Academic journal article North American Journal of Psychology

Attachment Style and Decoding of Nonverbal Cues

Academic journal article North American Journal of Psychology

Attachment Style and Decoding of Nonverbal Cues

Article excerpt

The present study examined individual differences in adult attachment style and nonverbal decoding accuracy. It was predicted that secure and preoccupied participants (i.e., those with a positive view of others) would demonstrate more nonverbal accuracy than would those with a negative view of others (i.e., dismissing and fearful participants). Participants were 59 single, female college students with a mean age of 19.20 years. Attachment styles were categorized using Bartholomew and Horowitz's two-dimensional model (1991) and nonverbal decoding accuracy was assessed through four subtests of the Diagnostic Analysis of Nonverbal Accuracy-2 (DANVA-2) (Nowicki, 2004). Thirty-six students were identified as "positive other" and 23 as "negative other." These groups did not differ in the nonverbal decoding of faces (adult or child) or child voices, p's > .05. However, there were significant group differences in the decoding of adult voices, t (57) = 2.24, p < .02 and total nonverbal accuracy scores, t (57) = 2.48, p < .01, with the "positive other" group demonstrating more nonverbal accuracy than the "negative other" group.

Largely influenced by early parent-child relationships, children develop characteristic styles of attachment that persist into adulthood and influence interpersonal relationships (Hazan & Shaver, 1987). Bartholomew and Horowitz (1991) have proposed a two-dimensional model representing four possible styles of attachment based on views of the self (positive vs. negative) and views of other people (positive vs. negative). The four attachment styles are secure (positive self, positive other), preoccupied (negative self, positive other), dismissing (positive self, negative other), and fearful (negative self, negative other). Attachment and attachment styles are important constructs for exploring adult functioning because insecure styles of attachment have been shown to be associated with psychopathology and dysfunctional attitudes (Roberts, Gotlib, & Kassel, 1996), as well as stress-related coping styles (Mikulincer & Florian, 1998).

Because the attachment system begins in infancy, early attachment relationships and patterns of emotional communication may differentially impact the skills children develop in the processing of nonverbal cues. Over the course of development, different levels of nonverbal skills are likely to be associated with specific styles of attachment, and this variability in skills may, in turn, impact adult relationships. Magai, Hunziker, Mesias, and Culver (2000) noted that our early socialization and attachment experiences impact how we express and interpret emotions. Niedenthal, Brauer, Robin, and Innes-Ker (2002) suggested that facial expressions of emotion are particularly critical for providing information that promotes attachment-related goals, and the individual's style of attachment shapes the type of information that is decoded. In other words, differences in communication skills, particularly nonverbal skills, during the course of development may help to mediate the relationship between attachment history and adult differences in interpersonal success. However, research is limited regarding the relationship between attachment styles and nonverbal information processing.

Tucker and Anders (1998) found that a secure style of attachment was associated with the expression of positive nonverbal behaviors indicative of closeness in dating couples. These same authors (Tucker & Anders, 1999) also reported that males with an anxious attachment style were less accurate in interpreting the feelings of their dating partners. Magai et al. (2000) examined the relationship between attachment and the decoding of facial expressions as part of a larger study on attachment and emotional biases. They showed that specific styles of attachment were associated with biases for processing particular emotions; for example, a secure style was associated with an increased likelihood of identification of the emotion of shame. However, they did not provide information on the basic accuracy of decoding and did not address other forms of nonverbal communication such as tones of voice. Furthermore, the average age of their participants was 63, so the results may not generalize to other age groups. Niedenthal et al. (2002) examined attachment style differences by examining the point at which a facial expression was perceived as shifting from a happy, sad, or angry face to a neutral facial expression in a computerized task. They found attachment style differences in the perception of emotional shifts in facial expressions. However, again these authors did not provide information regarding attachment style differences in basic nonverbal accuracy and limited their study to facial expressions of emotion.

