Paternal involvement and infant-father attachment: a Q-set study.

The primary purpose of this report was to investigate the correlates of attachment security with fathers and the concordance of mother-infant and father-infant attachment as measured by the Attachment Q-set (AQS, Waters, 1987). Sixty fathers and mothers of 14-month old infants independently described their child using the 90-item AQS and completed questionnaires about their involvement in and attitudes toward child-rearing and self-esteem. Mother-child interactions were observed in a play situation. Fathers reporting greater engagement in child caretaking activities described their children as more securely attached. Significant concordance between parents' AQS was obtained; however, this concordance was not due to generalization of interaction patterns across parents. That is, observed maternal sensitivity predicted maternal but not paternal security scores from the AQS.

Keywords: father-infant attachment, Attachment Q-Set, father involvement, mother-father attachment concordance

**********

In the theoretical writing that constituted the major impetus for studying attachment, Bowlby (1969) proposed that infants construct attachment relationships with many of the salient persons in their lives, especially their parents. Nevertheless, because mothers are typically children's primary caregivers, and because fathers have proven to be difficult to recruit and maintain as subjects in parent-infant research programs, research effort has concentrated on the mother-infant relationship. Although great strides have been made in our knowledge base regarding the father-child relationship (see Lamb, 1997; Tamis-LeMonda & Cabrera, 2002) the database of empirical findings regarding infants' attachments with mothers is far richer than the analogous base of empirical findings for infants' attachments with fathers.

This absence of research on infant-father attachment is particularly intriguing given the likelihood that mothers and fathers play markedly different roles in the lives of infants. Fathers engage in greater active social play (proportion of their total interaction time) with and less caregiving (both absolute amount and as a relative proportion of interaction time) of infants than do mothers; mothers perform more caretaking tasks and are more nurturing and accessible to their infants than are fathers (Clarke-Stewart, 1978; Crawley & Sherrod, 1984; Frascarolo-Moutinot, 1994; Lamb, 1977a; Parke, 1979). When involved in caretaking, fathers tend to be less engaged and reinforcing (Lamb & Oppenheim, 1989) and less sensitive to infant cues than are mothers (Donate-Bartfield & Passman, 1985; Heermann, Jones, & Wikoff, 1994). These differences between parental-interaction profiles notwithstanding, most attachment researchers endorse the notion that infants construct attachment relationships with both parents between the ages of eight and 18 months (Kotelchuck, 1972, 1976; Lamb, 1977b, 1997, 2002; Lamb, Frodi, Hwang, & Frodi, 1983; Main & Weston, 1981; Schaffer & Emerson, 1964).

Whether fathers might serve as attachment figures at all was the central question addressed in the earliest research reports on father-infant relationships (e.g., Kotelchuck, 1976; Pedersen & Robson, 1969; Schaffer & Emerson, 1964). These early studies established that infants direct attachment behavior to fathers and generally prefer fathers to strangers in times of stress. Results reported by Pedersen and Robson (1969) and by Chibucos and Kail (1981) indicated that the intensity of the infant's attachment behavior varied as a function of qualities of the father's interaction style, as is true in analyses of relations between mother-infant interaction quality and infant attachment behavior (e.g., Ainsworth, Blehar, Waters, & Wall, 1978; Belsky, Rovine, & Taylor, 1984; Smith & Pederson, 1988). However, the intensity of attachment behavior directed to fathers and the emergence of these behaviors in developmental time suggested that fathers might not be first (or primary) attachment figures for most infants (e.g., Lamb, 1977b; Spelke, Zelazo, Kagan, & Kotelchuck, 1973).

The second ripple of father-infant attachment studies considered the possibility that attachment patterns across parents would show concordance (e.g., Beslky, Garduque, & Hrncir, 1984; Chase-Lansdale & Owen, 1987; Grossmann, Grossmann, Huber, & Wartner, 1981; Lamb, Hwang, Frodi, & Frodi, 1982; Lamb, 1978; Main & Weston, 1981). The results of these early studies and others since then (e.g., Rosen & Rothbaum, 1993; Steele, Steele, & Fonagy, 1996) suggested that attachment security assessed in the Ainsworth Strange Situation (Ainwsorth et al., 1978) might have low to moderate degrees of concordance between parents of a given infant. These findings are of some conceptual importance for attachment theory because there is considerable evidence to the effect that mother-infant attachment quality is an important determinant of later modes of adaptation with other persons (e.g., Kavesh, 1991; Main, Kaplan, & Cassidy, 1985; Pastor, 1981; Sroufe, 1983; Suess, Grossmann, & Sroufe, 1992). If attachment with the mother occurs earlier and is the primary attachment for most infants, it is curious that the quality of attachment with the mother would show little significant relation to the quality of attachment with father (Fox, Kimmerly, & Schafer's [1991] and later Van Ijzendoorn & De Wolff's [1997] meta-analyses of these studies, notwithstanding). Presently, the explanation for the failure to find high concordance between parents for attachment quality is that relationships are co-constructions of a parent-infant dyad that are linked to the unique interaction patterns characteristic of the dyad and to the parent's own history of attachment relationships (e.g., Main & Goldwyn, in press).

