- Email: [email protected]
Attachment styles, interpersonal relationships and psychotic phenomena in a non-clinical student sample
Personality and Individual Differences 41 (2006) 707–718 www.elsevier.com/locate/paid
Attachment styles, interpersonal relationships and psychotic phenomena in a non-clinical student sample Katherine Berry *, Alison Wearden, Christine Barrowclough, Tom Liversidge School of Psychological Sciences, University of Manchester, Rutherford House, Manchester Science Park, Lloyd Street North, Manchester M15 6SZ, United Kingdom Received 28 October 2005; received in revised form 4 March 2006; accepted 16 March 2006 Available online 6 May 2006
Abstract This paper describes the development and initial validation of a measure of adult attachment styles based on Bartholomew’s (1990, 1997) model and adapted for use with individuals with psychosis. The associations between the new measure, interpersonal problems, self-concept, relationships with significant others during childhood, traumatic life events and non-clinical psychotic phenomena were examined in a student sample. Three hundred and twenty three students completed study measures via an internet website. In support of the construct validity of the measure, insecure attachment styles were associated with interpersonal problems and anxiety in attachment relationships was associated with low self-esteem. There were weak but significant associations between insecure attachment and more negative experiences of being parented. There was no evidence of associations between attachment and traumatic life events. As predicted, there were significant associations between insecure attachment and non-clinical psychotic phenomena; positive psychotic phenomena were associated with anxiety in attachment relationships, and social anhedonia with avoidance in attachment relationships. Findings suggest that the attachment measure is a valid instrument with good internal consistency, and can be used to explore associations between attachment styles and psychotic symptoms in clinical samples. Ó 2006 Elsevier Ltd. All rights reserved.
*
Corresponding author. Tel.: +44 0 161 275 8498; fax: +44 0 161 275 8487. E-mail address: [email protected] (K. Berry).
0191-8869/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2006.03.009
708
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Keywords: Adult attachment; Analogue sample; Interpersonal problems; Psychosis
1. Introduction Recent models of psychosis implicate earlier experiences with significant others and interpersonal traumas in the development and maintenance of psychotic symptoms (Garety, Kuipers, Fowler, Freeman, & Bebbington, 2001). Since the seminal work of Bowlby (1982) on early childhood attachments, much research has suggested that earlier interpersonal experiences influence attachments to others in adulthood, and adult attachment styles predict interpersonal functioning, self-esteem and psychopathology (Bartholomew & Horowitz, 1991; Platts, Tyson, & Mason, 2002). Psychosis is associated with interpersonal difficulties (Penn et al., 2004) and low self-esteem (Barrowclough et al., 2003). Attachment theory may therefore provide a useful framework for conceptualizing the influence of interpersonal relationships on the development and course of psychosis (Dozier, Stovall, & Albus, 1999). There is, however, currently limited empirical research investigating attachment styles and their correlates in clinical samples (Goodwin, 2003) and there are difficulties assessing the attachment styles of patients with psychosis with available measures. Interview attachment measures, such as the Adult Attachment Interview (AAI; Main & Goldwyn, 1984) assess attachment states of mind on the basis of the individual’s coherence of narrative in describing parenting experiences. Such measures are time consuming to complete, require extensive training, and, when administered to individuals with psychosis, may be confounded by the presence of psychotic symptoms (Dozier et al., 1999). Self-report attachment measures, although easier to administer, tend to focus on close interpersonal relationships with some items referring specifically to romantic relationships. They may therefore be less relevant to individuals with psychosis, who are often socially isolated (Randolph, 1998). Cognitive problems associated with psychosis may also make it difficult for patients to understand negatively worded items and rate themselves on scales with relatively wide-ranging Likert scales and few anchor points (Kelly, Sharkey, Morrison, Allardyce, & McCreadie, 2000). There is therefore a need to develop and validate a measure of attachment that is more appropriate for use with this group. Two dimensions have been found to underlie self-report measures (Brennan, Clark, & Shaver, 1998), which can be conceptualized in cognitive terms as: ‘model of self’ and ‘model of other’; or affective and behavioural terms as: ‘anxiety’ and ‘avoidance’ (Crowell, Fraley, & Shaver, 1999). Attachment anxiety is associated with a negative self-image and an excessive need for approval from others, coupled with a fear of rejection and abandonment. Attachment avoidance is associated with a negative image of others and is defined in terms of either an excessive need for selfreliance or a fear of depending on others. Bartholomew (1990, 1997) has also described these dimensions in terms of four prototypes: secure; fearful; dismissing; and preoccupied (see Fig. 1). Preoccupied and fearful attachment styles, which are characterized by high anxiety in attachment relationships, have been associated with low self-esteem (Bartholomew & Horowitz, 1991; Collins & Read, 1990). Dismissing attachment, characterized by low anxiety but high avoidance, has been associated with hostility, interpersonal coldness and emotional detachment. Preoccupied attachment, characterized by high anxiety and
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
709
MODEL OF SELF (ANXIETY) Positive (Low)
SECURE Positive (Low)
High self-worth, believes that others are responsive, comfortable with autonomy and in forming close relationships with others.
Negative (High)
PREOCCUPIED A sense of self -worth that is dependent on gaining the approval and acceptance of others.
MODEL OF OTHER (AVOIDANCE) DISMISSING
Negative (High)
Overt positive self-view, denies feelings of subjective distress and dismisses the importance of close relationships.
FEARFUL Negative self-view, lack of trust in others, subsequent apprehension about close relationships and high levels of distress.
Fig. 1. Bartholomew’s (1990) model of attachment (adapted from Bartholomew and Horowitz, 1991, p. 227).
low avoidance, has been associated with over intrusive behaviour, dominance and aggression. Fearful attachment, characterized by both high anxiety and high avoidance, has been associated with difficulties in assertiveness, social inhibition, anger and aggression (Bartholomew & Horowitz, 1991; Kobak & Sceery, 1988; Mallinckrodt, 2000). High levels of anxiety or avoidance in attachment relationships may be involved in the maintenance of specific psychotic symptoms. One study involving a non-clinical sample found associations between self-reported anxiety in attachment relationships and positive schizotypal characteristics such as bizarre beliefs and anomalous experiences, and avoidance and negative schizotypal characteristics such as withdrawal and loss of pleasure (Wilson & Costanzo, 1996). The authors found an additional relationship between attachment avoidance and positive schizotypy, and theorized, following Feeney, Noller, and Hanrahan (1994), that avoidance in attachment relationships is partially anxiety-driven. Both paranoia and avoidant attachment, are thought to be associated with negative interpersonal beliefs developed as a result of earlier interpersonal experiences (Bartholomew, 1990; Bentall, Kinderman, & Kaney, 1994). However, Wilson and Costanzo (1996) did not specifically investigate the relationship between avoidant attachment and the schizotypal characteristic of paranoid thinking. They also failed to control for the influence of negative affect which has been shown to predict the extent to which individuals report psychological symptoms (Watson & Pennebaker, 1989). The relationship between attachment and specific schizotypal symptoms therefore warrants further investigation. The use of analogue or non-clinical samples in psychosis research is becoming increasingly popular due to the recognition that psychotic symptoms are on
710
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
a continuum with normal experiences. Psychotic phenomena in non-clinical populations are therefore an interesting area of research in their own right and can inform the understanding of psychosis (Peters, Joseph, & Garety, 1999; van Os, Hanssen, Bijl, & Ravelli, 2000). The first aim of the present study is to assess the psychometric properties of a new attachment style questionnaire, the ‘Psychosis Attachment Measure’ (PAM), in an analogue sample. Construct validity will be demonstrated by associations between high levels of attachment anxiety and avoidance and interpersonal difficulties, an association between high levels of attachment anxiety and low self-esteem, and associations between high levels of attachment anxiety and avoidance and negative interpersonal experiences, including poor relationships with significant others during childhood and interpersonal traumas. The second aim of the study is to investigate the associations between attachment and non-clinical psychotic phenomena in an analogue sample. It is hypothesized that anxiety and avoidance in attachment relationships will be associated with higher levels of non-clinical psychotic phenomena. Specifically, anxiety will be associated with positive psychotic phenomena, and avoidance with negative symptom traits and paranoia. Following the validation of the measure in a large analogue sample, it can then be validated in more extensive samples with psychosis and used to develop theoretical models of interpersonal difficulties and symptoms.