The present study examined individual differences in attachment style and nonverbal decoding accuracy including both facial expressions of emotion and tones of voice with both child and adult stimuli. Using Bartholomew and Horowitz's (1991) attachment scheme, it was hypothesized that higher levels of nonverbal decoding accuracy would be found among individuals with secure or preoccupied attachment styles, as these persons maintain positive views of other people. In contrast, individuals with dismissing or fearful attachment styles maintain a negative view of others and are less likely to accurately "read" nonverbal expressions of emotion. It was therefore predicted that secure and preoccupied participants (i.e., those with a positive view of others) would demonstrate more accuracy in nonverbal decoding tasks than would participants with a negative view of others i.e., dismissing and fearful participants.



Participants were 59 single, female students attending a small liberal arts college for women in a large southeastern city. The students ranged in age from 17 to 26 (M = 19.20, SD = 1.34). The majority of the participants were Caucasian (73%) while 20% were African-American, and 7% were Hispanic, Asian, or indicated "other" with regard to racial identity. With regard to their present relationship status, 46% were not currently dating or involved in a serious romantic relationship, 41% indicated they were participating in a serious relationship, and 14% reported they were casually dating.


Diagnostic Analysis of Nonverbal Accuracy-2 (Nowicki, 2004; Nowicki & Duke, 1994). Participants were individually administered four subtests from the Diagnostic Analysis of Nonverbal Accuracy Test (DANVA-2) to assess the ability to accurately decode nonverbal cues. The adult and child faces tasks each consist of 24 color photographs of adult or child faces, representing happy, sad, angry, or fearful expressions of emotions. Facial expressions include an equal number of male and female faces and high and low intensity faces. The voice subtests involve two audiotapes; one presents 24 adult voices while the second has 24 recordings of children's voices. In these subtests the speaker recites the sentence, "I am going out of the room now and I will be back later" in a happy, angry, sad, or fearful tone of voice. Nowicki and Duke (1994) reported that performance on the original DANVA improves with age and is related to indices of social adjustment such as an internal locus of control, self-esteem, and classroom sociometric status, but not to measures of intelligence.

Over 250 research articles have used the DANVA and updated DANVA-2 to assess nonverbal skills. Nowicki (2004) has reviewed the research regarding the reliability and validity of the DANVA-2. For the adult faces subtest, internal consistency (coefficient alpha) ranged from .77 to .90 in college student samples with a test-retest reliability of .81-.84 over a one or 2-month interval (McIntire, Danforth, & Schneider, 1997; Nowicki & Carton, 1993). McIntire et al. (1997) reported that this subtest was significantly correlated with the Japanese and Caucasian Facial Expressions of Emotion Test for a sample of college students, r = .80 and this subtest is significantly related to the original DANVA, r = .54 (Nowicki & Carton, 1993). McIntire et al. (1997) found the DANVA adult faces negatively correlated with emotional control and Carton, Kessler, and Pape (1999) found this subtest was associated positively with well-being and negatively with depression.

Nowicki and Carton (1993) provided data on the child facial expressions subtest. Cronbach's coefficient alpha was .74 in 102 college students with a 2-month test-retest reliability of .88. This subtest was significantly related to the original DANVA subtest with this same sample of college students (r = .44) and scores on this subtest predicted higher achievement for second graders and social popularity in elementary school children (Rowe, 1993, cited in Nowicki & Carton, 1993).

The adult tones of voice subtest was studied in depth by Baum and Nowicki (1998). Nowicki (1995, cited in Baum & Nowicki, 1998) reported a 6-week test-retest reliability of .83 and a coefficient alpha of .78 with college students. Errors on this subtest were related to higher levels of external locus of control (Nowicki, 1995, cited in Baum & Nowicki, 1998) among college students and Baum and Nowicki (1998) concluded that errors on this subtest were related to a variety of measures of maladjustment for children and adults.

