Mind-mindedness in parents of pre-schoolers: A comparison between clinical and community samples

Abstract

Mind-mindedness relates to parents’ propensity to treat their young children as individuals with minds of their own. Research with community samples has demonstrated impressive findings regarding child development outcomes, leading to a suggestion that mind-mindedness should be considered in clinical interventions. This is the first mind-mindedness study to include parents of children referred to clinical services. A between group design (n=49) was used to investigate whether mind-mindedness differed between parents of a clinical group of pre-school children and parents of a community comparison group and to explore how mind-mindedness related to parental depression and stress, and child difficulties. The findings revealed that mind-mindedness was significantly lower in the clinical sample and was not related to depression in either group. In the clinical group mind-mindedness was related to parenting stress and in the community group it was related to children’s emotional and behavioural difficulties. Overall these findings provide preliminary evidence that mind-mindedness may be an important construct to consider in pre-school clinical interventions.

Keywords

Attachment clinical sample community sample mind-mindedness parent pre-school

Introduction

The parent-child attachment relationship has gained an important status in Western understanding of child development, because it has been found to play a primary role in children’s social, emotional and cognitive development (Thompson, 2008). Reflecting this, many key child policies and strategies now highlight the importance of stable attachments early in life (Department for Education and Skills, 2002, 2007; Department of Health, 2004, 2008, 2010; World Health Organization, 2004) and a range of clinical interventions have been developed to support the successful development of this attachment relationship (Cassidy & Shaver, 2008; British Psychological Society, 2007). A robust finding which further highlights the significance of attachment theory is that security of attachment is frequently transferred between generations (van Ijzendoorn, 1995).

Despite a vast amount of attachment-related literature, there is still much uncertainty regarding the antecedents of attachment and the mediators of this inter-generational transfer. It has been assumed since Ainsworth, Bell and Stayton’s (1974) early work that the primary mechanism in attachment is parental sensitivity in responding to children. However, influential meta-analyses have shown that sensitivity accounts for less than 25% of the influence of parental attachment on infant attachment security, leaving a ‘transmission gap’ (De Wolff & van Ijzendoorn, 1997; Raval et al., 2001). This problem has led researchers to look beyond sensitivity, as it is currently assessed, when searching for parental antecedents to security of attachment.

Parental capacity to represent the mental state of their children has been conceptualized by some researchers as a central antecedent to secure attachment and two main constructs, ‘reflective functioning’ and ‘mind-mindedness’, have emerged to reflect this capacity. Using a psychoanalytic framework Fonagy et al. (1991) argued that sensitive and responsive parenting was based on a capacity to make sense of one’s own attachment experiences in order to reflect on the infant’s mental world, and termed this concept reflective functioning (RF). More recently, from a cognitive developmental perspective, Meins (1997) developed the construct of mind-mindedness. She defined mind-mindedness (MM) as “the proclivity to treat one’s child as an individual with a mind from an early age” (Meins & Fernyhough, 1999:364). Therefore mind-mindedness relates to parents’ capacity to be sensitive to what is in their children’s minds, rather than just their physical or behavioural needs.

Both originators of these constructs acknowledge similarities and differences, but their theoretical background and operationalization are particularly different (Arnott & Meins, 2007; Sharp & Fonagy, 2008). Mind-mindedness focuses directly on how the parent perceives the child, rather than on how parents perceive their own attachment experiences. In terms of measurement differences RF is coded from a lengthy parent interview. MM uses a quicker and simpler representational measure than RF of parents’ tendency to conceive of their child as a mental agent. There are two mind-mindedness measures aimed at different age groups. The infant MM measure is for parents of infants in the first year of life and is coded from a videotaped parent-infant play session. The preschool MM measure is for parents of preschool children (three-five years) and is coded from a single question interview which simply asks parents to describe their child. The two MM measures are similar; they both assess the parents’ proclivity to use mind-related comments in relation to their child, the main difference being that the live measure includes a system for categorizing the appropriateness of mental comments (Meins & Fernyhough, 2006), by cross referencing the visual cues from the interaction with the parent’s comments. These simple MM measures however show good emerging validity and reliability and have yielded some striking findings regarding the importance of parental mind-mindedness for child development.

The theoretical underpinnings of mind-mindedness relate to the social development of cognition (Meins, 1997) and a rethinking of what constitutes parental sensitivity. Meins, Fernyhough, Fradley and Tuckey (2001) point out that Ainsworth, Bell and Stayton’s (1971) original definition of sensitivity included the mother’s ability to look at things from the child’s point of view. They note that most research into sensitivity has focused on behavioural responsiveness and neglected to consider this cognitive sensitivity. Drawing on a Vygotskian framework, Meins (1997, 1999) proposes not only that mind-mindedness is an antecedent to secure attachment but also that to treat the child as having a mind of their own stimulates the child’s ability to understand the minds of others. In this way mind-mindedness theoretically has the potential to explain how attachment is transferred from one generation to the next. Parental sensitivity to the child’s mind is therefore suggested to both underpin attachment and the development of this capacity to be mind-sensitive in the child, which the child then passes on to their own children.

The theory of MM is now becoming well supported by research evidence. Mind-mindedness has been shown to relate to, but be distinct from, traditionally measured maternal sensitivity in responding to children (Laranjo, Bernier & Meins, 2008; Lok & McMahon, 2006; Meins et al., 2002). Studies have demonstrated that parental MM of pre-schoolers correlates with child attachment security (Meins 1998; Meins, Fernyhough, Russell & Clark-Carter, 1998) while parental MM of six-month-olds predicts their child’s attachment security (Arnott & Meins, 2007; Lundy, 2003; Meins et al., 2001; Meins et al., 2011; Meins, Fernyhough, Wainwright, Gupta, Fradley & Tuckey, 2002). Furthermore, MM at either age predicts later Theory of Mind (ToM) development (Laranjo, Bernier, Meins & Carlson, 2010; Meins et al., 1998; Meins et al., 2003; Meins & Fernyhough, 1999), with most research showing medium to large effect sizes. Interestingly, the infant measure’s coding of appropriate and inappropriate mental comments has revealed that these two indices are not related (Meins et al., 2001; Meins et al., 2003; Meins et al., 2011), and although both indices predict attachment, only appropriate mental comments predict child ToM skills. Impressively, two studies looking at the relationship between MM, parents’ own attachment style and children’s attachment also indicate that MM may mediate parent and child attachment, thus helping to bridge the transmission gap (Arnott & Meins, 2007; Bernier & Dozier, 2003).