2. Methods 2.1. Participants and procedure Participants were 323 undergraduate and postgraduate students at University of Manchester who responded to an e-mail inviting them to participate in a web-based study investigating ‘attachment styles, interpersonal relationships and personality’. The median age of the sample was 21 years (range = 17–67 years) and 72% were female. The majority of the sample was White British (82%). Approximately half were in a romantic relationship (46%). 2.2. Measures 2.2.1. Attachment measures The 16-item PAM was based on existing measures of attachment (Bartholomew & Horowitz, 1991; Brennan et al., 1998). Items referred to thoughts, feelings and behaviours in close interpersonal relationships, but did not refer specifically to romantic relationships. Participants were asked to rate the extent to which each item was characteristic of them using a four-point scale ranging from ‘not at all’ to ‘very much’. Preliminary data from a clinical sample of patients with a schizophrenia diagnosis (n = 27) revealed that the measure is acceptable to participants and suggested good reliability, with alphas of .96 for the anxiety subscale and .85 for the avoidance subscale. The Relationships Questionnaire (RQ; Bartholomew & Horowitz, 1991) was used to assess the concurrent validity of the PAM. It consists of four paragraphs describing secure, preoccupied, fearful and dismissing attachment styles. Respondents are asked to rate the degree to which each description characterizes them using a seven-point scale, ranging from ‘not at all like me’, to ‘very
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
711
much like me’. Scores on these four items were used to derive the two dimensions of ‘model of others’ and ‘model of self’ (Griffin & Bartholomew, 1994). The RQ has been shown to have reasonable reliability and validity (Griffin & Bartholomew, 1994). 2.2.2. Interpersonal problems The 32-item Inventory of Interpersonal Problems-32 (IIP-32; Barkham, Hardy, & Startup, 1996) assesses a range of interpersonal problems, including aggression, poor sociability and excessive dependence on others. Respondents are asked to rate how much each problem distresses them on a five-point scale, ranging from ‘not at all’ to ‘extremely’. It has good psychometric properties (Barkham et al., 1996), and the alpha in the present study was .85. 2.2.3. Self-esteem The 30-item Self-Concept Questionnaire (SCQ; Robson, 1989) assesses various aspects of selfconcept. Respondents are asked to rate items on an eight-point scale, ranging from ‘completely disagree’ to ‘completely agree’. A high total scale score indicates low self-esteem. The measure has good reliability and validity (Robson, 1989) and the alpha in the present study was .91. 2.2.4. Earlier interpersonal experiences The Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979) is a 25-item self-report measure which assesses individuals’ retrospective judgements of care and protection in their relationship with a significant caregiver during the first 16 years of life. In the present study, participants were asked to answer the questions in relation to their primary caregiver during childhood. Low scores on the care scale reflect parental neglect and rejection and high scores on the protection scale indicate excessive control and intrusive parenting. The PBI is a well-established measure (Parker et al., 1979). Alphas in the present study were .92 for the care subscale and .88 for the protection subscale. Experiences of traumatic events were assessed using a measure based on the Trauma History Questionnaire (THQ; Green, 1996), which asked participants about their experiences of a range of interpersonal traumas. 2.2.5. Positive schizotypy The 20-item paranoia scale (PS; Fenigstein & Vanable, 1992) assesses levels of paranoid thinking. The measure has good psychometric properties in non-clinical samples (Fenigstein & Vanable, 1992) and the alpha in the present study was .90. The revised version of the Launay–Slade Hallucinations Scale (LSHS; Launay & Slade, 1981) is a 24-item scale which assesses the frequency of clinical and sub-clinical hallucinatory phenomena (Morrison, Wells, & Nothard, 2000). The measure has good psychometric properties (Morrison et al., 2000) and the alpha in the present study was .90. 2.2.6. Negative schizotypy The 40-item revised Social Anhedonia Scale (SAS; Eckblad, Chapman, Chapman, & Mishlowe, 1982) assesses inability to experience positive feelings in interpersonal relationships. Total scores are derived by summing positively endorsed traits. It has good psychometrics in student samples (Eckblad et al., 1982) and the alpha in the present study was .85.
712
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
2.2.7. Affect The Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) was used to control for the influence of negative affect on other self-report measures. Respondents are asked to rate the extent to which they have experienced ten positive and ten negative affective states during the past week. The measure has good reliability and validity (Watson et al., 1988) and in the present study, the alpha for the negative affect scale was .87. 2.2.8. Data analysis Attempts were made to transform skewed data sets (Tabachnick & Fidell, 1996). As some transformations were unsuccessful, for consistency and ease of interpretation all data sets were left untransformed. Data were then analysed using both parametric and non-parametric correlations, and as the significance levels were the same in both sets of analyses only parametric findings are reported. Where there were missing data, all available data were analyzed, and the sample size for each analysis reported in the tables. Correlation coefficients were compared using procedures suggested by Meng, Rosenthal, and Rubin (1992). Due to the large number of analyses performed, a conservative p value of .01 was adopted. All analyses were conducted using SPSS version 11.
3. Results 3.1. Factor analysis The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was .85, suggesting that the pattern of correlations were relatively compact and that a factor analysis should produce distinct and reliable factors (Field, 2000). An initial exploratory factor analysis of items on the PAM extracted three factors with eigenvalues above one. A varimax orthogonal rotation was used to help interpret the findings (Field, 2000). As expected, one of the factors seemed to represent attachment anxiety and the other avoidance. One item loaded onto the third factor: ‘If I have a problem with other people, I usually let them know’. As this item did not load onto the first or second factors, a decision was made to exclude it from further analyses and run a two-factor solution. Following varimax rotation, the first factor, representing anxiety, accounted for 23.87% of the variance and the second factor, representing avoidance, accounted for 20.53% of the variance (Table 1). The use of orthogonal rotation was deemed appropriate as the factors were only weakly correlated at .22 (Tabachnick & Fidell, 1996). Inspection of the residuals suggested that the solution provided an acceptable fit to the data (Field, 2000). The internal consistency of each dimension was acceptable. Cronbach’s alphas for the anxiety and avoidance dimensions were .82 and .75, respectively. Anxiety and avoidance subscale scores were derived by averaging scores for the 9 anxiety items and 6 avoidance items (Table 1). 3.2. Concurrent validity There were significant negative correlations between RQ positivity of self-model and attachment anxiety (r = .56, p < .001) and between RQ positivity of other model and attachment
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
713
Table 1 Factor loadings for individual items Factor 1: Anxiety 3. I tend to get upset, anxious or angry if other people are not there when I need them. 5. I worry that key people in my life won’t be around in the future.