The child tones of voice subtest was evaluated and reviewed by Rothman and Nowicki (2004). Cronbach's coefficient alpha was .73 for children and .73 for 50 young adults (Carney, 1999, cited in Rothman & Nowicki, 2004). Test-retest reliability was .78 over a one-month interval (Nowicki, 1997, cited in Rothman & Nowicki, 2004). In an examination of this subtest in a variety of samples from preschool to adult age, Rothman and Nowicki demonstrated that this test is reliable over time and significantly correlated with measures of interpersonal functioning.

The number of correct responses on each subtest may vary from 0 to 24 with higher scores corresponding to more accuracy in nonverbal decoding. A total nonverbal accuracy (0-96) score also was computed by combining accuracy scores from each of the four subtests.

Relationship Styles Questionnaire. (RSQ; Bartholomew & Horowitz, 1991) The RSQ is a two-dimensional, four-category attachment style model used to determine whether the participant's attachment style was secure, dismissing, preoccupied, or fearful. The secure and preoccupied subtypes both have a positive view of other people while the dismissing and fearful subtypes both have a negative view of others. The secure and dismissing subtypes share a positive self-view and the preoccupied and fearful subtypes share a negative self perspective. Four brief paragraphs, corresponding to each of the four styles, were presented and participants indicated the one that best described their behavior in close relationships.

Bartholomew and Horowitz (1991) predicted and found that measures of sociability distinguished the secure and preoccupied groups from the fearful and dismissing participants, while self-concept measures distinguished the secure and dismissing styles from the preoccupied and fearful styles. These distinctions were evident across multiple sources including interviews, self-reports and peer-reports. Griffin and Bartholomew (1994) also used multiple sources of measurement to confirm the two-dimensional attachment model. This categorical measure of attachment has been found to correspond favorably with dimensional, continuous models of attachment (Lopez & Gormley, 2002) and there is evidence for moderate stability over time in self-reported attachment style (Davila, Burge, & Hammen, 1997).


After completing basic demographic information, students categorized their own attachment style using Bartholomew and Horowitz's (1991) four-category scheme. Participants were then individually administered four subtests of the Diagnostic Analysis of Nonverbal Accuracy (DANVA-2; Nowicki, 2004). Two groups of students were created based on their positive vs. negative view of others within their style of attachment. Consistent with the conceptualization provided by Batholomew and Horowitz (1991), secure and preoccupied respondents were categorized as having a "positive other" view in their relationships while dismissing and fearful participants were categorized as having a "negative other" view in interpersonal relationships.


Thirty-six percent of the students categorized themselves as secure, 12% as dismissing, 25% as preoccupied, and 27% as fearful. Based on these basic attachment styles, 36 students (61%) were identified as "positive other" (secure or preoccupied attachment) and 23 (39%) were identified as "negative other" (dismissing or fearful attachment). Table 1 presents the means and standard deviations for the number of correct identifications for each of the nonverbal subtests and the nonverbal total for participants in these two attachment clusters. Although all the group means were in the predicted direction, one-tailed t-tests showed that these groups did not differ in their nonverbal decoding of adult faces, t (57) = 1.33, child faces, t (57) = .88, or child voices, t (57) = 1.51, p's > .05. However, there were significant group differences in the decoding of adult voices, t (57) = 2.24, p < .02 and in the overall total DANVA-2 score, t (57) = 2.48, p < .01, with the "positive other" group of students demonstrating higher nonverbal accuracy scores than the "negative other" group.


These findings suggest that attachment styles may be associated with specific patterns of nonverbal skills. Participants with a positive view of other people, those who identified themselves as secure or preoccupied in their attachment styles, were more accurate in the decoding of adult voices and in their total nonverbal accuracy score than those participants with a negative view of others, i.e., those who identified themselves as dismissing or fearful in their attachment style. These findings support the "positive/negative other" scheme used by Bartholomew and Horowitz (1991) to categorize styles of attachment. As predicted, those who have a more positive perspective of other people were found to have greater accuracy overall in reading nonverbal cues, especially voice tones, sent by their adult peers. Early parent-child relationships may contribute to better nonverbal decoding skills via early socialization that impacts the understanding of emotions (Magai et al., 2000). Children who attend to the emotions of others may also develop positive views of other people.