As mind-mindedness has been shown to be a possible antecedent to secure attachment and also helps children develop the ability to understand their own and others’ minds, theoretically, parental mind-mindedness could be a prime target for interventions. Studies suggesting mind-mindedness might help bridge the transmission gap recommended that future research should consider MM as a focus for clinical interventions (Arnott & Meins, 2007; Bernier & Dozier, 2003). In addition, some theoretical literature regarding attachment interventions is beginning to include the potential role of mind-mindedness (Atkinson & Goldberg, 2004). However, in reviewing the current MM research it is striking that none of the published research to date has used parents of a clinical sample of children.

In order to understand if MM would be an appropriate target for interventions there is a need for research which includes a clinical sample: i.e. parents of children with significant emotional and behavioural difficulties. There is now good evidence that lower mind-mindedness is associated with insecure attachment (Meins et al., 1998; Meins et al., 2001; Meins et al., 2011) and many renowned interventions for under-fives assume that the emotional and behavioural difficulties young children are referred with are at least in part an expression of attachment difficulties (e.g. Solihull NHS Care Trust, 2005; The Tavistock & Portman Clinic, 2009; Lojkasek & Cohen, 1999). Therefore, it could be predicted that parents accessing services would be lower in MM. In order to test this hypothesis the MM of parents of children referred to early years services would need to be compared to the MM of a similar sample of parents of non-referred or community children.

The apparent need for research which considers whether mind-mindedness is lower in parents of a clinical sample of children who have been referred to early years services was the basis for the current study. If it is the case that MM is lower in parents of a clinical sample of children then this increases the likelihood that MM would be a useful focus for interventions. The fact that the MM measures are straightforward and quick to administer would also increase the clinical relevance of any such findings.

There are some other important factors which need to be considered when comparing MM in parents of a clinical sample of children with parents of a community sample of children. As most young children are referred to clinical services with emotional or behavioural difficulties (Berlin, Zeanah & Lieberman, 2008), it was considered important for the research to also explore how MM is related to these difficulties. Furthermore, lower mind-mindedness has been found to be associated with depressive symptoms in parents in some community samples (Lok & McMahon, 2006; Lundy, 2003; Meins et al., 2011), and parents accessing child services may experience higher levels of depression (Beck, 2001), an exploration of the relationship between depression and MM in clinical samples was warranted. Similarly, parents accessing clinical services are likely to be experiencing high levels of stress (Broadhead, Chilton & Crichton, 2009), which presented a need to begin to also consider the relationship between MM and parenting stress.

The current study

In order to begin to address a gap in the literature, the present study aimed to explore parental mind-mindedness in a clinical sample of children, by comparing mind-mindedness in parents whose children had been referred to Child and Adolescent Mental Health Services (CAMHS) with a community sample of parents. In addition the relationship between MM, parental depression and stress, and children’s difficulties were also considered in both groups. The two groups were matched as closely as possible on demographics such as parental education, socio-economic-status and living location. This study used CAMHS under-fives services as nearly all of the current mind-mindedness literature relates to parents of children under five. As this is the first study of its kind, pre-schoolers (three to five years old) were chosen as the focus. The reason for this age range was both the higher number of pre-schoolers referred to under-fives services compared to infants and also potential problems identified in previous research using the MM interview measure with younger children (Bernier & Dozier, 2003).

Previous mind-mindedness research has been carried out with both mothers and fathers (Lundy, 2003; Arnott & Meins, 2007), and therefore this study included whichever parent was considered the child’s primary care-giver. Finally, the distinction between the quality and the quantity of mind-related comments may be especially important in clinical samples, as while appropriate comments predict children’s ToM skills (Meins et al., 2003) it was anticipated that high levels of inappropriate comments would relate to outcomes such as child difficulties. As the existing pre-school MM measure did not include a screening for the quality of mental comments, an exploratory addition to the coding was considered appropriate in the present study.

Method

The general aim of the study was to investigate differences in mind-mindedness between a clinical and a community sample and also to explore the relationship between parental mind-mindedness and depression, parenting stress and children’s emotional and behavioural difficulties. The specific hypotheses of the current study were: (1) mind-mindedness will be lower in parents of pre-school children referred to clinical services compared to the community sample; (2) parental mind-mindedness in a clinical sample will be negatively correlated with parental depression, as previously found in non-clinical samples; (3) parental mind-mindedness will be negatively correlated with parenting stress in both groups, and (4) parental MM scores will be negatively correlated with children’s behavioural difficulties in both groups. Additionally, this study included an explorative addition to the pre-school measure.

Participants

Participants were 49 parents (47 mothers and two fathers) of children aged three-five years old, living in a large British city. The clinical sample consisted of 24 parents whose children were referred to under-fives counselling clinics within a Child and Adolescent Mental Health Service (CAMHS) for a range of emotional and behavioural issues, such as aggression, tantrums, anxiety, and sleeping and eating difficulties. All but two of the clinical sample’s children attended nursery. The comparison group of 25 participants were a community sample of parents with children not currently attending CAMHS, recruited from local nurseries or Early Years centres in the same locality as the CAMHS. Parents were excluded if their command of the English language was not sufficient to understand and complete the interview and questionnaires.

In the clinical sample a total of 39 eligible parents were invited to take part during the data collection phase, and 31 agreed to participate (79% response rate). Of these, seven were excluded or dropped out, primarily due to difficulties scheduling or attending appointments. With the community sample, three nurseries were selected on the basis of having similar demographics and geographical locations to the CAMHS clinics and all agreed to take part. A total of 60 parents were approached to take part in the research and 29 agreed to participate (48% response rate). Due to difficulties scheduling appointments or very high levels of educational attainment (uncommon in the CAMHS group), four parents were excluded. All participants considered themselves to be the child’s primary caregiver. Demographic characteristics of the samples will be given in the results section.

Procedure

All participants were recruited via letters from the main author, which was given to them via either their therapist (parents of the clinical sample) or nursery manager (community sample). Parents took part in a short audio-recorded interview asking them to describe their child, without their child present. In addition, they were asked to provide demographic information and complete three questionnaires that measured parental stress, parental depression, and children’s level of emotional and behavioural difficulties. Interviews took place at either one of the CAMHS clinics (parents of the clinical sample of children) or at the child’s place of day-care (community sample). For parents of the clinical sample, interviews were conducted before their fourth session at CAMHS in an attempt to avoid effective treatment negating potential differences between the two groups.