The finding that concordance of attachment between parents was low sparked interest in the antecedents to attachment for fathers and infants. The most widely studied antecedent of father-infant attachment investigated in the literature is father involvement. Paternal involvement is now understood as a multidimensional construct. Lamb et al. (1985, 1986) have proposed three primary components of father involvement with children. One type of involvement refers to the quantity of time in absolute or relative terms that fathers spend with their children or are available to them if the children so desire or need. Lamb has labeled this dimension as the accessibility component of father involvement. A second dimension, labeled engagement, refers to the time, nature, and quality of father-child interaction. This dimension includes the actual time spent interacting of engaging with their children (the quantitative component of engagement) in either caretaking activities, playing, having fun, etc., and the quality of these interactions (the qualitative component). Finally, the third dimension is labeled responsibility and refers to the extent that fathers are in charge of their children including making decisions about childcare and other important aspects of the child's life. These dimensions of father involvement are different enough that studies investigating correlates of father involvement have found different correlates for each of the three dimensions of father involvement (Barnett & Baruch, 1987; Lamb, 1997; Volling & Belsky, 1991). Investigations of the different dimensions of father involvement as antecedents of attachment security, as measured by the Strange Situation, have yielded inconclusive findings. Studies have shown a minimal relation between father accessibility and infant-father attachment (Lamb, 1987; 1997; Lamb, Pleck, Charnov, & Levine, 1985). In terms of the qualitative component of engagement, initial research efforts have focused on aspects of father-infant interaction that are analogous to those aspects of mother-infant interaction associated with the quality of infant-mother attachment assessed in the Strange Situation (e.g., Braungart-Rieker, Courtney, & Garwood, 1999; Caldera, Huston, & O'Brien, 1995; Cox, Owen, Henderson, & Margand, 1992; Grossmann, Grossmann, Fremmer-Bombik, Kindler, Scheuerer-Englisch, & Zimmermann, 2002; Notaro & Volling, 1999; Volling & Belsky, 1991). In general, the findings of these studies show inconsistent results. For example, in a longitudinal sample of fathers, mothers, and their three-month-old infants, Cox et al. (1992) found that the composite paternal positive interaction in play (a composite of sensitivity, positive affect, animation, attitude toward play, activity, encouragement of achievement, amount of vocalizing to the child, and amount of reciprocal play) when infants were three months old predicted attachment security to father at 12 months. The quantitative component of engagement, however, yielded a negative association to attachment security. Fathers who spent more time with their infants were less likely to enjoy secure attachment relationships with them. It is of interest to note that these same composite variables were not significantly correlated with the quality of infant-mother attachment in this study.

Volling and Belsky (1991) conducted another longitudinal study of fathers, mothers, and their three-month-old infants. Infants and fathers were observed in interaction at three and nine months, and the father's behavior was coded for responsivity, stimulation, caregiving, and affection. None of these qualitative measures of fathering was significantly associated with infant-father security as measured by the Strange Situation.

A final factor considered in studies of antecedents of father-infant attachment is paternal characteristics. In the Cox et al. (1992) study, attitudes toward the infant and the parental role when infants were three months old predicted attachment security to father at 12 months. Psychological factors such as depression and self-esteem also have been investigated as predictors of attachment security in the literature on mothers (e.g., National Institute of Child Health and Human Development Early Child Care Research Network, NICHD ECCR, 1999). The role of these factors in the development of the father-infant relationship has received much less attention. In two studies, fathers with mild depression interacted with their infants less than non-depressed fathers (Field, Hossain, & Malphurs, 1999; Zaslow, Pederson, Cain, Suwalsky, & Kramer, 1985).

The failure to find significant predictive associations between fathers' behavior, attitudes toward child-rearing, psychological characteristics, and infant-father attachment classifications prompted Volling and Belsky (1991) to speculate that the Strange Situation may not be an appropriate tool for evaluating the quality of attachment relationships between infants and fathers.

Given that the interactions between fathers and infants differ from those of mothers and infants in terms of frequency and distribution across domains of interaction (e.g., bodily maintenance, play, affection, etc.), it is possible that the transactions producing secure attachment to fathers are different from those that lead to secure infant-mother attachment. Furthermore, if we accept (at least provisionally) Volling and Belsky's (1991) speculation concerning the validity of Strange Situation classifications for infants and fathers, then it is important that new studies approach the problem of father-infant attachment from different measurement perspectives than has been true in past studies. Consequently, the primary purpose of this report is to investigate the correlates of the attachment relationship between infants and their fathers and to investigate the concordance of mother-infant and father-infant attachment measured by the Attachment Q-set (AQS) (Waters & Deane, 1985; Waters, 1987).

To date, DelCarmen-Wiggins, Huffman, Pedersen, & Bryan's (2000) and Lundy's (2002) studies are the only two other investigations of attachment security with father using the AQS. In Del Carmen-Wiggins et al.'s (2000) study, mothers and fathers completed sorts on their children at three years of age. The authors found a moderate degree of concordance between the sorts of mothers and fathers. In addition, they found that the antecedents investigated, parental affective symptoms and marital quality, were better predictors of father-child than of mother-child attachment security. Lundy's study, however, did not include mothers.

The first purpose of this study was to investigate the relations between paternal psychological characteristics (i.e., self-esteem); child-rearing attitudes; the father's accessibility, engagement, and responsibility; and Q-sort descriptions of the infant's attachment behavior with the father. It was anticipated that father Q-sort descriptions of attachment would be related to levels of personality functioning (i.e., more optimal self-esteem) and to levels of involvement with child-rearing and socialization. Similar measures assessing psychological characteristics were available for the mother to allow a determination of cross-parent influences on Q-sort descriptions. In addition, mothers and infants were observed in a structured interaction and maternal behavior was scored for sensitivity.

A second purpose of this report was to evaluate the degree of congruence in the Q-sort security scores for mothers and fathers. Although studies using the Strange Situation to assess attachment quality have provided equivocal results regarding concordance of attachment across parents, Fox et al. (1991) suggested that many of these studies suffer from problems associated with power (i.e., small samples) and problems of measurement. That is, the Strange Situation does not easily lend itself to a continuous measure of attachment security. The AQS may help address this problem because it does provide a continuous index of attachment security (e.g., Vaughn & Waters, 1990; Waters & Deane, 1985). In addition, the assessments of mother-infant interaction can help establish the validity of Q-sort descriptions and derived attachment security scores for mothers and fathers. To the extent that mothers' interaction variables are related to their own descriptions of their infant's attachment behavior and attachment security, they support current interpretations of the processes leading to attachment security for infants and mothers. However, it maternal interaction qualities also are predictive of fathers' descriptions of child attachment behaviors, we might interpret the father Q-sort descriptions (and by implication the father-infant relationship) as derivative of the mother-infant relationship.

METHOD

PARTICIPANTS

Families with a 14-month-old toddler were recruited from birth records in a mid-sized Midwestern city. Families included in the study were involved in a longitudinal study of infant childcare (32 of these families were also involved in a language study). A total of 60 mothers and fathers participated. There were equal numbers of girls and boys. Twenty-six of the infants were first-born, 24 were second-born, and the rest had three or more older siblings. Twenty infants were in childcare full-time, 20 in part-time care, and 20 were at home full-time with mother. Ninety percent of the families were non-Hispanic White. The ages of the mothers ranged from 21 to 41 years with a mean of 30; fathers' ages ranged from 24 to 42 years with a mean of 32 years. Mothers' and fathers' education ranged from a high school diploma to postgraduate degrees with a mean of four years of post high school education for both parents. The participants were not necessarily students at the university. Occupation was not obtained.