Factor 2: Avoidance .67 .60
6. I frequently ask other people to reassure me that they care about me. 7. If other people disapprove of something I do, I get very upset. 9. I frequently wonder whether I can trust other people.
.70
10. I find it hard to believe that other people will be there for me if I need them. 11. I worry that if other people get to know me better, they won’t like who I really am. 15. I believe that other people will never be concerned about me. 16. I worry that if I displease other people, they won’t want to know me anymore.
.59
.62 .54
1. I prefer not to let other people know my ‘true’ thoughts and feelings. 2. I find it easy to depend on other people for support with problems or difficult situations. (reverse item) 4. I usually discuss my problems and concerns with other people. (reverse item) 8. I find it hard to accept help from other people when I have problems or difficulties. 12. When I’m feeling stressed, I prefer being on my own to being in the company of other people. 14. I try to cope with stressful situations on my own.
.62 .66
.69 .63 .56
.71
.71 .56 .75
avoidance (r = .45, p < .001). These correlations were similar to those obtained in previous research investigating relationships between self-report measures of attachment (Feeney et al., 1994; Bartholomew & Shaver, 1998). The relationships between anxiety and self-model (r = .49, p < .001) and between avoidance and other model remained significant when controlling for the influence of negative affect (r = .46, p < .001). 3.3. Attachment, interpersonal problems, self-esteem and earlier interpersonal experiences The relationships between attachment anxiety and avoidance, interpersonal problems, self-esteem and earlier interpersonal experiences were investigated as a further test of the measure’s validity (see Table 2). Both attachment anxiety and avoidance were significantly positively correlated with the total level of interpersonal problems. However, the correlation between anxiety and interpersonal difficulties was significantly higher than the correlation between avoidance and interpersonal difficulties (z = .40, p < .01). There was a significant negative correlation between attachment anxiety and self-esteem, suggesting that participants who rated themselves as high in attachment anxiety had lower self-esteem. Participants who rated themselves as high in attachment avoidance also had lower self-esteem but, as would be predicted on the basis of Bartholomew’s (1990) model, the correlation between anxiety and self-esteem was significantly higher than the correlation between avoidance and self-esteem (z = .51, p < .01). The above relationships were all still significant when controlling for the influence of negative affect.
714
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Table 2 Correlations between attachment, psychosocial functioning, interpersonal relationships and non-clinical psychotic phenomena Scale
n
Inventory of interpersonal problems Robson self-concept questionnaire Parental bonding instrument care subscalea Parental bonding instrument protection subscale Total number of traumatic events reported Total trauma distress scores Paranoia scale Launay hallucinations scale Social anhedonia scalea
280 248 266 266 258 258 244 222 230
a
Attachment anxietya
Attachment avoidance
Anxiety partial correlations for negative affect
Avoidance partial correlations for negative affect
r
r
r
r
p .60 .60 .23 .24 .02 .07 .51 .37 .19
<.001 <.001 <.001 <.001 .73 .258 <.001 <.001 .004
p .28 .18 .20 .18 .10 .12 .18 .20 .44
<.001 .004 .001 .003 .137 .072 .004 .003 <.001
p .52 .48 .15 .15 .04 .01 .41 .27 .12
<.001 <.001 .035 .037 .574 .894 <.001 <.001 .09
p .29 .20 .13 .12 .08 .08 .19 .16 .42
<.001 .006 .075 .091 .28 .263 .006 .025 <.001
Skewed variables.
There were significant associations between attachment and memories of relationships with caregivers. Participants who were higher in both attachment anxiety and avoidance tended to rate parents as less caring and more controlling. Although the findings were statistically significant due to the large sample size, the strength of the correlations was relatively weak, and they became nonsignificant when negative affect was controlled. There were no significant relationships between attachment and the trauma measure. 3.4. Attachment and non-clinical psychotic phenomena There were positive associations between all non-clinical psychotic phenomena and insecure attachment (see Table 2). The correlation between anxiety and paranoia was significantly greater than the correlation between avoidance and paranoia (z = .38, p < .01). The correlation between avoidance and social anhedonia was significantly greater than the correlation between anxiety and social anhedonia (z = .28, p < .01). The correlation between anxiety and hallucinations was greater than that between avoidance and hallucinations. When negative affect was controlled for, the significant relationships between attachment anxiety and paranoia, and anxiety and hallucinatory experiences were maintained; the significant relationships between attachment avoidance and social anhedonia, and avoidance and paranoia were similarly maintained, but the relationship between avoidance and hallucinations no longer reached significance.
4. Discussion The study suggests that the PAM has good psychometric properties. As predicted, there were significant associations between insecure attachment and non-clinical psychotic phenomena, with
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
715
specific associations between positive psychotic phenomena and anxiety in attachment relationships and social anhedonia and avoidance in attachment relationships. The methodological limitations of the study need to be taken into consideration in interpreting the findings and will be outlined prior to discussing the results in further detail. First, the cross-sectional design of the study limits the conclusions that can be drawn concerning the causal directions of the relationships between associated variables and relationships are most likely bi-directional. Second, the sample consisted of students who were self-selected, recruited by e-mail, and contained a preponderance of female participants. It is possible that people who self-select for such studies may be more prone to psychological problems (Freeman et al., 2005). Third, the use of a questionnaire to measure attachment introduces the possibility of self-reporting biases, including social desirability biases. Arguably, the anonymity of internet research may reduce the influence of social desirability (Freeman et al., 2005), but it is not possible to rule it out completely. Finally, as the study relied solely on self-report measures the findings may be confounded by common method variance. Despite these limitations, the fact that the findings are consistent with existing literature and theoretical models, supports their validity. Consistent with current models of attachment, the factor analysis suggested that two dimensions underlie the PAM (Bartholomew, 1990, 1997). The measure appears to be a reliable assessment tool with internally consistent subscales and good concurrent validity. As expected, there were relationships between insecure attachment and interpersonal problems (Bartholomew & Horowitz, 1991; Mallinckrodt, 2000), and we were able to show that these relationships were not entirely due to negative affect. It would, however, be interesting to replicate the study using an informant measure of interpersonal problems to investigate the influence of reporting biases on interpersonal problems. Arguably, significant others may be more likely to report negative interpersonal traits than individuals themselves (Kobak & Sceery, 1988). There is also evidence of attachment-style differences in the reporting of interpersonal problems. The relatively low correlation between avoidance and interpersonal problems compared to anxiety and interpersonal problems may be due to individuals with dismissing attachment under-reporting distress (Dozier & Lee, 1995). In line with previous research and in support of the validity of the measure, there were significant relationships between attachment anxiety and self-esteem (Bartholomew & Horowitz, 1991; Collins & Read, 1990). However, there were no strong or consistent relationships between attachment and parental relationships, nor between attachment and experiences of trauma. As participants tended to report positive parental relationships and relatively few traumatic events, it may have been difficult to identify strong relationships between negative experiences and insecure attachment. The relationships between these constructs therefore warrants investigation in samples of clinical participants who are more likely to report difficulties in parental relationships (Parker, Fairley, Greenwood, Jurd, & Silove, 1982) and traumatic events (Mueser et al., 1998). There were positive associations between anxiety and avoidance in attachment relationships and schizotypal characteristics (Wilson & Costanzo, 1996). As predicted, there were significant relationships between anxiety and the positive schizotypal characteristics of paranoia and hallucinations, and between avoidance and the negative schizotypal characteristic of social anhedonia. Correlations between anxiety and positive schizotypy were significantly stronger than those between avoidance and positive schizotypy; correlations for avoidance and social anhedonia were stronger than those between anxiety and social anhedonia. As predicted on the basis of cognitive
