Sources of Perceived Responsiveness in Family Relationships

Similar documents
The Paradox of Received Social Support

Commitment and Trust in Young Adult Friendships

Attachment Working Models Twist Memories of Relationship Events

Internal Consistency and Reliability of the Networked Minds Social Presence Measure

S P O U S A L R ES E M B L A N C E I N PSYCHOPATHOLOGY: A C O M PA R I SO N O F PA R E N T S O F C H I LD R E N W I T H A N D WITHOUT PSYCHOPATHOLOGY

Optimal Flow Experience in Web Navigation

Running head: INFLUENCE OF LABELS ON JUDGMENTS OF PERFORMANCE

Personality Traits Effects on Job Satisfaction: The Role of Goal Commitment

Chapter 9. Youth Counseling Impact Scale (YCIS)

In this chapter we discuss validity issues for quantitative research and for qualitative research.

HARRISON ASSESSMENTS DEBRIEF GUIDE 1. OVERVIEW OF HARRISON ASSESSMENT

Examining the efficacy of the Theory of Planned Behavior (TPB) to understand pre-service teachers intention to use technology*

Collecting & Making Sense of

Using the STIC to Measure Progress in Therapy and Supervision

The role of partner characteristics in attachment insecurity and depressive symptoms

A Hierarchical Comparison on Influence Paths from Cognitive & Emotional Trust to Proactive Behavior Between China and Japan

Development and Psychometric Properties of the Relational Mobility Scale for the Indonesian Population

Social Determinants and Consequences of Children s Non-Cognitive Skills: An Exploratory Analysis. Amy Hsin Yu Xie

Internal Consistency and Reliability of the Networked Minds Measure of Social Presence

Links between attachment orientations and dispositional and diary-based measures of disclosure in dating couples: A study of actor and partner effects

ABSTRACT PARTNER POSITIVE BEHAVIOR AND RECIPIENT SATISFACTION IN MILD TO MODERATELY ABUSIVE COUPLES AS MODERATED BYATTACHMENT STYLE

Adult attachment, emotional control, and marital satisfaction

Perceptions of Conflict and Support in Romantic Relationships: The Role of Attachment Anxiety

Doing Quantitative Research 26E02900, 6 ECTS Lecture 6: Structural Equations Modeling. Olli-Pekka Kauppila Daria Kautto

Self-Oriented and Socially Prescribed Perfectionism in the Eating Disorder Inventory Perfectionism Subscale

Measurement of communal strength

GOLDSMITHS Research Online Article (refereed)

The moderating effects of direct and indirect experience on the attitude-behavior relation in the reasoned and automatic processing modes.

This self-archived version is provided for scholarly purposes only. The correct reference for this article is as follows:

EMPATHY AND COMMUNICATION A MODEL OF EMPATHY DEVELOPMENT

One-Way ANOVAs t-test two statistically significant Type I error alpha null hypothesis dependant variable Independent variable three levels;

ABSOLUTE AND RELATIVE JUDGMENTS IN RELATION TO STRENGTH OF BELIEF IN GOOD LUCK

Scoring Adults Secure Base Use And Support

Data and Statistics 101: Key Concepts in the Collection, Analysis, and Application of Child Welfare Data

The Logic of Causal Order Richard Williams, University of Notre Dame, Last revised February 15, 2015

A longitudinal study of conflict in new parents: The role of attachment

Web-Based Radio Show. Structure and Behavioral Goals of the DIR /Floortime Program

TERMINOLOGY INSECURITY? APPLICATION OF ATTACHMENT THEORY TO THE HUMAN-COMPANION ANIMAL BOND, AND REVIEW OF ITS EFFECTS

POLI 343 Introduction to Political Research

A Cross-cultural Analysis of the Structure of Subjective Well-Being

Everyday Problem Solving and Instrumental Activities of Daily Living: Support for Domain Specificity

How to use Social Relation Model components as predictor or outcome

3 CONCEPTUAL FOUNDATIONS OF STATISTICS

COACH WORKPLACE REPORT. Jane Doe. Sample Report July 18, Copyright 2011 Multi-Health Systems Inc. All rights reserved.

MULTIPLE LINEAR REGRESSION 24.1 INTRODUCTION AND OBJECTIVES OBJECTIVES

Alternative Methods for Assessing the Fit of Structural Equation Models in Developmental Research

Do You See What I See? Actor and Partner Attachment Shape Biased Perceptions of Partners

Extraversion. The Extraversion factor reliability is 0.90 and the trait scale reliabilities range from 0.70 to 0.81.