However alternative explanations of the relationship between attachment styles and nonverbal decoding also need to be considered, since other attributes may vary along with styles of attachment. For example, individuals with a preoccupied style of attachment may be higher in attentiveness as they are highly sensitive to issues of rejection and abandonment. Mood states such as depression and anxiety also correlate with various attachment styles and may impact processing of interpersonal information (Van Buren & Cooley, 2002). The relationship between nonverbal decoding and attachment styles could be more indirect than might be suggested by these data. Those with a positive view of others may differ from those with a negative view in other personality traits that may more directly impact the accuracy of nonverbal processing. Thus, additional research is needed to examine the relationship between attachment styles and nonverbal decoding.

This proposed association between nonverbal accuracy and attachment style was evident primarily for the processing of tones of voice and overall nonverbal accuracy. Varying tones of voice are communicated even when people attempt to "hide" their true facial expressions, as voice tones are more difficult to control or inhibit (Rosenthal, Hall, diMatteo, Rogers, & Archer, 1979). Therefore differences in the ability to process tones of voice may have a particularly important impact on interpersonal relationships. However, other researchers have also found attachment style differences in the processing of facial expressions of emotion (Magai et al., 2000; Niedenthal et al., 2002). Task and stimuli differences may explain varying results across studies. For each of the four DANVA tasks, the group differences were in the expected direction with "positive other" participants demonstrating higher mean accuracy scores than "negative other" participants. A larger sample size may show that processing of both tones of voice and facial expressions differed among these groups.

Another explanation for the differing results for tones of voice and facial expressions reflects the timing of the facial affect stimuli. In actual interactions, facial expressions are often fleeting and may last for a shorter period of time than the one-second used in the current administration of the DANVA-2. Future research using a more rapid mode of presentation may demonstrate processing differences. Those with a "negative" view of others may be less accurate at processing emotional faces when the facial affect stimuli are presented more rapidly, thus requiring greater skill.

These findings are, of course, limited in scope by the nature and size of the sample that included only college-aged women at a small liberal arts college. These results may not extend to men and may not apply to other groups of participants. It was interesting that only 36% of this sample reported a "secure" style of attachment while other researchers have reported that 50 or 60% of their participants were "secure" (e.g., Bartholomew & Horowitz, 1991; Hazan & Shaver, 1987). These differences suggest that this sample may not be typical. Replication of these findings with other samples is clearly indicated.

More research is needed to study nonverbal correlates of different attachment styles. As we come to understand the skills and behaviors that may accompany different attachment styles, we may develop more opportunities for changing patterns of relationships at the level of basic skills and information processing. Individuals with less-than-ideal attachment histories may be taught to improve their skills and this skill development may later contribute to more productive interpersonal relationships.

Note: The author would like to thank Amrita Dhamoon and Alison Maclean for their assistance in data collection and Steve Nowicki, Jr. for his support of this project.


Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244.

Baum, K. M., & Nowicki, S. Jr. (1998). Perception of emotion: Measuring decoding accuracy of adult prosodic cues varying in intensity. Journal of Nonverbal Behavior, 22, 89-107.

Carton, J. S., Kessler, E. A., & Pape, C. L. (1999). Nonverbal decoding skills and relationship well-being in adults. Journal of Nonverbal Behavior, 23, 91-100.

Davila, J., Burge, D., & Hammen, C. (1997). Why does attachment style change? Journal of Personality and Social Psychology, 73, 826-838.

Griffin, D., & Bartholomew, K. (1994). Models of the self and other: Fundamental dimensions underlying measures of adult attachment. Journal of Personality and Social Psychology, 67, 430-445.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511-524.