716
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
models of paranoia, there was also a weak association between avoidance and paranoid thinking, which remained significant when controlling for the influence of negative affect. In summary, the findings of this study suggest that the PAM is a reliable and valid instrument, which can be used to assess attachment styles in analogue samples and develop theoretical models of psychosis. Investigations of the associations between attachment and psychosis in clinical samples are now warranted to assess whether the relationships between attachment and specific symptoms are replicated. Acknowledgements The authors thank all the students who participated in the research. References Barkham, M., Hardy, G. E., & Startup, M. (1996). The IIP-32: A short version of the Inventory of Interpersonal Problems. British Journal of Clinical Psychology, 35, 21–35. Barrowclough, C., Tarrier, N., Andrews, B., Humphreys, L., Ward, J., & Gregg, L. (2003). Self-esteem in schizophrenia: Relationships between self-evaluation, family attitudes, and symptomatology. Journal of Abnormal Psychology, 112, 92–99. Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of Social and Personal Relationships, 7, 147–178. Bartholomew, K. (1997). Adult attachment processes: Individual and couple perspectives. British Journal of Medical Psychology, 70, 249–263. 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. Bartholomew, K., & Shaver, P. R. (1998). Methods of assessing adult attachment. Do they converge?. In J. A. Simpson & W. S. Rholes (Eds.) Attachment theory and close relationships (pp. 22–45). London: Guilford Press. Bentall, R. P., Kinderman, P., & Kaney, S. (1994). The self, attributional processes and abnormal beliefs: Towards a model of persecutory delusions. Behaviour Research and Therapy, 32, 331–341. Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York: Basic Books. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). London: Guilford Press. Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, 644–663. Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual differences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 434–465). New York: Guilford Press. Dozier, M., & Lee, S. (1995). Discrepancies between self- and other-report of psychiatric symptomatology: Effects of dismissing attachment strategies. Development and Psychopathology, 7, 217–226. Dozier, M., Stovall, K. C., & Albus, K. E. (1999). Attachment and psychopathology in adulthood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 497–519). New York: Guilford Press. Eckblad, M. L., Chapman, L. J., Chapman, J. P., & Mishlowe, M. (1982). Unpublished test reported in Mishlove, M., & Chapman, L. J. (1985). Journal of Abnormal Psychology 94, 384–396. Feeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment. In M. B. Sperling & W. H. Berman (Eds.), Attachment in adults: Clinical and developmental perspectives (pp. 128–152). New York: Guilford Press.
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
717
Fenigstein, A., & Vanable, P. A. (1992). Paranoia and self-consciousness. Journal of Personality and Social Psychology, 62, 129–138. Field, A. (2000). Discovering statistics using SPSS for windows. London: Sage Publications. Freeman, D., Garety, P. A., Bebbington, P. E., Smith, B., Rollinson, R., Fowler, D., et al. (2005). Psychological investigation of the structure of paranoia in a non-clinical sample. British Journal of Psychiatry, 186, 427– 435. Garety, P. A., Kuipers, E., Fowler, D., Freeman, D., & Bebbington, P. E. (2001). A cognitive model of the positive symptoms of psychosis. Psychological Medicine, 31, 189–195. Goodwin, I. (2003). The relevance of attachment theory to the philosophy, organization and practice of adult mental health services. Clinical Psychology Review, 23, 35–56. Green, B. L. (1996). Trauma History Questionnaire. In B. H. Stamm (Ed.), Measurement of stress, self-report trauma and adaptation (pp. 366–368). Lutherville: Sidran Press. Griffin, D., & Bartholomew, K. (1994). The metaphysics of measurement: The case of adult attachment. In K. Bartholomew & D. Perlman (Eds.). Advances in personal relationships: Vol 5. Attachment processes in adulthood (pp. 17–52). London: Jessica Kingsley. Kelly, C., Sharkey, V., Morrison, G., Allardyce, J., & McCreadie, R. G. (2000). Nithsdale Schizophrenia Surveys 20. Cognitive function in a catchment area based population of patients with schizophrenia. British Journal of Psychiatry, 177, 348–353. Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working models, affect regulation and representations of self and others. Child Development, 59, 135–146. Launay, G., & Slade, P. (1981). The measurement of hallucinatory predisposition in male and female prisoners. Personality and Individual Differences, 2, 221–234. Main, M. & Goldwyn, R. (1984). Adult attachment scoring and classification system. University of California, unpublished manuscript. Mallinckrodt, B. (2000). Attachment, social competencies, social support, and interpersonal processes in psychotherapy. Psychotherapy Research, 10, 239–266. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172–175. Morrison, A. P., Wells, A., & Nothard, S. (2000). factors in predisposition to auditory and visual hallucinations. British Journal of Clinical Psychology, 39, 67–78. Mueser, K. T., Goodman, L. B., Trumbetta, S. L., Rosenberg, S. D., Osher, F. C., Vidaver, R., et al. (1998). Trauma and posttraumatic stress disorder in severe mental illness. Journal of Consulting and Clinical Psychology, 66, 493–499. Parker, G., Fairley, M., Greenwood, J., Jurd, S., & Silove, D. (1982). Parental representation of schizophrenics and their association with onset and course of schizophrenia. British Journal of Psychiatry, 141, 573–581. Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument. British Journal of Medical Psychology, 52, 1–10. Penn, D. L., Mueser, K. T., Tarrier, N., Gloege, A., Cather, C., Serrano, D., et al. (2004). Supportive therapy for schizophrenia. Possible mechanisms and implications for adjunctive psychosocial treatments. Schizophrenia Bulletin, 30, 101–112. Peters, E. R., Joseph, S. A., & Garety, P. A. (1999). Measurement of delusional ideation in the normal population: Introducing the PDI (Peters et al. Delusions Inventory). Schizophrenia Bulletin, 25, 553–576. Platts, H., Tyson, M., & Mason, O. (2002). Adult attachment style and core beliefs: Are they linked?. Clinical Psychology and Psychotherapy 9, 332–348. Randolph, E. T. (1998). Social networks and schizophrenia. In K. T. Mueser & N. Tarrier (Eds.), Handbook of social functioning in schizophrenia (pp. 238–246). Needham Heights: Allyn Bacon. Robson, P. (1989). Development of a new self-report questionnaire to measure self esteem. Psychological Medicine, 19, 513–518. Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New York: Harper Collins. van Os, J., Hanssen, M., Bijl, R. V., & Ravelli, A. (2000). Straus (1969) revisited: A psychosis continuum in the general population? Schizophrenia Research, 45, 11–20.
718
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063–1070. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, and distress: Exploring the central role of negative affectivity. Psychology Review, 96, 234–254. Wilson, J. S., & Costanzo, P. R. (1996). A preliminary study of attachment, attention, and schizotypy in early adulthood. Journal of Social and Clinical Psychology, 15, 231–260.