PROCEDURES

One week prior to a visit to their homes, parents were mailed a package of questionnaires including a background history survey, the Clinical Measurement Package, and the Child Rearing Practices Report. Parents were asked to fill out the questionnaires on their own time but to have them ready a week later when the home visit took place.

The home visit began with videotaping the mother and infant while the father completed a Father Care-Taking Questionnaire. Next, the visitor videotaped mothers and their toddlers for 15 minutes while playing with a variety of toys provided. Finally, mothers and fathers were given instructions on how to complete the AQS. Once parents felt comfortable with the procedure, they were each asked to complete independently the initial sort of the revised Waters (1987) Attachment Q-Sort while the investigator was still present. Parents were encouraged to ask questions about the items and the procedures.

MEASURES

Attachment. Each parent completed the 90-item AQS by Waters (1987). During the home visit, mothers and fathers were carefully instructed on how to conduct the sort following the procedures described by Waters (1987). The AQS consists of 90 cards, each describing an attachment-related child behavior. The first step involved conducting an initial sort separating the cards into three piles according to how similar each item was to the target child. One pile of cards had statements that were similar to the target child (A), another pile included items neither similar nor dissimilar or not applicable (B), and a third pile consisted of items that were dissimilar (C). Each parent completed this initial sort on the day of the home visit in order to insure that all parents were equally exposed to all the items prior to the final sort. The cards and written procedures were then left with the parents who were instructed to try to keep the items in mind for one week while observing their child. After a week, parents were asked to complete the final sort by sorting all 90 cards into nine piles of 10 cards each. That is, piles A, B, and C were further divided into three piles each to arrive at nine final piles. The A pile was subdivided into a pile that contained cards that were "most like my child" (pile 9), "like my child" (pile 8), and "somewhat like my child" (pile 7). The B pile was subdivided into "more like than unlike my child" (pile 6), "neither like nor like my child" (pile 5), and "more unlike than like my child" (pile 4). The C pile was subdivided into "very much unlike my child" (pile 1), "unlike my child" (pile 2), and "somewhat unlike my child" (pile 3). The final nine piles had to contain 10 cards in each. At the end of the week, the interviewer picked up the sorted cards from the parents.

Security scores were computed by correlating the parents' sorts with the criterion sort for security following the procedures recommended by Water and Deane (1985). The resulting correlation was the security score for that child. Note that four fathers did not complete the Q-sort (see Table 1 for means).

Child-rearing attitudes. The revised Block (1965) Child Rearing Practices Report (CRPR; Rickel & Biasatti, 1982) was administered to both mothers and fathers. Rickel and Biasatti condensed the CRPR in order to make it less time-consuming to administer and "to derive a more general and meaningful scoring strategy based on a smaller number of factors" (p. 130) than the original. This revised questionnaire includes 40 items measured on a six-point Likert scale, with responses ranging from 1 = not at all descriptive of me to 6 = highly descriptive of me. The revised CRPR includes two dimensions. Nurturing represents believing in flexible child-rearing practices and attitudes. Examples include statements such as: "I joke and play with my child"; "I trust my child will behave as he should even when I am not there"; "I find some of my greatest satisfactions in my child." Restrictive represents a child-rearing style that is control-oriented. Examples include: "I believe that scolding and criticisms make a child improve"; "I believe that a child should be seen and not heard"; "I instruct my child not to get dirty when he is playing." The factor scores were obtained by selecting each of the items that corresponded to each factor, and multiplying the item's score by the factor loading provided by Rickel and Biasatti. The weighted items were then summed to arrive at the total score for each of the two dimensions (see Table 1 for means). Chronbach's alpha was .70 for nurturing and .81 for restrictive.

Self-esteem. Both parents were given the Clinical Measurement Package: A Field Manual (Hudson, 1982) to ascertain parental self-esteem. The scale consists of 25 items designed to measure how a person feels about him- or herself. Subjects are asked to rate on a five-point Likert-type scale from rarely or never to most of all of the time how often each item applies to them. After reversing appropriate items a total score was calculated so that a high score indicated high self-esteem (see Table 1 for means). Chronbach's alpha was .92.

Father's involvement. Fathers were asked to fill out the Father Care-Taking Questionnaire, which was developed for this study to assess the extent of fathers' accessibility and engagement with the child on a daily basis (see Table 2 for items and corresponding means). For accessibility fathers were asked to state how many hours per week (weekdays and weekends) they are home when the child is home and awake. For engagement, fathers were asked to state the numbers of hours they spend playing and reading during weekdays and weekends as well as how many times per week (weekdays and weekends) they perform specific caregiving activities such as bathing, changing diapers, etc. (see Table 3). The questionnaire also asked fathers about the distribution of responsibility in socializing the target child. That is, fathers were asked to state who is primarily responsible for various tasks, such as choosing a pediatrician and teaching the child social skills. Possible responses included mother, father, or both. See Table 4 for distribution of responsibilities. Note that not all items were completed by the 60 subjects. This is because some of the items were not applicable to all families, and fathers left those items blank. For example, not all families had begun teaching manners to their child; thus only 40 fathers responded to this question.

Mother-child interaction. Mother-child pairs were videotaped at their home during a series of structured tasks designed to measure compliance. Mothers were instructed to have the child complete the following tasks: (1) put seven cubes in a cup for one minute; (2) stack five wooden blocks for 30 seconds; (3) complete a puzzle for two minutes; (4) listen to mother read a book for 3 minutes. Mothers were told to try to get the child to perform each task repeatedly for the specified duration of time (e.g., if the child put all seven cubes in the cup before the minute was up, the mother was instructed to dump the cubes out and repeat the task). The same order of presentation was used for each subject.