CHAPTER 1 Understanding Social Behavior

Field Study Prospectus. Mythical Cliff: You Know It s Cold Outside When You Go Outside and It s Cold. Submitted by

Pearson Education Limited Edinburgh Gate Harlow Essex CM20 2JE England and Associated Companies throughout the world

CURRENT RESEARCH IN SOCIAL PSYCHOLOGY

PERCEIVED TRUSTWORTHINESS OF KNOWLEDGE SOURCES: THE MODERATING IMPACT OF RELATIONSHIP LENGTH

S.A.F.E.T.Y. TM Profile for Joe Bloggs. Joe Bloggs. Apr / 13

PERSON PERCEPTION September 25th, 2009 : Lecture 5

The Effect of Attachment and Sternberg s Triangular Theory of Love on Relationship Satisfaction

Running Head: PERSONALITY AND SOCIAL MEDIA 1

SEMINAR ON SERVICE MARKETING

International Journal of Research and Review E-ISSN: ; P-ISSN:

2018 Gatlinburg Conference Symposium Submission SS-20

How to Manage Seemingly Contradictory Facet Results on the MBTI Step II Assessment

CHAPTER 3 METHOD AND PROCEDURE

An Empirical Study of the Roles of Affective Variables in User Adoption of Search Engines

Confirmatory Factor Analysis of Preschool Child Behavior Checklist (CBCL) (1.5 5 yrs.) among Canadian children

Chapter 7: Descriptive Statistics

11/18/2013. Correlational Research. Correlational Designs. Why Use a Correlational Design? CORRELATIONAL RESEARCH STUDIES

Is Facebook an Accurate Representation of Personality?

Predictors of Caregiving in Adult Intimate Relationships: An Attachment Theoretical Perspective

CHAPTER - 6 STATISTICAL ANALYSIS. This chapter discusses inferential statistics, which use sample data to

TLQ Reliability, Validity and Norms

Perception. Chapter 8, Section 3

Collecting & Making Sense of

Structural Validation of the 3 X 2 Achievement Goal Model

Psychological Experience of Attitudinal Ambivalence as a Function of Manipulated Source of Conflict and Individual Difference in Self-Construal

Basic concepts and principles of classical test theory

Problem-solving behaviors in college relationships

Glossary From Running Randomized Evaluations: A Practical Guide, by Rachel Glennerster and Kudzai Takavarasha

11/24/2017. Do not imply a cause-and-effect relationship

Models of Information Retrieval

The Interactive Effects of Ethical Norms and Subordinate Recommendations on Accounting Decisions

Making a psychometric. Dr Benjamin Cowan- Lecture 9

PEER REVIEW HISTORY ARTICLE DETAILS TITLE (PROVISIONAL)

THE RELATIONSHIP OF ADULT ATTACHMENT STYLE AND INTERACTIVE CONFLICT STYLES TO MARITAL SATISFACTION. A Dissertation ANNE KATHERINE CROWLEY

Original Article. Authors

Feedback Mechanisms and Self-Regulatory Processes in the Social Sciences

Attachment and the investment model: Predictors of relationship commitment, maintenance, and persistence

Answers to end of chapter questions

CURRENT RESEARCH IN SOCIAL PSYCHOLOGY

Examining the Role of Self-esteem in the Association between Emotional Vulnerability and Psychological Well-being

Students Perception and Attitude Toward Hiring a Former Offender

TITLE OF THE ARTICLE: ASSESSING THE RELATIONSHIP BETWEEN EMOTIONAL QUOTIENT AND FRIENDSHIP QUOTIENT AMONG YOUTH OF JAMMU CITY, J&K, INDIA

1. Before starting the second session, quickly examine total on short form BDI; note

TTI Success Insights Emotional Quotient Version

Attachment and styles of conflict resolution in close relationships.

Blind and Deaf to Acceptance: The Role of Self-Esteem in Capitalizing on Social Acceptance. Anna Maud Luerssen

Statistical Methods and Reasoning for the Clinical Sciences

Chapter 3 Perceiving Ourselves and Others in Organizations

Chapter 11 Nonexperimental Quantitative Research Steps in Nonexperimental Research

Notes Unit I Theories, Perspectives and Background Information

Transcription:

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/322959930 Sources of Perceived Responsiveness in Family Relationships Article in Journal of Family Psychology February 2017 DOI: 10.1037/fam0000411 CITATIONS 0 READS 4 3 authors, including: William Cook University of New England (USA) 55 PUBLICATIONS 5,159 CITATIONS Jan De Mol Université Catholique de Louvain 47 PUBLICATIONS 227 CITATIONS SEE PROFILE SEE PROFILE Some of the authors of this publication are also working on these related projects: Grief processes View project Parenting View project All content following this page was uploaded by William Cook on 06 February 2018. The user has requested enhancement of the downloaded file.

