verbal Fluency in Adolescents

• In Event: Attachment and Language Use in Adolescence: Integrating Research on Parent, Peer, and Romantic Partner Relationships

Abstract

Attachment narratives provide rich data regarding speakers’ mental representations of their attachment relationships. Typically, attachment narratives are analyzed using a top-down, narrative coherence coding process, yielding valuable information regarding the way people organize and make meaning of relationship-relevant memories, emotions, and cognitions. Variability in the way adolescent speakers describe their relationships on semi-structured attachment interviews predicts psychopathology (Grant et al., 2003), social functioning (Van Meter et al., 2019), and emotion regulation (Berger et al., 2005), concurrently as well as prospectively into adulthood years.

Yet, these attachment narratives also contain a wealth of additional data that remain untapped. Natural language provides information about a speaker’s vocabulary, grammar, social skills, fluency, and intelligibility. In the current project, using a methodology and theoretical framework borrowed from speech and language science (Hill et al., 2020), we examine verbal disfluency as a correlate of attachment security, arguing that adolescents, for whom accessing information regarding their attachment relationships requires less effort, attachment narratives will produce narratives containing fewer disfluencies. In addition, fillers typically increase during adolescence, pragmatically serving to “hold the speaking turn” in conversations as cognition develops (Laserna et al., 2014). Conversely, disfluent speech may occur when cognitive resources are limited during emotional regulation and cognitive-linguistic tasks (Roche & Arnold, 2018).

We collected data from adolescents (N=200, Mage=15.31, SDage=1.27, 28.5% male), half of whom were hospitalized, in an inpatient unit, with half from a matched community control group, all whom reported on their attachment security to their mothers and fathers using the Kerns Security Scale (Van Ryzin & Leve, 2012). Adolescents provided a language sample by speaking about their attachment relationships using the Childhood Attachment Interview (CAI; Shmueli-Goetz, 2014). First, (H1) we anticipated that non-hospitalized adolescents would use fewer disfluencies than the hospitalized adolescents. Next, (H2) we hypothesized that attachment security would be inversely related to fillers, empty words, phrase repetitions, and total disfluencies. These were examined separately among the clinical groups.

(H1) After accounting for age and gender, the results of an ANCOVA indicated there was no difference between non-hospitalized and hospitalized adolescents’ use of empty words (Mdiff=.006, p=.069 ), phrase repetitions (Mdiff=.000, p=.773) and total disfluencies (Mdiff=.009, p=.277). However, there was a significant difference between non-hospitalized and hospitalized adolescents' use of fillers (Mdiff=.009, p=.003). Next, (H2), among non-hospitalized adolescents, greater security within their relationships with their fathers, specifically in dependability, is associated with lower frequency of fillers (b=-.010, p=.006), phrase repetitions (b=-.001, p=.049), total disfluencies (b=-.022, p=.012), and marginally with empty words (b=-.007, p=.58). In contrast, among hospitalized adolescents there were no associations between security within their relationships with their fathers and any of the language variables (p>.05). Likewise, there was no association of any of the language variables with their perceived relationship with their mothers in either clinical group. In future analyses, we will examine speech fluency in specific sections of the CAI interview for their association with attachment security.

Authors

• Abby M Rozenberg, University of California - Irvine
  Presenting Author

• Carla Sharp, University of Houston
  Non-Presenting Author

• Jessica L Borelli, University of California - Irvine
  Non-Presenting Author