Maternal ratings. The mother's behavior during the compliance tasks in the larger study was rated on a five-point Likert-type scale (with 1 = never to 5 = always) for various behaviors. For this report, only sensitivity is of concern. Sensitivity referred to how appropriate for the age of the child were her commands and how responsive she was to her child's behavior. Examples of sensitive behaviors included following the child's lead with the toys, responding when the child vocalized, and understanding the child's behavior if the child appeared uninterested in toys. Each toy was coded separately and then collapsed to arrive at one score per mother (see Table 1 for means).

The author coded all subjects. A coder who was unfamiliar with the hypothesis of the study coded 25% of the sample for reliability. Reliability was calculated by computing Pearson correlations between the scores of the two coders. The correlation between raters for sensitivity was r = 91, p <.001.

RESULTS

PRELIMINARY ANALYSES

Preliminary analyses tested associations between the infants' gender, demographic characteristics, and attachment security. None of these analyses yielded significant results; therefore, the demographic variables and gender were not included in further analyses.

Four composite scores were created from the Father Care-Taking Questionnaire: (1) engagement in care included the number of times the father dressed, fed, bathed changed, and put to bed the child per week (summed weekdays and weekends) (M= 24.28, SD = 12.61); (2) engagement in play/read was the number of hours the father spent reading and playing with the child during a week (summed weekdays and weekends) (M = 17.87, SD = 9.26); (3) accessibility was the number of hours per week (seven days) that the father was home when the child was awake (M = 38.35, SD = 12.12); and (4) shared responsibility was a tally of tasks for which both parents were equally responsible in socializing the child (M= 5.87; SD = 1.57). The engagement variable was divided into play/read and caretaking because each might correlate differently with attachment. The shared responsibility variable was chosen rather than paternal responsibility because so few fathers indicated being primarily responsible for any of the given decisions. The variable was obtained by totaling the number of socializing items for which fathers said both parents were equally responsible.

Finally, correlations were conducted between the predictor variables for fathers (Table 5) and mothers (Table 6) separately in order to guide subsequent analyses. The restrictive and nurturing dimensions were significantly correlated for fathers. Therefore, two separate regression analyses were conducted, each including only one of the child-rearing attitudes. These variables were not significantly correlated for mothers.

PATERNAL INVOLVEMENT AND CHARACTERISTICS AS PREDICTORS OF SECURITY OF ATTACHMENT TO FATHER

The four composite variables, engagement in play, engagement in caretaking, accessibility, and shared responsibility, as well as the nurturing attitudes about child-rearing and self-esteem were submitted to a multiple regression analysis with security with father as the dependent variable. A stepwise multiple regression procedure was used since there was no theoretical reason to give any of the variables priority. The final equation was significant. Engagement in caregiving entered first as a significant predictor of attachment security, F (1,55) = 5.10, p < .05, and accounted for 9% of the variance. Fathers who engaged in more feeding, dressing, and diaper-changing had children who scored higher on security of attachment, b = .30, p < .05. Engagement in play/reading, accessibility, shared responsibility, and nurturing child-rearing attitudes were not significantly related to security of attachment to father in the regression analyses or individually in the correlations. Another regression including restrictive attitudes was conducted yielding almost identical results. This latter regression is not discussed further.

To further explore which Q-set items were responsible for the relation between engagement of fathers and security, a post hoc analysis of the items was conducted. The correlations between each of the 90 items of the Q-set and the fathers' frequency of caretaking tasks were investigated. Correlation coefficients appear in Table 7. Fathers who provided care for their children reported having children who laugh and smile easily with many people (item 7), talk to new people if father asks (15), do not lose interest in new adults who annoy child (17), laugh when father teases (27), do not easily become angry with toys (30), want to be the center of father's attention (31), stop misbehaving when told "no" by father (32), try new activities before going to father first (36), are not demanding and impatient with father (38), obey when father asks child to bring something (41), put arms around father or put hands on his shoulder when picked up (53), do not expect father to be interfering with activities when father is simply trying to help (54), go for help before trying something on (63), do not want father's attention for themselves (64), do not want visitors to pay a lot of attention to them (67), do not have cuddly toys or security blankets (73), do not easily become angry with father (79), and do not cry to get what they want from father(81). Thirteen of these items are heavily weighted security items. That is, the distribution of the "ideally" secure child has items 15, 32, 36, 41, 53, and 64 in the three highest piles (7, 8, 9) and items 30, 31, 38, 54, 63, 79, and 81 in the three lowest piles (1, 2, 3).

PARENTAL DISTRIBUTION OF RESPONSIBILITY AND SECURITY OF ATTACHMENT TO FATHER AND MOTHER

To test whether sharing responsibility for the socialization of the child was related to security of attachment with that parent, these variables were submitted to t-tests with security as the dependent measure. Because so few fathers reported being the primary teacher or decision-maker, fathers responding with this option were deleted from the analysis. A total of 11 cases were deleted. For each item, the attachment scores of children whose mothers were the teacher or decision-maker were compared to those for whom both parents were teacher or decision maker using two-tailed t tests. None of the tests yielded significant results (see Table 8).

MATERNAL ATTRIBUTES AS PREDICTORS OF SECURITY OF ATTACHMENT TO FATHER AND MOTHER

To test the prediction of attachment theory that maternal sensitivity predicts security of attachment, this variable and maternal self-esteem, nurturing and restrictive child-rearing attitudes, and the shared responsibility variables were entered into a stepwise multiple regression analysis with attachment as the dependent variable. Maternal sensitivity was a significant predictor of attachment security with mother F(1,58) = 10.35, p < .01, accounting for 15% of the variance. The more sensitive the mothers' behavior was rated, b = .30, p < .01, the higher the security score of their children. Self-esteem entered as a significant predictor in the second step, F Change (1, 57) = 7.28, p < .01, and accounted for 10% of the variance. The higher that mothers rated themselves on self-esteem, b = .38, p < .01, the higher the security scores of their children. None of the other variables entered into the equation.

As a final test of the validity of the Q-set, maternal sensitivity and s self-esteem were submitted to a multiple regression analysis with attachment security to father as the dependent variable. The equation was significant, F(1,58) = 4.81, p < .05. Maternal self-esteem accounted for 8% of the variance. Children whose fathers reported having a secure attachment relationship with father had mothers with higher self esteem (b = .28, p < .05). Maternal sensitivity was not related to attachment security to father.