Measures

Mind-Mindedness (MM)

Parental MM was measured using Meins et al.’s (1998) interview, consisting of one main question: “Can you describe [child] for me?”. If parents asked for guidance on the kind of response required they were told that there were no right or wrong answers and they should simply describe whichever characteristics came to mind. The audio was transcribed verbatim and coded for MM using Meins and Fernyhough’s (2006) coding scheme. As per the manual each attribute which the parent mentioned was placed into one of four categories: mental, behavioural, physical or general.

Mental attributes included any reference to the child’s mental life, such as the child’s mind, will, imagination, knowledge, memory or meta-cognition. Comments coded as mental therefore might be those such as “wilful”, “clever”, “conscientious” or “has a mind of their own”. Also included in this category were comments relating to children’s likes, dislikes and wishes or desires, as well as their emotions, such as ““happy”, “manipulative” or “loving”. Behavioural attributes included any reference to behaviour, such as activities, games and interactions on a behavioural level, such as “talkative”, “energetic” or “loud”. Comments using descriptions like “lively”, “bubbly”, “fun”, “naughty”, “cheerful” or “well-mannered” were also coded as behavioural as they could be interpreted as non-mentalistic. Physical attributes included any physical descriptions such as comments about the child’s age, appearance, health or position in the family. General comments were those which related to the child but did not fit into any of the other three categories such as “lovely”, “normal child”, “a pleasure”. The number of mental attributes was then calculated as a proportion of the total number of attributes mentioned in the interview, in order to control for differences in verbosity between participants (Meins & Fernyhough, 2006).

In addition to the standard coding scheme this study added an exploratory extra component to the coding: all mental attributes were then further coded as positive, negative or neutral. Therefore comments such as “considerate”, “clever”, or “loving” would be coded positive, comments such as “wilful” or “mind of their own” would be neutral and comments such as “manipulative” or “spiteful” would be considered negative. This aspect of coding had not been used in previous research at the time this study was carried out however it is of note that since completion a similar addition of emotional valence has been included in another research study (Demers, Bernier, Tarabulsy & Provost (2010). The infant version of the MM measure does include criteria for determining if a mental comment is appropriate or inappropriate. Although the coding used in this study is clearly different to the infant coding as it cannot measure accuracy, the study designed this addition to the pre-school measure to give some indication of emotional quality of comments.

All the interviews were conducted by the first author, who then transcribed and coded them. All anonymized transcriptions were also coded by a second coder blind to the groups and the study hypotheses. Inter-rater agreement for the assignment of a comment to one of the four categories was high (k = 0.89) and inter-rater agreement for the coding of mental comments as positive, neutral or negative was equally high (k = 0.91), indicating a sufficient level of reliability in both cases (Peat, 2001).

Despite the simplicity of the MM measure, evidence from a range of studies shows good indicators of the measure’s validity and reliability. Temporal stability of the concept of MM was partially demonstrated by Meins and Fernyhough’s (1999) longitudinal study, as several measures of MM were taken over a number of months, resulting in significant correlations across the different time points. Further support for the temporal continuity of the construct arises from the finding that MM at six months strongly predicts MM at four years (Meins et al., 2003). The inter-rater reliability has been shown to be constantly high in all studies (k = 0.90). In terms of construct validity, this measure has generally been shown to interact with related constructs in the expected ways, based on the theory. For instance, this measure is positively related to maternal sensitivity (Meins et al., 1998), non-hostility (Lok & McMahon, 2006) and children’s attachment security (Meins et al., 1998) and predicts children’s ability to understand others’ minds (Meins, 1998; Meins & Fernyhough, 1999). Furthermore, research has shown MM to be unrelated to maternal education, socio-economic status (SES), maternal age, marital status (Beiner & Dozier, 2003) or children’s general cognitive abilities (Meins & Fernyhough, 1999). Although further work could be done, the findings from the research available indicate reasonable levels of reliability and validity.

Child behavioural and emotional difficulties

The Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997) was used to measure the severity of children’s difficulties. The SDQ is a brief behavioural screening questionnaire which has been shown to provide a valid and reliable measure of potential behavioural and emotional difficulties in children 3-16 years old (Goodman, 2001; Goodman & Scott, 1999). The questionnaire gives both a total score and five subscales: emotional symptoms, conduct problems, hyperactivity, peer-problems and pro-social behaviour.

Depression

Current levels of parental depressive symptoms were assessed using the Beck Depression Inventory II; a well validated 21-item self-report measure (BDI-II; Beck, Brown & Steer, 1996).

Parenting stress

The Parenting Stress Index—Short Form (PSI/SF; Abidin, 1995) was used to measure parents’ stress. It is not a measure of general stress, but rather of stress regarding the parenting of a particular child. The PSI/SF is a 36-item self-report measure; this gives a total measure of perceived overall parenting stress, and three subscales: parent distress, child difficulty and parent-child interaction. For each of the 36 items respondents choose their level of agreement with a statement on a five-point scale which ranged from “strongly agree” to “strongly disagree”. The psychometric properties of the PSI are good (Abidin, 1997) and the feasibility of the PSI short form has recently been ascertained (Reitman, Currier & Stickle, 2002). The short form shows high internal consistency for all its scales and good test-retest reliability. Conclusions regarding the validity for the short form are mostly drawn from the exceptionally high correlation between the two measures (r = .94).

Demographic information

Participants were asked to complete basic demographic information concerning themselves and their child. This included the child’s gender, age, birth order and whether they attended nursery, as well as parent’s age, ethnicity, occupation, education level and marital status. Socio-economic status was based on parental occupation and education (Mueller & Parcel, 1981). A family was considered to be ‘status 1’ if parents had manual or unskilled jobs and the parent interviewed had left school at the minimum age for leaving. A family was considered to be ‘status 2’ if parents had skilled/managerial/professional positions and if the interviewed parent had gone on to further education. Only one problematic complication arose where a parent had left school at minimum leaving age and also held a managerial position; this case was left out of analysis involving SES.

Results

Overview

The characteristics and normality of variables were explored before preliminary analyses were undertaken comparing the clinical group and the community group in order to determine the demographic similarity of the groups. Next, analyses were conducted to explore differences on the other variables between the two groups and whether any demographic information was related to mind-mindedness. Data were then analysed in relation to the four hypotheses. The three hypotheses relating to whether MM was negatively correlated with depression, stress, and children’s emotional and behavioural characteristics were considered first. This was to ensure that any factors related to MM in both groups could be controlled for when looking at differences between the two groups. Lastly, exploratory analysis relating to the development of the measure was carried out.