Attachment styles, interpersonal relationships and psychotic phenomena in a non-clinical student sample Katherine Berry *, Alison Wearden, Christine Barrowclough, Tom Liversidge School of Psychological Sciences, University of Manchester, Rutherford House, Manchester Science Park, Lloyd Street North, Manchester M15 6SZ, United Kingdom Received 28 October 2005; received in revised form 4 March 2006; accepted 16 March 2006 Available online 6 May 2006
Abstract This paper describes the development and initial validation of a measure of adult attachment styles based on Bartholomew’s (1990, 1997) model and adapted for use with individuals with psychosis. The associations between the new measure, interpersonal problems, self-concept, relationships with significant others during childhood, traumatic life events and non-clinical psychotic phenomena were examined in a student sample. Three hundred and twenty three students completed study measures via an internet website. In support of the construct validity of the measure, insecure attachment styles were associated with interpersonal problems and anxiety in attachment relationships was associated with low self-esteem. There were weak but significant associations between insecure attachment and more negative experiences of being parented. There was no evidence of associations between attachment and traumatic life events. As predicted, there were significant associations between insecure attachment and non-clinical psychotic phenomena; positive psychotic phenomena were associated with anxiety in attachment relationships, and social anhedonia with avoidance in attachment relationships. Findings suggest that the attachment measure is a valid instrument with good internal consistency, and can be used to explore associations between attachment styles and psychotic symptoms in clinical samples. Ó 2006 Elsevier Ltd. All rights reserved.
*
Corresponding author. Tel.: +44 0 161 275 8498; fax: +44 0 161 275 8487. E-mail address: [email protected] (K. Berry).
0191-8869/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2006.03.009
708
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Keywords: Adult attachment; Analogue sample; Interpersonal problems; Psychosis
1. Introduction Recent models of psychosis implicate earlier experiences with significant others and interpersonal traumas in the development and maintenance of psychotic symptoms (Garety, Kuipers, Fowler, Freeman, & Bebbington, 2001). Since the seminal work of Bowlby (1982) on early childhood attachments, much research has suggested that earlier interpersonal experiences influence attachments to others in adulthood, and adult attachment styles predict interpersonal functioning, self-esteem and psychopathology (Bartholomew & Horowitz, 1991; Platts, Tyson, & Mason, 2002). Psychosis is associated with interpersonal difficulties (Penn et al., 2004) and low self-esteem (Barrowclough et al., 2003). Attachment theory may therefore provide a useful framework for conceptualizing the influence of interpersonal relationships on the development and course of psychosis (Dozier, Stovall, & Albus, 1999). There is, however, currently limited empirical research investigating attachment styles and their correlates in clinical samples (Goodwin, 2003) and there are difficulties assessing the attachment styles of patients with psychosis with available measures. Interview attachment measures, such as the Adult Attachment Interview (AAI; Main & Goldwyn, 1984) assess attachment states of mind on the basis of the individual’s coherence of narrative in describing parenting experiences. Such measures are time consuming to complete, require extensive training, and, when administered to individuals with psychosis, may be confounded by the presence of psychotic symptoms (Dozier et al., 1999). Self-report attachment measures, although easier to administer, tend to focus on close interpersonal relationships with some items referring specifically to romantic relationships. They may therefore be less relevant to individuals with psychosis, who are often socially isolated (Randolph, 1998). Cognitive problems associated with psychosis may also make it difficult for patients to understand negatively worded items and rate themselves on scales with relatively wide-ranging Likert scales and few anchor points (Kelly, Sharkey, Morrison, Allardyce, & McCreadie, 2000). There is therefore a need to develop and validate a measure of attachment that is more appropriate for use with this group. Two dimensions have been found to underlie self-report measures (Brennan, Clark, & Shaver, 1998), which can be conceptualized in cognitive terms as: ‘model of self’ and ‘model of other’; or affective and behavioural terms as: ‘anxiety’ and ‘avoidance’ (Crowell, Fraley, & Shaver, 1999). Attachment anxiety is associated with a negative self-image and an excessive need for approval from others, coupled with a fear of rejection and abandonment. Attachment avoidance is associated with a negative image of others and is defined in terms of either an excessive need for selfreliance or a fear of depending on others. Bartholomew (1990, 1997) has also described these dimensions in terms of four prototypes: secure; fearful; dismissing; and preoccupied (see Fig. 1). Preoccupied and fearful attachment styles, which are characterized by high anxiety in attachment relationships, have been associated with low self-esteem (Bartholomew & Horowitz, 1991; Collins & Read, 1990). Dismissing attachment, characterized by low anxiety but high avoidance, has been associated with hostility, interpersonal coldness and emotional detachment. Preoccupied attachment, characterized by high anxiety and
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
709
MODEL OF SELF (ANXIETY) Positive (Low)
SECURE Positive (Low)
High self-worth, believes that others are responsive, comfortable with autonomy and in forming close relationships with others.
Negative (High)
PREOCCUPIED A sense of self -worth that is dependent on gaining the approval and acceptance of others.
MODEL OF OTHER (AVOIDANCE) DISMISSING
Negative (High)
Overt positive self-view, denies feelings of subjective distress and dismisses the importance of close relationships.
FEARFUL Negative self-view, lack of trust in others, subsequent apprehension about close relationships and high levels of distress.
Fig. 1. Bartholomew’s (1990) model of attachment (adapted from Bartholomew and Horowitz, 1991, p. 227).
low avoidance, has been associated with over intrusive behaviour, dominance and aggression. Fearful attachment, characterized by both high anxiety and high avoidance, has been associated with difficulties in assertiveness, social inhibition, anger and aggression (Bartholomew & Horowitz, 1991; Kobak & Sceery, 1988; Mallinckrodt, 2000). High levels of anxiety or avoidance in attachment relationships may be involved in the maintenance of specific psychotic symptoms. One study involving a non-clinical sample found associations between self-reported anxiety in attachment relationships and positive schizotypal characteristics such as bizarre beliefs and anomalous experiences, and avoidance and negative schizotypal characteristics such as withdrawal and loss of pleasure (Wilson & Costanzo, 1996). The authors found an additional relationship between attachment avoidance and positive schizotypy, and theorized, following Feeney, Noller, and Hanrahan (1994), that avoidance in attachment relationships is partially anxiety-driven. Both paranoia and avoidant attachment, are thought to be associated with negative interpersonal beliefs developed as a result of earlier interpersonal experiences (Bartholomew, 1990; Bentall, Kinderman, & Kaney, 1994). However, Wilson and Costanzo (1996) did not specifically investigate the relationship between avoidant attachment and the schizotypal characteristic of paranoid thinking. They also failed to control for the influence of negative affect which has been shown to predict the extent to which individuals report psychological symptoms (Watson & Pennebaker, 1989). The relationship between attachment and specific schizotypal symptoms therefore warrants further investigation. The use of analogue or non-clinical samples in psychosis research is becoming increasingly popular due to the recognition that psychotic symptoms are on
710
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
a continuum with normal experiences. Psychotic phenomena in non-clinical populations are therefore an interesting area of research in their own right and can inform the understanding of psychosis (Peters, Joseph, & Garety, 1999; van Os, Hanssen, Bijl, & Ravelli, 2000). The first aim of the present study is to assess the psychometric properties of a new attachment style questionnaire, the ‘Psychosis Attachment Measure’ (PAM), in an analogue sample. Construct validity will be demonstrated by associations between high levels of attachment anxiety and avoidance and interpersonal difficulties, an association between high levels of attachment anxiety and low self-esteem, and associations between high levels of attachment anxiety and avoidance and negative interpersonal experiences, including poor relationships with significant others during childhood and interpersonal traumas. The second aim of the study is to investigate the associations between attachment and non-clinical psychotic phenomena in an analogue sample. It is hypothesized that anxiety and avoidance in attachment relationships will be associated with higher levels of non-clinical psychotic phenomena. Specifically, anxiety will be associated with positive psychotic phenomena, and avoidance with negative symptom traits and paranoia. Following the validation of the measure in a large analogue sample, it can then be validated in more extensive samples with psychosis and used to develop theoretical models of interpersonal difficulties and symptoms.