CONCORDANCE OF FATHER-INFANT AND MOTHER-INFANT ATTACHMENT

Attachment security to father was significantly correlated with attachment security to mother (r = .48, p < .001). Children who were reported to have secure attachment with fathers also scored higher on security of attachment with mothers.

DISCUSSION

This study investigated correlates of infant-father attachment and the convergence of infant-parent attachment as measured by the Attachment Q-set. The findings lend support to the continued investigation of infant-father attachment, to the theoretical implications of attachment theory, and to the validity of the AQS. As could be predicted from the theory, paternal involvement as measured by engagement in caregiving activities was significantly and positively related to attachment security with fathers. Fathers who provided regular care to their children had children whose security scores were higher than children whose fathers were less involved. Although the data on fathers that yielded significant results was obtained by self-report, the findings ate consistent with the implications of the theory. At the very least, it can be stated that fathers who characterized themselves as playing an active role in the caretaking of their infant also described their relationship with their infantas positive and secure.

Examination of the relations between the Q-set items and fathers' caregiving involvement provide further insight into the important features of the infant-father relationship and, in particular, those relationship variables that lead to attachment security. Q-set items with significant associations to father's engagement in caregiving suggest a relationship characterized by sociability with others, independence, obedience, and warmth on the part of the child. The direction of the relations indicates that fathers who are involved in the caregiving of their infants describe their infants as more likely to engage socially with others, play independently with toys, be compliant with the father, and enjoy a warm relationship with the father. These findings are consistent with previous research linking paternal engagement and child outcomes (for complete reviews see Palkovitz, 2002; Pleck, 1997). Specifically, studies have found that children of engaged fathers are more competent in interaction with others (e.g., Amato, 1987; Mosley & Thompson, 1995), exhibit higher levels of self-control (e.g., Amato, 1987), are more able to carry out responsibilities and follow parents' directions (e.g., Mosley & Thompson, 1995), and are generally better adjusted (e.g., McKeown, Ferguson, & Rodney, 1998). These findings also warrant further exploration of the dependence/independence dimension (Waters & Deane, 1985) in relation to fathers. Specifically, the dependence/independence dimension refers to the correlation between a given child's Q-sort description and the Q-sort description of the "ideally" independent child provided by a set of "experts" in the field.

There was no relation between the fathers' child-rearing attitudes, accessibility, or engagement in play with their children and Q-sort security scores. Neither having a nurturing nor restrictive attitude toward child rearing was significantly correlated to attachment security with the father. Perhaps the children in this study were too young for fathers to have accurately completed the child-rearing instrument. Many of the items in the Child-Rearing Practices Report refer to older children requiring the fathers to speculate about their responses rather than to reflect on actual behavior with the child. This failure to find relations between conceptually salient child-rearing attitudes and values is consistent with Volling and Belsky's (1991) report and suggests again that attributes of mothers that are important influences on attachment security may not be as important when father-infant attachment is the focus of research.

The lack of relation between father accessibility and security of attachment to the father was surprising in view of the findings of Cox and colleagues (1992). Cox et al. found a negative relation between time spent with the infant and secure attachment with father when controlling for the behavioral and attitudinal variables. This relation was not anticipated, and they speculated that fathers might be spending time with their infants when the family is under stress, perhaps due to the (un) employment status of the father, hence the negative relation. In this study, the lack of relation between accessibility and security of attachment with father may be due to the relatively low number of hours fathers were accessible to their children. When one considers that there ate 120 hours in a week, a person who works 40 hours per week (eight hours five days a week) and who might sleep for 40 hours in a week (eight hours per night, five days per week) has left 40 hours per week to be accessible to his/her children. The fathers in this study reported an average of 18.62 accessible hours per week, a number that may not be a sufficiently salient factor in the development of attachment security.

The finding of congruence between attachment security with father and mother is consistent with the interpretations offered by Fox et al. (1991) and Del-Carmen-Wiggins et al. (2000) and suggests the possibility that the attachment relationship with one parent may influence the relationship with the other. Of course, the direction of such effects cannot be addressed here. Caution is taken in interpreting these findings, however, as parents are likely to perceive their child similarly, which may be contributing to this concordance.

The relation found between the maternal observations and the Q-Sort provides strength and support to the findings with fathers, the theory, and the measure. According to Bowlby and Ainsworth, the most essential maternal characteristic leading to secure attachment is that of sensitivity to infant communicative signals. That sensitivity was significantly related to security in this study suggests the AQS is measuring behavioral attributes of the infant that are intrinsically linked to maternal sensitivity. To the extent that the father's engagement may serve as an estimate of sensitivity, this finding lends support to the relation found between father involvement and security to father. Additionally, the finding that maternal sensitivity did not relate to attachment security to father strengthens the validity of the Q-Set and suggests that attachment security reflects a relationship attribute and not a temperamental characteristic of the infant.

The results of this study and their interpretation are of course limited by the fact that all of the father data come via father self-reports. They would be strengthened considerably if they were corroborated by observations of fathers with their infants at home by independent observers. Nevertheless, the relations obtained here can be accepted as consistent with current conceptual speculation and empiricism concerning attachments between infants and fathers. As noted earlier in this report, fathers have proven to be expensive as targets of parent-infant interaction research in terms of time and research resources. Data such as these indicate that the expenditure of such resources in the future is likely to prove fruitful.