Planned Analysis and Power

Hypotheses 1, 2 and 3, predicting that parental mind-mindedness would be negatively correlated with depression, parenting stress and children’s emotional and behavioural difficulties in both groups, were assessed using Pearson’s correlations. Hypothesis 4, predicting that mind-mindedness would be lower in parents of clinical pre-school children compared to the community sample, was assessed using an independent samples t-test.

Power calculations were computed for the proposed analyses using Clark-Carter (2004) and assuming a p-value of 0.05 and power of 0.80. For the t-test a large effect size was assumed, based on the findings of Meins et al. (1998), leading to a sample requirement of 20 people in each group, using a one-tailed test as the direction of the finding is specified. For the correlations a large effect size was also assumed, since only a large effect size was considered to be clinically interesting, as this indicates that 25% of the variance on the other variables is accounted for by the variance in mind-mindedness. Again a one-tailed test was used as the direction of findings is specified, this level of power required a sample size of 24 people in each group. Therefore, full power was reached in both groups for the tests required.

Preliminary Analysis

Interval data were tested for normality and homogeneity of variance in each group. All outcome variables were considered interval data, as suggested by Fife-Schaw (1995). Parental age and education, child age, scores for depression and mind-mindedness, as well as total scores for parental stress and children’s emotional and behavioural difficulties all appeared to meet the criteria. As suggested by Field (2000) skewness and kurtosis of the data were analysed by transforming values provided by SPSS into z-scores. The z-score values for each variable fell below 2.5 and were therefore considered to be acceptably distributed for the relatively small sample. This was confirmed by inspecting the histograms and the non-significant Kolmogorov-Smirnov statistics for each variable. Although the box plots showed two outliers in the clinical group on the PSI and two outliers in the community group on the MM score, when the 5% trimmed means were inspected there was very little difference between the two means in each case, therefore these cases were left in the data file as suggested by Tabachnick and Fidel (2007).

Demographic Information

The demographic characteristics of both groups are shown below in Table 1.

Table 1.

Demographic Characteristics of Clinical and Community Groups

Variable	Clinical group (N = 21)	Community group (N = 25)
Caregiver:
Mother	23 (96%)	24 (96%)
Father	1	1
Ethnicity:
White British	20 (84%)	16 (64%)
Any other white background	2 (8%)	3 (12%)
African	1 (4%)	2 (8%)
Caribbean	0 (0%)	2 (8%)
White/African	1 (4%)	1 (4%)
Asian	0 (0%)	1 (4%)
Mean Age in years:	28.92 (SD = 4.53)	34.56 (SD = 6.08)
Range in years	21-37	23-48
Marital Status:
Single parent	14 (58%)	6 (24%)
Living with partner	10	19
Level of Education:
No Exams	5 (21%)	1 (4%)
GCSEs	8 (33%)	8 (32%)
A-levels	7 (29%)	6 (24%)
Undergraduate	3 (13%)	7 (28%)
Postgraduate	1 (4%)	3 (12%)
Socio-economic-status
Status 1	13 (54%)	9 (36%)
Status 2	11	15
Child gender:
Female	11 (46%)	14 (56%)
Child age in months:	47.21 (SD = 7.72)	44.68 (SD = 5.39)
Range	34-59	36-54
First-born:	15 (63%)	11 (44%)

Comparison of Clinical and Community Samples

A series of chi-square tests for independence (with Yates continuity correction) were used to check for demographic differences between the two groups on categorical variables. Due to the small numbers of participants from a range of different ethnic minorities these categories were collapsed for statistical reasons into two groups, White British or any ethnic minority. The chi-square tests indicated no significant differences in terms of child gender, child being first-born, parent ethnicity or SES. However, the chi-square test for independence indicated that there was a significant association between being a single-parent family and group, χ² (1, 49) = 4.64, p = .03, phi = −.35 (see Table 1).

Independent-samples t-tests were conducted to compare the two groups on the continuous variables of child age, parent age and parent educational level, as well as their scores for depression, parenting stress and children’s strengths and difficulties. There was no significant difference between the clinical and the community sample in terms of child age or parental education level. There was however a significant age difference for parents in the clinical sample and the community sample, t (47) = −3.67, p <.01 (see Table 1). The magnitude of the difference in means (mean difference = −5.64, 95% CI: −8.74 to −2.55) was large (eta squared = .22). There was also a significant difference between the two groups on depression, parenting stress and children’s emotional and behaviour difficulties as shown in Table 2.

Table 2.

Comparing Scores on BDI, PSI and SDQ in the Clinical and Community Groups

Variable	Clinical		Community		t value
Variable	M	SD	M	SD	t value
BDI	17.74	12.11	9.13	6.97	2.97 **
PSI	101.96	24.01	69.28	14.48	5.74 **
SDQ	18.68	7.57	8.64	4.85	5.43 **

Note: A lower number indicates a more positive score. ** p < .01

In summary there appears to be no statistically significant differences between the two samples in terms of child’s age, sex, or whether they were first-born, neither were there significant differences between the two samples in terms of parent’s ethnicity, education or SES. However, the parents in the clinical sample were significantly younger and more likely to be single-parent families. The clinical sample was also likely to rate themselves as having higher depressive symptoms and more parenting stress. They also rated their children as having more emotional and behavioural difficulties.

Mind-Mindedness and Demographics

The relationship was investigated between MM and the variables of child’s age, sex and birth-order, as well as parent’s age, education level, SES, ethnicity, and joint/single parental status. A series of Pearson’s correlations, biserial and point-biserial correlations indicated the MM was unrelated to any of the demographic variables.

Hypotheses

The relationships between mind-mindedness (as measured by the proportional score of mental comments), depression (as measured by the BDI-II), parenting stress (as measured by the PSI-SF), and the child’s emotional and behavioural difficulties (as measured by the SDQ) were investigated using Pearson product-moment correlation coefficients unless otherwise stated. Preliminary analyses were performed to ensure no violation of the assumption of normality, linearity and homoscedasticity.

Hypothesis 1 – MM will be negatively related to depression

In the clinical group there was no significant correlation between MM and depression, r = −.24, n = 24, p = .28. In the community group there was also no significant correlation between the two variables, r = −.20, n = 24, p = .36.