2. Methods 2.1. Participants and procedure Participants were 323 undergraduate and postgraduate students at University of Manchester who responded to an e-mail inviting them to participate in a web-based study investigating ‘attachment styles, interpersonal relationships and personality’. The median age of the sample was 21 years (range = 17–67 years) and 72% were female. The majority of the sample was White British (82%). Approximately half were in a romantic relationship (46%). 2.2. Measures 2.2.1. Attachment measures The 16-item PAM was based on existing measures of attachment (Bartholomew & Horowitz, 1991; Brennan et al., 1998). Items referred to thoughts, feelings and behaviours in close interpersonal relationships, but did not refer specifically to romantic relationships. Participants were asked to rate the extent to which each item was characteristic of them using a four-point scale ranging from ‘not at all’ to ‘very much’. Preliminary data from a clinical sample of patients with a schizophrenia diagnosis (n = 27) revealed that the measure is acceptable to participants and suggested good reliability, with alphas of .96 for the anxiety subscale and .85 for the avoidance subscale. The Relationships Questionnaire (RQ; Bartholomew & Horowitz, 1991) was used to assess the concurrent validity of the PAM. It consists of four paragraphs describing secure, preoccupied, fearful and dismissing attachment styles. Respondents are asked to rate the degree to which each description characterizes them using a seven-point scale, ranging from ‘not at all like me’, to ‘very
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
711
much like me’. Scores on these four items were used to derive the two dimensions of ‘model of others’ and ‘model of self’ (Griffin & Bartholomew, 1994). The RQ has been shown to have reasonable reliability and validity (Griffin & Bartholomew, 1994). 2.2.2. Interpersonal problems The 32-item Inventory of Interpersonal Problems-32 (IIP-32; Barkham, Hardy, & Startup, 1996) assesses a range of interpersonal problems, including aggression, poor sociability and excessive dependence on others. Respondents are asked to rate how much each problem distresses them on a five-point scale, ranging from ‘not at all’ to ‘extremely’. It has good psychometric properties (Barkham et al., 1996), and the alpha in the present study was .85. 2.2.3. Self-esteem The 30-item Self-Concept Questionnaire (SCQ; Robson, 1989) assesses various aspects of selfconcept. Respondents are asked to rate items on an eight-point scale, ranging from ‘completely disagree’ to ‘completely agree’. A high total scale score indicates low self-esteem. The measure has good reliability and validity (Robson, 1989) and the alpha in the present study was .91. 2.2.4. Earlier interpersonal experiences The Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979) is a 25-item self-report measure which assesses individuals’ retrospective judgements of care and protection in their relationship with a significant caregiver during the first 16 years of life. In the present study, participants were asked to answer the questions in relation to their primary caregiver during childhood. Low scores on the care scale reflect parental neglect and rejection and high scores on the protection scale indicate excessive control and intrusive parenting. The PBI is a well-established measure (Parker et al., 1979). Alphas in the present study were .92 for the care subscale and .88 for the protection subscale. Experiences of traumatic events were assessed using a measure based on the Trauma History Questionnaire (THQ; Green, 1996), which asked participants about their experiences of a range of interpersonal traumas. 2.2.5. Positive schizotypy The 20-item paranoia scale (PS; Fenigstein & Vanable, 1992) assesses levels of paranoid thinking. The measure has good psychometric properties in non-clinical samples (Fenigstein & Vanable, 1992) and the alpha in the present study was .90. The revised version of the Launay–Slade Hallucinations Scale (LSHS; Launay & Slade, 1981) is a 24-item scale which assesses the frequency of clinical and sub-clinical hallucinatory phenomena (Morrison, Wells, & Nothard, 2000). The measure has good psychometric properties (Morrison et al., 2000) and the alpha in the present study was .90. 2.2.6. Negative schizotypy The 40-item revised Social Anhedonia Scale (SAS; Eckblad, Chapman, Chapman, & Mishlowe, 1982) assesses inability to experience positive feelings in interpersonal relationships. Total scores are derived by summing positively endorsed traits. It has good psychometrics in student samples (Eckblad et al., 1982) and the alpha in the present study was .85.
712
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
2.2.7. Affect The Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) was used to control for the influence of negative affect on other self-report measures. Respondents are asked to rate the extent to which they have experienced ten positive and ten negative affective states during the past week. The measure has good reliability and validity (Watson et al., 1988) and in the present study, the alpha for the negative affect scale was .87. 2.2.8. Data analysis Attempts were made to transform skewed data sets (Tabachnick & Fidell, 1996). As some transformations were unsuccessful, for consistency and ease of interpretation all data sets were left untransformed. Data were then analysed using both parametric and non-parametric correlations, and as the significance levels were the same in both sets of analyses only parametric findings are reported. Where there were missing data, all available data were analyzed, and the sample size for each analysis reported in the tables. Correlation coefficients were compared using procedures suggested by Meng, Rosenthal, and Rubin (1992). Due to the large number of analyses performed, a conservative p value of .01 was adopted. All analyses were conducted using SPSS version 11.
3. Results 3.1. Factor analysis The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was .85, suggesting that the pattern of correlations were relatively compact and that a factor analysis should produce distinct and reliable factors (Field, 2000). An initial exploratory factor analysis of items on the PAM extracted three factors with eigenvalues above one. A varimax orthogonal rotation was used to help interpret the findings (Field, 2000). As expected, one of the factors seemed to represent attachment anxiety and the other avoidance. One item loaded onto the third factor: ‘If I have a problem with other people, I usually let them know’. As this item did not load onto the first or second factors, a decision was made to exclude it from further analyses and run a two-factor solution. Following varimax rotation, the first factor, representing anxiety, accounted for 23.87% of the variance and the second factor, representing avoidance, accounted for 20.53% of the variance (Table 1). The use of orthogonal rotation was deemed appropriate as the factors were only weakly correlated at .22 (Tabachnick & Fidell, 1996). Inspection of the residuals suggested that the solution provided an acceptable fit to the data (Field, 2000). The internal consistency of each dimension was acceptable. Cronbach’s alphas for the anxiety and avoidance dimensions were .82 and .75, respectively. Anxiety and avoidance subscale scores were derived by averaging scores for the 9 anxiety items and 6 avoidance items (Table 1). 3.2. Concurrent validity There were significant negative correlations between RQ positivity of self-model and attachment anxiety (r = .56, p < .001) and between RQ positivity of other model and attachment
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
713
Table 1 Factor loadings for individual items Factor 1: Anxiety 3. I tend to get upset, anxious or angry if other people are not there when I need them. 5. I worry that key people in my life won’t be around in the future.