Table 1
Means and Standard Deviations for Measures of Fathers and Mothers

Variable                          Fathers *

                             Mean         SD

Attachment Security           .29         .20
Self-Esteem                 77.22        14.08
Nurturing                   46.24         5.11
Restrictive                 30.40         7.09
Sensitivity                    -            -

Variable                          Mothers

                             Mean         SD

Attachment Security           .34          .19
Self-Esteem                 73.77        12.58
Nurturing                   47.64         3.24
Restrictive                 28.78         6.21
Sensitivity                  3.72          .79

* N = 56

Table 2
Means and Standard Deviations for Number of Hours per
Week Father is Accessible and Engaged in Play/Reading

Variable                            Mean       SD

Accessibility: Number of hours Father is

    at home weekdays             18.62         8.23
    at home weekends             19.73         7.73

Engagement in play/reading: Number of hours Father is engaged in

    playing weekdays              7.88         5.00
    playing weekends              7.92         4.75
    reading weekdays              1.13         1.27
    reading weekends               .93         1.19

Table 3
Means and Standard Deviations for Number of Times per
Week Father is Engaged in Caregiving with the Child

Variable
                           Mean        SD

dresses child              4.13      3.02
feeds child                5.17      3.80
changes diapers            9.58      6.51
puts child to bed          3.50      2.49
bathes child               1.90      1.76

Table 4

Number of Families in Which Mother, Father, or Both
Are Involved in the Socialization of the Child

                               Mothers    Fathers
Shared responsibility

Teaching the child:

  to relate to others            17          1
  manners                        13          1
  to self feed                   33          0

Which parent chooses:
  childcare for the child        11          0
  the baby-sitters               12          1
  the pediatrician               17          2
  the child's diet               43          1
  clothes for the child          46          1
  toys for the child             15          4

                               Both
Shared responsibility

Teaching the child:

  to relate to others            40
  manners                        26
  to self feed                   26

Which parent chooses:
  childcare for the child        26
  the baby-sitters               47
  the pediatrician               41
  the child's diet               16
  clothes for the child          11
  toys for the child             40

Table 5
Intercorrelations of Paternal Variables

     Variable                     2       3        4

1.   Father Security             .17     .13      .04
2.   Father Self-Esteem                 -.10     -.22
3.   Father Accessibility                         .54 ***
4.   Father Engagement
       in Playing/Reading
5.   Father Engagement
       in caregiving
6.   Shared Responsibility
7.   Nurturing Father
8.   Restrictive Father

     Variable                     5       6        7

1.   Father Security             .29 **  -.07     .12
2.   Father Self-Esteem          .01     -.02     .16
3.   Father Accessibility        .08      .14     .29 *
4.   Father Engagement
       in Playing/Reading        .24      .05     .29 *
5.   Father Engagement
       in caregiving                     -.01    -.02
6.   Shared Responsibility                        .05
7.   Nurturing Father
8.   Restrictive Father

     Variable                      8

1.   Father Security             -.07
2.   Father Self-Esteem          -.12
3.   Father Accessibility         .25
4.   Father Engagement
       in Playing/Reading         .05
5.   Father Engagement
       in caregiving             -.53
6.   Shared Responsibility       -.05
7.   Nurturing Father             .29 *
8.   Restrictive Father

* p<.05. ** p <.01. *** p <.001.

Table 6
Intercorrelations of Maternal Variables

Variable                         2          3

1. Maternal Security           .33 **    -.08
2. Maternal Self-esteem                   .18
3. Nurturing Mother
4. Restrictive Mother
5. Maternal Sensitivity

Variable                         4          5

1. Maternal Security          -.20        .40 ***
2. Maternal Self-esteem       -.04        .04
3. Nurturing Mother            .08       -.13
4. Restrictive Mother                    -.00
5. Maternal Sensitivity

** p<.01. *** p<.001.

Table 7

Significant Correlations Between Q-Set Items and
Paternal Engagement in Caregiving

Item                                       r

7.  Child laughs and smiles easily
    with many people.                     .34 **
15. Child will talk to new
    persons if father asks.               .30 **
17. Child looses interest in new
    adults who annoy child.              -.36 **
27. Child laughs when father teases.      .30 *
30. Child easily becomes angry
    with toys.                           -.32 *
31. Child wants to be the center
    of father's attention.               -.27 *
32. Child stops misbehaving
    when told "no" by father.             .28 *
36. Child tries new activities
    without going to father first.        .29 *
38. Child is demanding and
    impatient with father, fusses
    and persists unless he
    does what child wants.               -.41 ***
41. If father says to bring
    something, child obeys               -.43 ***
53. Child puts arms around
    father or puts hand.
    on his shoulder when picked up.       0.29 **
54. Child expects father to be
    interfering with activities when
    he is simply trying to help.          -.33 **
63. Child goes for help
    before trying something on.           -.30 **
64. Child wants father's
    attention for him/herself.            -.27 *
67. Child wants visitors to
    pay a lot of attention.               -.25 *
73. Child has cuddly toy
    or security blanket.                  -.37 **
79. Child easily becomes
    angry at father.                      -.40 **
81. Child cries to get what
    she/he wants from father.             -26. *

* p<.05. ** p<.01. *** p<.001.

Table 8
Mean Security Scores with Fathers and
Mothers of Families in Which Mother or Both
Parents Are Involved in the Socialization of the Child

                           Security to Father

Which parent               Mothers      Both

Teaching the child ...
  to relate to others      .22(.18)      .33(.20)
  manners                  .24(.22)      .31(.19)
  to self feed             .31(.22)      .29(.15)

Which parent chooses:
  childcare                .38(.13)      .28(.21)
  the baby-sitters         .31(.21)      .29(.20)
  the pediatrician         .31(.18)      .30(.21)
  the child's diet         .29(.21)      .30(.18)
  clothes for the child    .29(.20)      .29(.18)
  toys for the child       .23(.17)      .31(.21)

                           Security to Mother

Which parent               Mothers      Both
Teaching the child ...
  to relate to others      .29(.20)      .36(.19)
  manners                  .35(.18)      .34(.20)
  to self feed             .33(.21)      .36(.18)

Which parent chooses:
  childcare                .41(.16)      .33(.19)
  the baby-sitters         .36(.14)      .34(.21)
  the pediatrician         .35(.17)      .34(.21)
  the child's diet         .32(.22)      .42(.10)
  clothes for the child    .32(.20)      .43(.15)
  toys for the child       .32(.17)      .36(.20)

REFERENCES

Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S.N. (1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, NJ: Erlbaum.

Amato, P.R. (1987). Children in Australian families: The growth of competence. Upper Saddle River, N J: Prentice Hall.

Barnett, R.C., & Baruch, G.K. (1987). Determinants of fathers' participation in family work. Journal of Marriage and the Family, 49, 29-40.

Belsky, J., Garduque, L., & Hrncir, E. 1984). Assessing performance, competence, and executive capacity in infant play: Relations to home environment and security of attachment. Developmental Psychology, 20, 406-417.

Belsky, J., Rovine, M., & Taylor, D. (1984). The Pennsylvania infant and family development project II: Origins of individual differences in infant-mother attachment: Maternal and infant contributions. Child Development, 55, 706-717.