Hypothesis 2 – MM will be negatively related to parenting stress

In the clinical group there was a significant negative correlation between MM and parenting stress, r = −.49, n = 24, p = .01, with higher levels of parenting stress associated with lower levels of mind-mindedness. Using Cohen’s (1988) interpretation this represents a large correlation, indicating a strong relationship between MM and parenting stress. The variance shared by the two variables is therefore 24%. In the community group there was no significant correlation between MM and parenting stress, r = −.19, n = 25, p = .37.

The relationship between MM and the subscales of the PSI were then investigated. In the clinical group there was a strong negative correlation between MM and the subscale of parent distress, r = −.46, n = 24, p = .02 and between MM and the subscale of difficult parent-child interaction, r = −.54, n = 24, p <.01. There was no significant correlation between MM and the subscale of child difficulty, r = −.29, n =24, p = .17. As with the overall PSI score none of the subscales were significantly correlated with MM in the community group.

Hypothesis 3 – MM will be negatively related to children’s behavioural and emotional difficulties

In the clinical group there was no significant correlation between MM and children’s behavioural and emotional difficulties, r = −.23, n = 22, p = .30. In the community group however there was a strong negative correlation between the two variables, r = −.56, n = 25, p < . 01. The variance shared by the two variables is therefore 31%.

The relationship between MM and the subscales of the SDQ were investigated using Spearman rho correlation coefficients as preliminary analysis indicated that these subscales did not meet the assumptions of parametric data. In the community group there was a medium negative correlation between MM and the subscale of conduct problems, r = .41, n = 25, p = .04, and a strong negative correlation between MM and the subscale of hyperactivity, r = .65, n = 25, p < .01. There were no significant correlations between MM and subscales of emotional symptoms, r = −.31, n = 25, p = .13, peer problems, r = .34, n =25, p = .10, or pro-social behaviour, r = .11, n =25, p = .59. As with the overall SDQ score none of the subscales were significantly correlated with MM in the clinical sample.

As can be seen from the correlation matrix in Table 3, depression, parenting stress and children’s emotional and behavioural difficulties were all significantly correlated with each other.

Table 3.

Correlations among BDI, PSI, SDQ and MM for Clinical (N = 21) and Community (N = 25) groups

	BDI	PSI	SDQ	MM
Clinical
BDI	–	.53**	.45*	−.24
PSI		–	.82**	−.49**
SDQ			–	−.23
Community
BDI	–	.42*	.41*	−.20
PSI		–	.50*	−.19
SDQ			–	−.56**

Note: Personal correlation coefficients (two-tailed test of significance). * p < .05 ** p < .01

Hypothesis 4 – MM will be lower in the clinical sample compared to the community sample

As none of the other variables were correlated with MM in both groups, an independent-samples t-test was conducted to compare the mind-mindedness scores of the clinical and community groups. Leven’s test indicated that assumption of homogeneity of variance had been met, F = (1, 47) = .44, p = .51. As expected there was a significant difference in scores for the clinical group (M = .21, SD = .12) and the community group (M = .47, SD = .16), t (44) = −6.39, p < .01. The magnitude of the difference in the means (mean difference = −.26, 95% CI: −.34 to −.18) was very large (eta squared = .46). An ANCOVA was originally considered to partial out the effects of the BDI, SDQ and PSI, however due to the assumption of homogeneity of regression slopes being violated this was not appropriate (Clark-Carter, 2004; Field, 2000). Possible interaction effects will be addressed in the discussion.

Exploratory analysis

The support for Hypothesis 4, that MM was lower in the clinical group compared to the community group, was explored further by looking at the type of mental comments parents made. When considering the difference between the two groups according to whether the mental comments made were positive, neutral and negative, an interesting pattern emerged. These subscales failed to meet the assumptions necessary for parametric statistics, as they were not interval data or normally distributed, and therefore non-parametric tests were used. A Mann-Whitney U test revealed a significant difference in the number of positive comments made by the clinical group (Md = 1.00, n = 24) and the community group (Md = 4.00, n = 25), U = 118.00, z = −3.68, p = <.01, r = .53. Given that the community group made more mind-related comments overall this is unsurprising. However, a second Mann-Whitney U test revealed the difference in the number of negative mental comments made by the clinical group (Md = 1.00, n = 24) and the community group (Md = .00) approached significance, U = 190.00, z = −1.83, p = .06, r = . 27. What is particularly striking is that inspection of the means and medians revealed that the clinical group had a higher number of negative mental comments than the community group, despite making fewer overall mental comments.

In order to control for overall differences in the number of mental comments made, the number of positive, negative and neutral mental comments was also calculated as a proportion of the total number of mental comments. Using this figure a Mann-Whitney U test revealed no significant difference in the proportion of positive comments made in the clinical group (Md = .45, n = 24) compared to the community group (Md = 63), U = 234.00, z = −1.32, p = .19. There was however a significant difference in the proportion of negative mental comments between the clinical group (Md = 0.25, n = 24) and the community group (Md = .00, n = 25), U = 211.50, z = −2.05, p = .04, r = .29. Using Cohen (1988) criteria this represents a medium effect size. There was also a significant difference in the proportion of neutral mental comments between the clinical (Md = .15, n = 24) and the community group (Md = .35, n = 25), U = 192.50, z = −2.17, p = .03, r = .31 which again represents a medium effect size. Therefore the clinical group used a significantly higher proportion of negative mental comments and a significantly lower proportion of neutral comments than the community group, when describing their child. Although the community group used significantly more positive comments overall, when positive comments were calculated as a proportion of all mental comments there was no significant difference between the groups.

Summary

The overall proportion of mental comments, and therefore mind-mindedness, was significantly lower in the clinical group compared to the community group. Overall ratings of depression, parenting stress and child difficulties were significantly higher in the clinical compared to the community group. Interestingly, mind-mindedness did not appear to be correlated with depression in either group. However, mind-mindedness was negatively correlated with parenting stress in the clinical group, with no correlation in the community group. In addition, mind-mindedness was negatively correlated with children’s difficulties in the community group, with no correlation in the clinical group. With regard to the more exploratory examination of MM ‘mental’ comments, whilst there was, no difference in the proportion of positive comments between the two groups, the proportion of negative mind-related comments was significantly higher in the clinical group and the proportion of neutral comments was significantly lower in the clinical group.