Factor 2: Avoidance .67 .60
6. I frequently ask other people to reassure me that they care about me. 7. If other people disapprove of something I do, I get very upset. 9. I frequently wonder whether I can trust other people.
.70
10. I find it hard to believe that other people will be there for me if I need them. 11. I worry that if other people get to know me better, they won’t like who I really am. 15. I believe that other people will never be concerned about me. 16. I worry that if I displease other people, they won’t want to know me anymore.
.59
.62 .54
1. I prefer not to let other people know my ‘true’ thoughts and feelings. 2. I find it easy to depend on other people for support with problems or difficult situations. (reverse item) 4. I usually discuss my problems and concerns with other people. (reverse item) 8. I find it hard to accept help from other people when I have problems or difficulties. 12. When I’m feeling stressed, I prefer being on my own to being in the company of other people. 14. I try to cope with stressful situations on my own.
.62 .66
.69 .63 .56
.71
.71 .56 .75
avoidance (r = .45, p < .001). These correlations were similar to those obtained in previous research investigating relationships between self-report measures of attachment (Feeney et al., 1994; Bartholomew & Shaver, 1998). The relationships between anxiety and self-model (r = .49, p < .001) and between avoidance and other model remained significant when controlling for the influence of negative affect (r = .46, p < .001). 3.3. Attachment, interpersonal problems, self-esteem and earlier interpersonal experiences The relationships between attachment anxiety and avoidance, interpersonal problems, self-esteem and earlier interpersonal experiences were investigated as a further test of the measure’s validity (see Table 2). Both attachment anxiety and avoidance were significantly positively correlated with the total level of interpersonal problems. However, the correlation between anxiety and interpersonal difficulties was significantly higher than the correlation between avoidance and interpersonal difficulties (z = .40, p < .01). There was a significant negative correlation between attachment anxiety and self-esteem, suggesting that participants who rated themselves as high in attachment anxiety had lower self-esteem. Participants who rated themselves as high in attachment avoidance also had lower self-esteem but, as would be predicted on the basis of Bartholomew’s (1990) model, the correlation between anxiety and self-esteem was significantly higher than the correlation between avoidance and self-esteem (z = .51, p < .01). The above relationships were all still significant when controlling for the influence of negative affect.
714
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Table 2 Correlations between attachment, psychosocial functioning, interpersonal relationships and non-clinical psychotic phenomena Scale
n
Inventory of interpersonal problems Robson self-concept questionnaire Parental bonding instrument care subscalea Parental bonding instrument protection subscale Total number of traumatic events reported Total trauma distress scores Paranoia scale Launay hallucinations scale Social anhedonia scalea
280 248 266 266 258 258 244 222 230
a
Attachment anxietya
Attachment avoidance
Anxiety partial correlations for negative affect
Avoidance partial correlations for negative affect
r
r
r
r
p .60 .60 .23 .24 .02 .07 .51 .37 .19
<.001 <.001 <.001 <.001 .73 .258 <.001 <.001 .004
p .28 .18 .20 .18 .10 .12 .18 .20 .44
<.001 .004 .001 .003 .137 .072 .004 .003 <.001
p .52 .48 .15 .15 .04 .01 .41 .27 .12
<.001 <.001 .035 .037 .574 .894 <.001 <.001 .09
p .29 .20 .13 .12 .08 .08 .19 .16 .42
<.001 .006 .075 .091 .28 .263 .006 .025 <.001
Skewed variables.
There were significant associations between attachment and memories of relationships with caregivers. Participants who were higher in both attachment anxiety and avoidance tended to rate parents as less caring and more controlling. Although the findings were statistically significant due to the large sample size, the strength of the correlations was relatively weak, and they became nonsignificant when negative affect was controlled. There were no significant relationships between attachment and the trauma measure. 3.4. Attachment and non-clinical psychotic phenomena There were positive associations between all non-clinical psychotic phenomena and insecure attachment (see Table 2). The correlation between anxiety and paranoia was significantly greater than the correlation between avoidance and paranoia (z = .38, p < .01). The correlation between avoidance and social anhedonia was significantly greater than the correlation between anxiety and social anhedonia (z = .28, p < .01). The correlation between anxiety and hallucinations was greater than that between avoidance and hallucinations. When negative affect was controlled for, the significant relationships between attachment anxiety and paranoia, and anxiety and hallucinatory experiences were maintained; the significant relationships between attachment avoidance and social anhedonia, and avoidance and paranoia were similarly maintained, but the relationship between avoidance and hallucinations no longer reached significance.
4. Discussion The study suggests that the PAM has good psychometric properties. As predicted, there were significant associations between insecure attachment and non-clinical psychotic phenomena, with
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
715
specific associations between positive psychotic phenomena and anxiety in attachment relationships and social anhedonia and avoidance in attachment relationships. The methodological limitations of the study need to be taken into consideration in interpreting the findings and will be outlined prior to discussing the results in further detail. First, the cross-sectional design of the study limits the conclusions that can be drawn concerning the causal directions of the relationships between associated variables and relationships are most likely bi-directional. Second, the sample consisted of students who were self-selected, recruited by e-mail, and contained a preponderance of female participants. It is possible that people who self-select for such studies may be more prone to psychological problems (Freeman et al., 2005). Third, the use of a questionnaire to measure attachment introduces the possibility of self-reporting biases, including social desirability biases. Arguably, the anonymity of internet research may reduce the influence of social desirability (Freeman et al., 2005), but it is not possible to rule it out completely. Finally, as the study relied solely on self-report measures the findings may be confounded by common method variance. Despite these limitations, the fact that the findings are consistent with existing literature and theoretical models, supports their validity. Consistent with current models of attachment, the factor analysis suggested that two dimensions underlie the PAM (Bartholomew, 1990, 1997). The measure appears to be a reliable assessment tool with internally consistent subscales and good concurrent validity. As expected, there were relationships between insecure attachment and interpersonal problems (Bartholomew & Horowitz, 1991; Mallinckrodt, 2000), and we were able to show that these relationships were not entirely due to negative affect. It would, however, be interesting to replicate the study using an informant measure of interpersonal problems to investigate the influence of reporting biases on interpersonal problems. Arguably, significant others may be more likely to report negative interpersonal traits than individuals themselves (Kobak & Sceery, 1988). There is also evidence of attachment-style differences in the reporting of interpersonal problems. The relatively low correlation between avoidance and interpersonal problems compared to anxiety and interpersonal problems may be due to individuals with dismissing attachment under-reporting distress (Dozier & Lee, 1995). In line with previous research and in support of the validity of the measure, there were significant relationships between attachment anxiety and self-esteem (Bartholomew & Horowitz, 1991; Collins & Read, 1990). However, there were no strong or consistent relationships between attachment and parental relationships, nor between attachment and experiences of trauma. As participants tended to report positive parental relationships and relatively few traumatic events, it may have been difficult to identify strong relationships between negative experiences and insecure attachment. The relationships between these constructs therefore warrants investigation in samples of clinical participants who are more likely to report difficulties in parental relationships (Parker, Fairley, Greenwood, Jurd, & Silove, 1982) and traumatic events (Mueser et al., 1998). There were positive associations between anxiety and avoidance in attachment relationships and schizotypal characteristics (Wilson & Costanzo, 1996). As predicted, there were significant relationships between anxiety and the positive schizotypal characteristics of paranoia and hallucinations, and between avoidance and the negative schizotypal characteristic of social anhedonia. Correlations between anxiety and positive schizotypy were significantly stronger than those between avoidance and positive schizotypy; correlations for avoidance and social anhedonia were stronger than those between anxiety and social anhedonia. As predicted on the basis of cognitive