Block, J. (1965). The child-rearing practices report. Berkeley: University of California, Institute of Human Development.

Bowlby, J. (1969). Attachment and loss (Vol. 1). New York: Basic Books.

Braungart-Rieker, J., Courtney, S., & Garwood, M. M. (1999). Mother- and father-infant attachment: Families in context. Journal of Family Psychology, 13, 535-553.

Caldera, Y.M, Huston, A., & O'Brien, M. (1995, April). Antecedents of father-infant attachment: A longitudinal study. Poster presented to the Society for Research in Child Development, Indianapolis, IN.

Chase-Lansdale, P.L., & Owen, M.T. (1987). Maternal employment in a family context: Effects on infant-mother and infant-father attachments. Child Development, 55, 1505-1512.

Chibucos, T., & Kail, P. (1981). Longitudinal examination of father-infant interaction and infant-father interaction. Merril-Palmer Quarterly, 27, 81-96.

Clarke-Stewart, A. (1978). And daddy makes three: The father's impact on mother and young child. Child Development, 49, 466-478.

Cox, M.J., Owen, M.T., Henderson, V.K., & Margand, N.A. (1992). Prediction of infant-father and infant-mother attachment. Developmental Psychology, 28, 474-483.

Cox, M.J., Owen, M.T., Lewis, J.M., & Henderson, V.K. (1989). Marriage, adjustment, and early parenting. Child Development, 60, 1015-1024.

Crawley, S.B., & Sherrod, R.B. (1984). Parent-infant play during the first year of life. Infant Behavior and Development, 7, 65-75.

DelCarmen-Wiggins, R., Huffman, L.C., Pedersen, F.A., & Bryan, Y.E. (2000). Mothers' and fathers' perceptions of three year olds' attachment behavior. Developmental and Behavioral Pediatrics, 21, 97-106.

Donate-Bartfield, D., & Passman, R.H. (1985). Attentiveness of mothers and fathers to their baby' s cries. Child Development, 59, 506-511.

Erickson, M.F., Sroufe, L.A., & Egeland, B. (1985). The relationship between quality of attachment and behavior problems in preschool in a high-risk sample. In I. Bretherton & E. Everett (Eds.), Growing points of attachment theory and research (pp. 147- 166). Monographs of the Society for Research in Child Development, 50 (1-2, Serial No. 209).

Field, T.M., Hossain, Z., & Malphurs, J. (1999). "Depressed" fathers' interactions with their infants. Infant Mental Health Journal, 20, 322-332.

Fox, N., Kimmerly, N.L., & Schafer, W.D. (1991). Attachment to mother/attachment to father: A meta-analysis. Child Development, 62,210-225.

Frascarolo-Moutinot, F. (1994). Engagment paternal quotidian et relations parents-enfant [Daily paternal involvement and parent-child relationships]. Unpublished doctoral dissertation. University of Geneva.

Grossmann, K., Grossmann, K.E., Fremmer-Bombik, E., Kindler, H., Scheuerer-Englisch, H., & Zimmermann, P. (2002). The uniqueness of the child-father attachment relationship: Fathers' sensitive and challenging play as a pivotal variable in a 16-year longitudinal study. Social Development, 11, 307-331.

Grossmann, K.E., Grossmanm K., Huber, F., & Wartner, U. (1981). German children's behavior towards their mothers at 12 months and their fathers at 18 months in Ainsworth's Strange Situation. International Journal of Behavioral Development, 4, 157-181.

Heermann, J.A., Jones, L.C., & Wikoff, R.L. (1994). Measurement of parent behavior during interactions with their infants. Infant Behavior and Development, 17, 311-321.

Hudson, W.W. (1982). A measurement package for clinical workers. Journal of Applied Behavioral Science, 18, 229-238.

Kavesh, L. (1991). Antecedents of peer competence in childhood. Unpublished doctoral dissertation, University of Illinois at Chicago.

Kotelchuck, M. (1972). The nature of the child's tie to his father. Unpublished doctoral dissertation, Harvard University.

Kotelchuck, M. (1976). The infant's relationship to the father: Experimental evidence. In M.E. Lamb (Ed.), The role of the father in child development (pp. 329-344). New York: Wiley.

Lamb, M.E. (1977a). The development of mother-infant and father-infant attachments in the second year of life. Developmental Psychology, 13, 631-648.

Lamb, M.E. (1977b). Father-infant and mother-infant interaction in the first year of life. Child Development, 48, 167-181.

Lamb, M.E. (1978). Qualitative aspects of mother- and father-infant attachments. Infant Behavior and Development, 1, 1-10.

Lamb, M.E. (1986). The changing roles of fathers. In M.E. Lamb (Ed.), The father's role: Applied perspectives (pp. 3-27). New York: Wiley.

Lamb, M.E. (Ed.). (1997). The role of the father in child development (3rd ed.). New York: Wiley.

Lamb, M.E. (2002). Infant-father attachments and their impact on child development. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 93- 117). Mahwah, NJ: Lawrence Earlbum Associates.

Lamb, M.E., Hwang, C.P., Frodi, A., & Frodi, M. (1982). Security of mother- and father-infant attachment and its relations to sociability with strangers in traditional and nontraditional Swedish families. Infant Behavior and Development, 5, 355-367.

Lamb, M.E., & Oppenheim, D. (1989). Fatherhood and father-child relationships: Five years of research. In S.H. Cath, A. Gurwitt, & L. Gunsberg (Eds.), Fathers and their families (pp. 11-26). Hillsdale, NJ: Erlbaum.

Lamb, M.E., Frodi, A.M., Hwang, C.P., & Frodi, M. (1983). Effects of paternal involvement on infant preferences for mothers and fathers. Child Development, 54, 450-458.

Lamb, M.E., Pleck, J.H., Charnov, E.L., & Levine, J.A. (1985). Paternal behavior in humans. American Zoologist. 25, 883-894.

Lundy, B.L. (2002). Paternal socio-psychological factors and infant attachment: The mediating role of synchrony in father-infant interactions. Infant Behavior & Development, 25, 221-236.