Discussion

The main aim of this study was to extend current parental mind-mindedness research to a clinical sample of pre-school children. The first specific aim was to establish if parents of pre-schoolers attending under-fives clinical services have lower mind-mindedness than a comparison group of parents not attending clinical services. The results supported this hypothesis and showed a large effect size. There were no differences in the samples on demographic variables such as child’s age, gender, parents’ educational level or SES, therefore differences are unlikely to be due to these variables. It should be noted however, that due to the correlational nature of this design, no causal links can be drawn from any of the findings in this study. Parents may be lower in mind-mindedness due to child difficulties/referrals to services, or alternatively mind-mindedness may influence these factors. Another possibility is that the relationship between these two factors may be due to a third variable. Although an effort was made to match the samples in terms of demographics, there were differences in the samples relating to both single parent status and parental age. It is possible that these differences may however be representative of the populations from which the samples were drawn, with children attending CAMHS being more likely to have young single parents than children in the general population. However, none of these variables appeared to be related to mind-mindedness, making it less likely that the difference in mind-mindedness was entirely due to any one of these factors. Caution must be taken, however, when interpreting the non-significant findings due to the small samples size meaning there was only enough power to detect larger effect sizes.

The disparity between the two groups on the BDI, SDQ and PSI may also be connected to the differences in MM, however due to the fact these variables related to MM in contrasting ways in the two groups it was not possible to simply partial out their effects. Instead the complex interactions between these variables would need to be considered separately for each group, and this was beyond the scope of this study. It is also possible some unmeasured variables may account for differences between the mind-mindedness of parents referred and not referred to child clinical services. The complex reasons why mind-mindedness is lower in the clinical sample are yet to be explored and need to be taken into consideration in future research.

In the clinical group the average proportion of mental attributes was 0.21, while it was 0.47 in the community group. This latter score is slightly higher than some previous findings in community samples, although still comparable: Meins et al. (2003) and Lok and McMahon (2006) reported rates of 0.44 and 0.36 respectively. The rates in this study are also remarkably similar to those reported in Meins et al. (1998) which reported rates of 0.48 for parents of securely attached children and 0.24 for the parents of insecurely attached children. Therefore, although previous research has not raised the issue of “good enough” mind-mindedness, perhaps categories of scores could be developed. Although such a system would need further consideration this initial research supports the possibility that a proportional score around 0.20 may be considered low, whereas a score around 0.40 may be more typical.

The second aim of this study was to establish whether mind-mindedness was related to depression in this clinical sample, as has previously been found in some community samples. Although depressive symptoms were significantly higher in the clinical group, surprisingly the results of this study found that depressive symptoms were not related to mind-mindedness in either group, therefore failing to replicate some previous findings (Lok & McMahon, 2006; Lundy, 2003; Meins et al., 2011). One possible reason for this may be due to measurement differences, as while the BDI has been shown to relate to mind-mindedness using the infant measure, the only study to find mind-mindedness was related to depression in parents of pre-schoolers (Lok & McMahon, 2006) used the Epidemiological Studies’ Depression Scale (CES-D; Radloff, 1977). Another possibility is that this study lacked the power to detect smaller effect sizes.

The third aim of this study was to explore whether mind-mindedness was related to parenting stress levels in either group of parents. Interestingly, results indicated that mind-mindedness was related to parenting stress levels in the clinical group but not in the community group. It was also found, however, that there was a significant difference in stress levels between the two groups, with the clinical group displaying higher levels of parenting stress. Therefore it is unclear from this study whether mind-mindedness’s relationship to parenting stress is due to other differences between the groups, or whether mind-mindedness is only related to parenting stress when the stress levels are fairly high, as displayed in the clinical group. Perhaps the most likely explanation is that there are bidirectional relationships between difficulties with children (indicated by referrals to services), higher parental stress and lower mind-mindedness.

Further analysis revealed that for the clinical sample only the parent distress subscale and the parent-child interaction difficulty subscale were related to mind-mindedness, while the child difficulty subscale was not. As the parent-child interaction subscale is closely related to attachment (Abidin, 1995) it is unsurprising that this was related to mind-mindedness. However, the finding that parents’ ratings of their parenting stress and difficulty in the interaction but not their child’s difficulties, were related to mind-mindedness in clinical samples, but not in community samples, adds new information to our understanding of mind-mindedness. This finding implies that being low in MM is not simply related to seeing the child as more difficult but rather that low MM is associated with the stress of the parent-child relationship. Again, due to the small sample size, it will be important for further research to explore the relationship between parenting stress and mind-mindedness in both parents accessing and not accessing clinical services.

The fourth aim of the study was to explore if mind-mindedness was related to parents’ ratings of their children’s emotional and behavioural difficulties. Interestingly, mind-mindedness was only found to be significantly related to this variable in the community group. As expected given the nature of the measure, the clinical group rated their children’s difficulties as significantly higher, therefore it is possible that mind-mindedness is only related to lower levels of child difficulties, as found in the community group.

This study also included an exploratory development to the pre-school measure in order to gain some indication of the quality as well as the quantity of mental comments. Unlike the video-based infant-interaction measure, no objective check for appropriateness of comments was possible, so this study included a simple coding of parents’ mind-related comments about their pre-schoolers as positive, negative or neutral. Results indicated that inter-rater reliability for this additional coding was very high (k = .91). There were a significantly higher proportion of negative mind-related comments, and a significantly lower proportion of neutral comments, in the clinical group compared to the community group. There was no difference between the groups’ proportion of positive comments. Interestingly, a similar pattern of results was found in a related area of research which explored infant disturbance and compared mothers with and without eating disorders. Stein, Woolley, Cooper and Fairburn (1994) found mothers with eating disorders expressed the same amount of positive comments as those without, but expressed a greater number of negative comments while feeding their infants. This pattern of findings may indicate the possibility that child difficulties are related to greater exposure to negative feedback, rather than lack of positive feedback. Future research might consider including this addition to the MM pre-school measure and also exploring further the finding that the quality, as well as the quantity, of mind-related comments varied between the clinical and the community samples.

As previously discussed, there are important limitations to this study such as the small sample size, medium power and non-causal nature of the findings, all of which should be taken into account when considering future research. It is also important to remember that all the measures in this study were self-report measures. No observational data or standardized measures were available about parents’ or children’s actual difficulties, therefore limiting this study to the consideration of the relationship between mind-mindedness and parents’ perceptions of these difficulties. As this study did not specifically measure attachment the relationship between attachment and MM is not known for either group, nor is it known how this relationship might account for differences between the two groups.