716
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
models of paranoia, there was also a weak association between avoidance and paranoid thinking, which remained significant when controlling for the influence of negative affect. In summary, the findings of this study suggest that the PAM is a reliable and valid instrument, which can be used to assess attachment styles in analogue samples and develop theoretical models of psychosis. Investigations of the associations between attachment and psychosis in clinical samples are now warranted to assess whether the relationships between attachment and specific symptoms are replicated. Acknowledgements The authors thank all the students who participated in the research. References Barkham, M., Hardy, G. E., & Startup, M. (1996). The IIP-32: A short version of the Inventory of Interpersonal Problems. British Journal of Clinical Psychology, 35, 21–35. Barrowclough, C., Tarrier, N., Andrews, B., Humphreys, L., Ward, J., & Gregg, L. (2003). Self-esteem in schizophrenia: Relationships between self-evaluation, family attitudes, and symptomatology. Journal of Abnormal Psychology, 112, 92–99. Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of Social and Personal Relationships, 7, 147–178. Bartholomew, K. (1997). Adult attachment processes: Individual and couple perspectives. British Journal of Medical Psychology, 70, 249–263. 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. Bartholomew, K., & Shaver, P. R. (1998). Methods of assessing adult attachment. Do they converge?. In J. A. Simpson & W. S. Rholes (Eds.) Attachment theory and close relationships (pp. 22–45). London: Guilford Press. Bentall, R. P., Kinderman, P., & Kaney, S. (1994). The self, attributional processes and abnormal beliefs: Towards a model of persecutory delusions. Behaviour Research and Therapy, 32, 331–341. Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York: Basic Books. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). London: Guilford Press. Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, 644–663. Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual differences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 434–465). New York: Guilford Press. Dozier, M., & Lee, S. (1995). Discrepancies between self- and other-report of psychiatric symptomatology: Effects of dismissing attachment strategies. Development and Psychopathology, 7, 217–226. Dozier, M., Stovall, K. C., & Albus, K. E. (1999). Attachment and psychopathology in adulthood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 497–519). New York: Guilford Press. Eckblad, M. L., Chapman, L. J., Chapman, J. P., & Mishlowe, M. (1982). Unpublished test reported in Mishlove, M., & Chapman, L. J. (1985). Journal of Abnormal Psychology 94, 384–396. Feeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment. In M. B. Sperling & W. H. Berman (Eds.), Attachment in adults: Clinical and developmental perspectives (pp. 128–152). New York: Guilford Press.
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
717
Fenigstein, A., & Vanable, P. A. (1992). Paranoia and self-consciousness. Journal of Personality and Social Psychology, 62, 129–138. Field, A. (2000). Discovering statistics using SPSS for windows. London: Sage Publications. Freeman, D., Garety, P. A., Bebbington, P. E., Smith, B., Rollinson, R., Fowler, D., et al. (2005). Psychological investigation of the structure of paranoia in a non-clinical sample. British Journal of Psychiatry, 186, 427– 435. Garety, P. A., Kuipers, E., Fowler, D., Freeman, D., & Bebbington, P. E. (2001). A cognitive model of the positive symptoms of psychosis. Psychological Medicine, 31, 189–195. Goodwin, I. (2003). The relevance of attachment theory to the philosophy, organization and practice of adult mental health services. Clinical Psychology Review, 23, 35–56. Green, B. L. (1996). Trauma History Questionnaire. In B. H. Stamm (Ed.), Measurement of stress, self-report trauma and adaptation (pp. 366–368). Lutherville: Sidran Press. Griffin, D., & Bartholomew, K. (1994). The metaphysics of measurement: The case of adult attachment. In K. Bartholomew & D. Perlman (Eds.). Advances in personal relationships: Vol 5. Attachment processes in adulthood (pp. 17–52). London: Jessica Kingsley. Kelly, C., Sharkey, V., Morrison, G., Allardyce, J., & McCreadie, R. G. (2000). Nithsdale Schizophrenia Surveys 20. Cognitive function in a catchment area based population of patients with schizophrenia. British Journal of Psychiatry, 177, 348–353. Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working models, affect regulation and representations of self and others. Child Development, 59, 135–146. Launay, G., & Slade, P. (1981). The measurement of hallucinatory predisposition in male and female prisoners. Personality and Individual Differences, 2, 221–234. Main, M. & Goldwyn, R. (1984). Adult attachment scoring and classification system. University of California, unpublished manuscript. Mallinckrodt, B. (2000). Attachment, social competencies, social support, and interpersonal processes in psychotherapy. Psychotherapy Research, 10, 239–266. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172–175. Morrison, A. P., Wells, A., & Nothard, S. (2000). factors in predisposition to auditory and visual hallucinations. British Journal of Clinical Psychology, 39, 67–78. Mueser, K. T., Goodman, L. B., Trumbetta, S. L., Rosenberg, S. D., Osher, F. C., Vidaver, R., et al. (1998). Trauma and posttraumatic stress disorder in severe mental illness. Journal of Consulting and Clinical Psychology, 66, 493–499. Parker, G., Fairley, M., Greenwood, J., Jurd, S., & Silove, D. (1982). Parental representation of schizophrenics and their association with onset and course of schizophrenia. British Journal of Psychiatry, 141, 573–581. Parker, G., Tupling, H., & Brown, L. B. (1979). A parental bonding instrument. British Journal of Medical Psychology, 52, 1–10. Penn, D. L., Mueser, K. T., Tarrier, N., Gloege, A., Cather, C., Serrano, D., et al. (2004). Supportive therapy for schizophrenia. Possible mechanisms and implications for adjunctive psychosocial treatments. Schizophrenia Bulletin, 30, 101–112. Peters, E. R., Joseph, S. A., & Garety, P. A. (1999). Measurement of delusional ideation in the normal population: Introducing the PDI (Peters et al. Delusions Inventory). Schizophrenia Bulletin, 25, 553–576. Platts, H., Tyson, M., & Mason, O. (2002). Adult attachment style and core beliefs: Are they linked?. Clinical Psychology and Psychotherapy 9, 332–348. Randolph, E. T. (1998). Social networks and schizophrenia. In K. T. Mueser & N. Tarrier (Eds.), Handbook of social functioning in schizophrenia (pp. 238–246). Needham Heights: Allyn Bacon. Robson, P. (1989). Development of a new self-report questionnaire to measure self esteem. Psychological Medicine, 19, 513–518. Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New York: Harper Collins. van Os, J., Hanssen, M., Bijl, R. V., & Ravelli, A. (2000). Straus (1969) revisited: A psychosis continuum in the general population? Schizophrenia Research, 45, 11–20.
718
K. Berry et al. / Personality and Individual Differences 41 (2006) 707–718
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063–1070. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, and distress: Exploring the central role of negative affectivity. Psychology Review, 96, 234–254. Wilson, J. S., & Costanzo, P. R. (1996). A preliminary study of attachment, attention, and schizotypy in early adulthood. Journal of Social and Clinical Psychology, 15, 231–260.