Lopez, F. G., & Gormley, B. (2002). Stability and change in adult attachment style over the first-year college transition: Relations to self-confidence, coping, and distress patterns. Journal of Counseling Psychology, 49, 355-364.

Magai, C., Hunziker, J., Mesias, W., & Culver, L. C. (2000). Adult attachment styles and emotional biases. International Journal of Behavioral Development, 24, 301-309.

McIntire, K. A., Danforth, M. M., & Schneider, H. G. (1997). Measuring cue perception: Assessment of reliability and validity. Poster presented at the Southeastern Psychological Association Annual Meeting, Atlanta, GA.

Mikulincer, M., & Florian, V. (1998). The relationship between adult attachment styles and emotional and cognitive reactions to stressful events. In J. A. Simpson, & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 143-165). New York: Guilford Press.

Niedenthal, P. M., Brauer, M., Robin, L., & Innes-Ker, A. H. (2002). Adult attachment and the perception of facial expression of emotion. Journal of Personality and Social Psychology, 82, 419-433.

Nowicki, S. (2004). The Diagnostic Analysis of Nonverbal Accuracy-2. Unpublished manuscript, Emory University, Atlanta, GA.

Nowicki, S. Jr., & Carton, J. (1993). The measurement of emotional intensity from facial expressions. The Journal of Social Psychology, 133, 749-750.

Nowicki, S. Jr., & Duke, M.P. (1994). Individual differences in the nonverbal communication of affect: The Diagnostic Analysis of Nonverbal Accuracy Scale. Journal of Nonverbal Behavior, 19, 9-35.

Roberts, J. E., Gotlib, I. H., & Kassel, J. D. (1996). Adult attachment security and symptoms of depression: The mediating roles of dysfunctional attitudes and low self-esteem. Journal of Personality and Social Psychology, 70, 310-320.

Rosenthal, R., Hall, J. A., diMatteo, M. R., Rogers, P. L., & Archer, D. (1979). Sensitivity to nonverbal communication: The PONS test. Baltimore: John Hopkins University Press.

Rothman, A. D., & Nowicki, S. Jr. (2004). A measure of the ability to identify emotion in children's tone of voice. Journal of Nonverbal Behavior, 28, 67-92.

Tucker, J. S., & Anders, S. L. (1998). Adult attachment style and nonverbal closeness in dating couples. Journal of Nonverbal Behavior, 22, 109-124.

Tucker, J. S., & Anders, S. L. (1999). Attachment style, interpersonal perception accuracy, and relationship satisfaction in dating couples. Personality and Social Psychology Bulletin, 25, 403-412.

Van Buren, A., & Cooley, E. L. (2002). Attachment styles, view of self, and negative affect. North American Journal of Psychology, 4, 417-430.

Author info: Correspondence should be addressed to Eileen L. Cooley, Department of Psychology, Agnes Scott College, Decatur, GA 30030,

North American Journal of Psychology, 2005, Vol. 7, No. 1, 25-34

Eileen L. Cooley

Agnes Scott College

TABLE 1 "Positive Other" vs. "Negative Other" Attachment Styles and
Nonverbal Accuracy Scores

                                 Mean       SD      t value

Total DANVA Score
Positive Other        n = 36     81.39     4.63
Negative Other        n = 23     77.39     7.83      2.48 *

Adult Faces
Positive Other        n = 36     19.94     2.34
Negative Other        n = 23     18.83     3.86      1.33

Child Faces
Positive Other        n = 36     22.22     2.13
Negative Other        n = 23     21.78     1.76      0.88

Adult Voices
Positive Other        n = 36     19.08     1.86
Negative Other        n = 23     17.48     3.79      2.24 *

Child Voices
Positive Other        n = 36     20.14     1.82
Negative Other        n = 23     19.30     2.40      1.51

* p < .05             One-tailed
Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.