Conclusions

Although preliminary, this is the first study to extend mind-mindedness research into the clinical field of young children experiencing emotional and behavioural difficulties. Despite not knowing precisely which factors contribute to lower mind-mindedness in the clinical sample, the findings are in line with previous research indicating more generally that parent cognitions may be important in mediating the relationship between parent and child difficulties (Dix, Ruble et al., 1986; Sadler, Slade & Mayes, 2006; Sharp, Fonagy & Goodyer, 2006; Wheatcroft & Creswell, 2007). The findings also add support to the assertions of previous mind-mindedness research (Arnott & Meins, 2007; Bernier & Dozier, 2003) suggesting this construct could have implications for clinical services working with under-fives. Currently, attachment-based interventions often focus on parental sensitivity in responding to children however a recent review of such interventions only found them to be moderately effective (Prior & Glaser, 2006). As some research has suggested that mind-mindedness might be a better predictor than sensitivity of attachment status (Meins et al., 2002) or at least be a prerequisite for sensitivity (Laranjo et al., 2008), it is worth considering whether focusing on mind-mindedness in interventions can improve outcomes for families. In future, if it is replicated that clinical samples of young children have parents with lower MM than their community comparisons, and if causality is established with parental MM predicting referrals to services, then this may indicate a need for early interventions that aim to help increase MM. Such interventions may turn out to be particularly significant in the future given that the other known antecedent of child attachment status, adult attachment stance, seems to be quite resistant to change (Korfmacher et al., 1997). However, it is important to note that as yet no information is available regarding whether mind-mindedness is amenable to intervention, therefore careful evaluation of any such interventions would be critical. A recent review of a wide range of interventions aimed at supporting the attachment relationship concluded that further work is needed on which aspects of intervention are effective and what outcomes should be used to measure effectiveness (Berlin et al., 2008). Therefore, it may be useful for future research to assess any changes in mind-mindedness during currently available attachment-based interventions, as mind-mindedness may already be an effective but unrecognized aspect of some interventions.

Footnotes

Author biographies

Tansy Walker is a Clinical Psychologist with a particular interest in parent-child interactions and attachment theory. She obtained her MSc in Developmental Psychology from Sussex University and her DClinPsy from Canterbury Christ Church University, Salomons Campus.

Rebecca Wheatcroft is a Clinical Psychologist in Greenwich Child and Adolescent Mental Health Services, London. She obtained her DClinPsy at UCL and her MSc in Child Development from the Institute of Education, and has previously worked in research on the children of mothers with eating disorders. She has a particular interest in working clinically with the parents of under fives.

Paul Camic is Reader in Clinical and Health Psychology and Research Director in the Department of Applied Psychology, Salomons Campus, Canterbury Christ Church University, Tunbridge Wells, Kent, UK.

References

Abidin

(1995). Parenting stress index –III: professional manual. Lutz, FL: Psychological Assessment Resources.

Abidin

(1997). Parenting stress index: A measure of the parent-child system. In Zalaquett

C. P.

Wood

(Eds.). Evaluating stress: A book of resources, pp. 277–291. Lanham, MD: Scarecrow Press, Inc.

Ainsworth

Bell

Stayton

(1971). Individual differences in Strange Situation behaviour of one-year-olds. In Schaffer

H. R.

(Ed.), The origins of human social relations. New York: Academic Press.

Ainsworth

Bell

Stayton

(1974). Infant-mother attachment and social development: Socialisation as a product of reciprocal responsiveness to signals. In Richards

M. P.

(Ed.), The introduction of the child into a social world. London: Cambridge University Press.

Arnott

Meins

(2007). Links among antenatal attachment representations, postnatal mind-mindedness, and infant attachment security: A preliminary study of mothers and fathers. Bulletin of the Menninger Clinic, 71, 132–149.

Atkinson

Goldberg

(Eds.). (2004). Attachment issues in psychopathology and intervention. London: Lawrence Erlbaum Associates.

Beck

A. T.

Brown

Steer

R. A.

(1996). Beck depression inventory II manual. San Antonio, TX: The Psychological Corporation.

Beck

C. T.

(2001). Maternal depression and child behaviour problems: A meta-analysis. Journal of Advanced Nursing, 29, 623–29.

Berlin

Zeanah

Lieberman

(2008). Prevention and intervention programs for supporting early attachment security. In Cassidy

Shaver

(Eds.), Handbook of Attachment: Theory, Research and Clinical Applications, pp. 745–61. New York: Guildford Press.

10.

Bernier

Dozier

(2003). Bridging the attachment transmission gap: The role of maternal mind-mindedness. International Journal of Behavioral Development, 27, 355–365.

11.

The British Psychological Society. (2007). Attachment theory into practice. Leicester: Author.

12.

Broadhead

Chilton

Crichton

(2009). Understanding parental stress within the Scallywags service for children with emotional and behavioural difficulties. Emotional and Behavioural Difficulties, 14, 101–115.

13.

Cassidy

Shaver

(Eds.). (2008). Handbook of attachment: Theory, research, and clinical applications (2nd ed.). New York: Guildford Press.

14.

Clark-Carter

(2004). Quantitative psychological research: A student’s handbook (2nd ed.). Hove: Psychology Press.

15.

Cohen

J. W.

(1988). Statistical power analysis for behavioural sciences (2nd ed.). New York: Erlbaum.

16.

Demers

Bernier

Tarabulsy

G. M.

Provost

M. A.

(2010). Maternal and child characteristics as antecedents of maternal mind-mindedness. Infant Mental Health Journal, 31, 94–112.

17.

Department for Education and Skills. (2002). Birth to three matters: A framework to support children in their earliest years. London: Author.

18.

Department for Education and Skills. (2007). The early years foundation stage. London: Author.

19.

Department of Health. (2004). National service framework for children, young people and maternity services. London: Author.

20.

Department of Health. (2008). Children and young people in mind: the final report of the National CAMHS Review. London: Author.

21.

Department of Health. (2010). Keeping children and young people in mind: the Government’s full response to the independent review of CAMHS. London: Author.

22.

De Wolff

van Ijzendoorn

(1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68, 571–591.

23.

Dix

Ruble

Crusec

Nixon

(1986). Social cognition in parents: Inferential and affective reactions to children in three age levels. Child Development, 57, 879–894.

24.

Field

(2000). Discovering statistics using SPSS for Windows. London: Sage.

25.

Fife-Schaw

(1995). Levels of measurement. In Breakwell

G. M.

Hammond

Fife-Schaw

(Eds.), Research methods in psychology. London: Sage.

26.