Main, M., & Goldwyn, R. (in press). Adult attachment classification system. In M. Main (Ed.), Behavior and the development of representational models of attachment: Five methods of assessment. Cambridge: Cambridge University Press.

Main, M., Kaplan, N, & Casidy, J. (1985). Security in infancy, childhood, and adulthood: Move to the level of representations. Monographs of Social Research in Child Development, 50, 66-104 (Serial No. 209).

Main, M., & Weston, D. (1981). Security of attachment to mother and father: Related to conflict behavior and readiness to establish new relationships. Child Development, 52, 932-940.

Marsiglio, W. (1995). Fatherhood: Contemporary theory, research, and social policy. Thousand Oaks, CA: Sage.

Matas, L., Arend, R., & Sroufe, L. A. (1978). Continuity of adaptation in the second year: The relationship between quality of attachment and later competence. Child Development, 49, 547-556.

McKeown, K., Ferguson, H., & Rodney, D. (1998). Changing fathers? Cork, Ireland: Collins Press.

Mosley, J., & Thomson, E. (1995). Fathering behavior and child outcomes: The role of race and poverty. In W. Marsiglio (Ed.), Fatherhood: Contemporary theory, research, and social policy (pp.148-165). Thousand Oaks, CA: Sage.

NICHD Early Child Care Research Network (1999). Chronicity of maternal depressive symptoms, maternal sensitivity, and child functioning at 36 months. Developmental Psychology, 35, 1297-1310.

Notaro, P.C., & Volling, B.L. (1999). Parental responsiveness and infant-parent attachment: A replication study with fathers and mothers. Infant Behavior and Development, 22, 534-352.

Palkovitz, R. (2002). Involved fathering and child development: Advancing our understanding of good fathering. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 119-140). Mahwah, NJ: Lawrence Earlbum Associates.

Parke, R.D. (1979). Perspectives of father-infant interaction. In J.D. Osofsky (Ed.), Handbook of infant development (pp. 549-590). New York: Wiley.

Pastor, D.L. (1981). The quality of mother-infant attachment and its relationship to toddlers' initial sociability with peers. Developmental Psychology, 17, 326-335.

Pedersen. F.A., & Robson, K. (1969). Father participation in infancy. American Journal of Orthopsychiatrv, 39, 466-472.

Pederson, D., Moran, G., Sitko, C., Campbell, K., Ghesquire, K., & Acton, H. (1990). Maternal sensitivity and the security of infant-mother attachment: A Q-sort study. Child Development, 61, 1974-1983.

Pleck, J.H. (1997). Paternal involvement: Levels, sources, and consequences. In M.E. Lamb (Ed.). The role of the father in child development (3rd ed., pp. 66-103). New York: Wiley.

Rickel, A.U., & Biasatti, L.L. (1982). Modification of the Block Child Rearing Practices Report. Journal of Clinical Psychology, 38, 129-134.

Rosen, K.S., & Burke, P.B. (1999). Multiple attachment relationships within families: Mothers and fathers with two young children. Developmental Psychology, 35, 436-444.

Rosen, K.S., & Rothbaum, F. (1993). Quality of parental caregiving and security of attachment. Developmental Psychology, 29, 358-367.

Schaffer, H.R., & Emerson, P.E. (1964). The development of social attachments in infancy. Monographs of Social Research in Child Development, 29, (Serial No. 94).

Smith, P., & Pederson, D. (1988). Maternal sensitivity and patterns of infant-mother attachment. Child Development, 50, 971-975.

Spelke, E., Zelazo, P., Kagan, J., & Kotelchuck, M. (1973). Father interaction and separation protest. Developmental Psychology, 9, 83-90.

Steele, H., Steele, M., & Fonagy, P. (1996). Associations among attachment classifications of mothers, fathers, and their infants. Child Development, 67, 541-555.

Stroufe, L.A. (1983). Infant-caregiver attachment and adaptation in the preschool: The roots of competence maladaptation. In M. Perlmutter (Ed.), Development of cognition, affect, and social relations (pp. 41-81). Hillsdale, NJ: Erlbaum.

Suess, G.J., Grossmann, K.E., & Sroufe, L.A. (1992). Effects of infant attachment to mother and father on quality of adaptation in preschool: From dyadic to individual organization of self. Infant Behavior and Development, 15, 43-65.

Tamis-LeMonda, C.S., & Cabrera, N. (Eds.). (2002). Handbook of father involvement: Multidisciplinary perspectives. Mahwah, NJ: Lawrence Earlbum Associates.

Van Ijzendoorn, M.H., & De Wolff M.S. (1997). In search of the absent father: Meta-analyses of infant-father attachment: A rejoinder to our discussants. Child Development, 68, 604-609.

Van Ijzendoorn, M.H., & Kroonenberg, P.M. (1988). Cross-cultural patterns of attachment: A meta-analysis of the Strange Situation. Child Development, 59, 147-156.

Vaughn, B.E., & Waters, E. (1990). Attachment behavior at home and in the laboratory: Q-sort observations and strange situation classifications of one-year-olds. Child Development, 61, 1965-1973.

Volling, B.L., & Belsky, J. (1992). Infant, father, and marital antecedents of infant-father attachment security in dual-earner and single-earner families. International Journal of Behavioral Development, 15, 83-100.

Volling, B.L., & Belsky, J. (1991). Multiple determinants of father involvement during infancy in dual-earner and single-earner families. Journal of Marriage and the Family, 53, 461-474.

Waters, E. (1987). Attachment Q-Set. Stony Brook: State University of New York.

Waters, E., & Deane, K.E. (1985). Defining and assessing individual differences in attachment relationships: Q-Methodology and the organization of behavior in infancy and early childhood. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50, 41-65.

Zaslow, M.J., Pedersen, F.A., Cain, R.L., Suwalsky, J.T.D., & Kramer, E.L. (1985). Depressed mood in new fathers: Associations with parent-infant interaction. Psychological Monographs, 111, 135-150.

This research was supported by a Program Project Grant to the University of Kansas and a National Institute of Health Traineeship to the author. The author is grateful to the families who participated in this study.

Correspondence concerning this article should be sent to Yvonne M. Caldera, Department of Human Development and Family Studies, Box 41162, Texas Tech University, Lubbock, TX 79409. Electronic mail: [e-mail].