Fonagy

Steele

Moran

G. S.

Higgitt

A. A.

(1991). The capacity for understanding mental states: The reflective self in parent and child and its significance for security of attachment. Infant Mental Health Journal, 12(3), 201–218.

27.

Goodman

(1997). The strengths and difficulties questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38, 581–586.

28.

Goodman

(2001). Psychometric properties of the strengths and difficulties questionnaire. Journal of American Academy of Child and Adolescent Psychiatry, 40, 1337–1345.

29.

Goodman

Scott

(1999). Comparing the strengths and difficulties questionnaire and the child behavior checklist: Is small beautiful? Journal of Abnormal Child Psychology, 27, 17–24.

30.

Korfmacher

Adam

Ogawa

Egeland

(1997). Adult attachment: Implications for the therapeutic process in a home visitation intervention. Applied Developmental Science, 1, 43–52.

31.

Laranjo

Bernier

Meins

(2008). Associations between maternal mind-mindedness and infant attachment security: Investigating the mediating role of maternal sensitivity. Infant Behavior and Development, 31, 688–695.

32.

Laranjo

Bernier

Meins

Carlson (2010). Early manifestations of children’s theory of mind: The roles of maternal mind-mindedness and infant security of attachment. Infancy, 15, 300–323.

33.

Lok

S. M.

McMahon

C. A.

(2006). Mothers’ thoughts about their children: Links between mind-mindedness and emotional availability. British Journal of Developmental Psychology, 24, 477–488.

34.

Lundy

(2003). Father- and mother-infant face-to-face interactions: Differences in mind-related comments and infant attachment? Infant Behavior & Development, 26, 200–212.

35.

Meins

(1997). Security of Attachment and the Social Development of Cognition. Hove: Psychology Press.

36.

Meins

(1998). The effects of security of attachment and maternal attribution of meaning on children’s linguistic acquisitional style. Infant Behavior and Development, 21, 237–252.

37.

Meins

(1999). Sensitivity, security and internal working models: Bridging the transmission gap. Attachment and Human Development, 1, 325–342.

38.

Meins

Fernyhough

Arnott

Leekam

Turner

(2011). Mother- versus infant-centered correlates of maternal mind-mindedness in the first year of life. Infancy, 16, 137–165.

39.

Meins

Fernyhough

(1999). Linguistic acquisitional style and mentalising development: The role of maternal mind-mindedness. Cognitive Development, 14, 363–380.

40.

Meins

Fernyhough

(2006). Mind-mindedness coding manual. Unpublished manuscript. Durham University, Durham, UK.

41.

Meins

Fernyhough

Fradley

Tuckey

(2001). Rethinking maternal sensitivity: Mothers’ comments on infants’ mental processes predict security of attachment at 12 months. Journal of Child Psychology and Psychiatry, 42, 637–648.

42.

Meins

Fernyhough

Russell

Clark-Carter

(1998). Security of attachment as a predictor of symbolic and mentalising abilities: A longitudinal study. Social Development, 7, 1–24.

43.

Meins

Fernyhough

Wainwright

Clark-Carter

Gupta

M. D.

Fradley

. (2003). Pathways to understanding mind: Construct validity and predictive validity of maternal mind-mindedness. Child Development, 74, 1194–1211.

44.

Meins

Fernyhough

Wainwright

Gupta

M. D.

Fradley

Tuckey

(2002). Maternal mind-mindedness and attachment security as predictors or theory of mind understanding. Child Development, 73, 1715–1726.

45.

Mueller

Parcel

(1981). Measures of socioeconomic status: Alternatives and recommendations. Child Development, 52, 13–30.

46.

Muir

Lojkasek

Cohen

(1999). Watch, wait, & wonder: A manual describing a dyadic infant-led approach to problems in infancy and early childhood. Toronto: Hincks-Dellcrest Centre.

47.

Peat

(2001). Health science research: A handbook of quantitative methods. Sydney: Allen & Unwin.

48.

Prior

Glaser

(2006). Understanding attachment and attachment disorders: Theory, evidence and practice. London: Jessica Kingsley Publishers.

49.

Radloff

(1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401.

50.

Raval

Goldberg

Atkinson

Benoit

Myhal

Poulton

. (2001). Maternal attachment, maternal responsiveness and infant attachment. Infant Behavior and Development, 24, 281–304.

51.

Reitman

Currier

Stickle

(2002). A critical evaluation of the parenting stress index-short form (PSI-SF) in a head start population. Journal of Clinical Child and Adolescent Psychology, 53, 384–92.

52.

Sadler

Slade

Mayes

(2006). Minding the baby: A mentalization-based parenting program. In Allen

Fonagy

(Eds.), Handbook of mentalization-based treatment, pp. 271–288. Winchester, UK: John Wiley and Sons Ltd.

53.

Sharp

Fonagy

(2008). The parent’s capacity to treat the child as a psychological agent: Constructs, measures and implications for developmental psychopathology. Social Development, 17, 737–754.

54.

Sharp

Fonagy

Goodyer

(2006). Imagining your child’s mind: Psychosocial adjustment and mothers’ ability to predict their children’s attributional response styles. British Journal of Developmental Psychology, 24, 197–214.

55.

Solihull NHS Care Trust (2005). The Solihull Approach.

Retrieved 18 June 2009

from: www.solihull.nhs.uk/solihullapproach/default.asp

56.

Stein

Woolley

Cooper

Fairburn

(1994). An observational study of mothers with eating disorders and their infants. Journal of Child Psychology and Psychiatry, 35, 733–748.

57.

Tabachnick

Fidel

(2007). Using multivariate statistics (5th ed.). Boston: Pearson Education.

58.

The Tavistock and Portman Clinic. (2009). The Tavistock Psychoanalytic Model of Work for Children Under Five.

Retrieved 18 June 2009

from: www.tavi-port.org/training/child-and-adolescent-mental-health.html

59.

Thompson

(2008). Early attachment and later development: Familiar questions, new answers. In Cassidy

Shaver

(Eds.), Handbook of attachment: Theory, research, and clinical applications (2nd ed.), pp. 348–365. New York: Guildford Press.

60.

van Ijzendoorn

(1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the adult attachment interview. Psychological Bulletin, 117, 387–403.

61.

Wheatcroft

Creswell

(2007). Parents’ cognitions and expectations about their pre-school children: The contribution of parental anxiety and child anxiety. British Journal of Developmental Psychology, 25, 435–441.

62.

World Health Organization. (2004). The importance of caregiver-child interactions for the survival and healthy development of young children. Geneva: Author.