Running Head: PERCEIVED RESPONSIVENESS 1 Sources of Perceived Responsiveness in Family Relationships 2017, American Psychological Association. This paper is not the copy of record and may not exactly replicate the final, authoritative version of the article. Please do not copy or cite without authors permission. The final article will be available, upon publication, via its DOI: 10.1037/fam0000411 William L. Cook University of New England Marie Dezangré and Jan De Mol Université Catholique de Louvain-La-Neuve Author Note William L. Cook, Ph.D., Center for Excellence in Neuroscience, University of New England. Marie Dezangré, M.A., Université Catholique de Louvain-La-Neuve. Jan De Mol, Ph.D., Psychological Sciences Research Institute, Université Catholique de Louvain-La-Neuve. This research was supported by a grant from the U.S. National Science Foundation, Grant Number 9696083. This manuscript is based on the masters degree thesis of Marie Dezangré, completed at Université Catholique de Louvain-La-Neuve, Belgium. The Adolescent Responsiveness Scale from which we adapted the questionnaire items for this study was developed by M. S. Losoff as part of his doctoral dissertation at The University of Texas at Austin, 1989, Ted L. Huston, dissertation advisor. The authors thank Dr. Huston for his assistance obtaining a copy of the dissertation. This material has not been presented at any conferences, posted on a website, or published in any other source. Accepted for publication in Journal of Family Psychology. Correspondence concerning this article should be addressed to William L. Cook, Center for Excellence in Neuroscience, University of New England, 11 Hills Beach Road, Biddeford, Maine, 04005. Contact: BillCook.PhD@gmail.com. February 6, 2017

PERCEIVED RESPONSIVENESS 2 Abstract Perceived Responsiveness has become one of the most important constructs in the relationship sciences. It is central to the development of a secure attachment style, the experience of social support, an internal locus of control, and the sense of control in close relationships. Conversely, an unresponsive environment is associated with learned helplessness and depression. Viewed through the lens of the social relations model (SRM), Perceived Responsiveness in family relationships could have multiple sources: the perceiver; the target or partner; the perceivertarget relationship; and the family group. This study used the SRM to determine the relative importance of these sources of Perceived Responsiveness in the relationships of 207 two-parent two-child families. Characteristics of the perceiver and the target each accounted for about 25% of the systematic variance in Perceived Responsiveness, whereas the perceiver-target relationship accounted for approximately 48%. At the individual level of analysis, reciprocity of Perceived Responsiveness was pervasive in the family relationships of the two children. Regardless of age, young people who generally perceived others as responsive were generally perceived by others as responsive. At the dyadic level of analysis, reciprocity was present in two dyads: mother father and older child younger child. Reliable target variances support the view that Perceived Responsiveness is not just inside the head of the perceiver, and reciprocity correlations suggest potentially useful systemic interventions. Keywords: Responsiveness, Interpersonal Perception, Family Processes, Social Relations Model

PERCEIVED RESPONSIVENESS 3 Sources of Perceived Responsiveness in Family Relationships Over the last two decades, Perceived Responsiveness has become a core organizing principal for the study of interpersonal relationships (Reis, Clark, & Holmes, 2004). It has a high degree of conceptual and definitional overlap with perceived social support (Branje, S. J. T., van Acken, M. A. G., & van Lieshout, 2002); Lakey, McCabe, Fisicara, & Drew, 1996), it is a key contributor to positive marital communication and therapy (Gottman, 1979), and it is a causal factor in research and theory on intimacy (Debrot, Cook, Perrez, & Horn, 2012; Laurenceau, Barrett, & Pietromonaco, 1998), adult attachment (Collins & Read, 1994), perceived control in family relationships (Cook, 1993), and learned helplessness (Fincham & Cain, 1986). Perceived Responsiveness is the perception that a partner comprehends and values one s internal state and point of view, and will act supportively and empathically to one s desires, needs, goals, values, and preferences (Reis, Clark & Holmes, 2004). Research on Perceived Responsiveness has generally treated it as a single-dyad phenomenon. However, it is also possible to study it within groups such as families. This allows one to determine how Perceived Responsiveness varies across different partners and to investigate the sources of this variability. This study fills gaps in our knowledge of Perceived Responsiveness in relationships by using the family version of the social relations model (SRM: Kenny, Kashy, & Cook, 2006) to identify the relative importance of perceiver, target, relationship, and group characteristics as sources of Perceived Responsiveness in family relationships and to evaluate the extent to which Perceived Responsiveness is reciprocated. Early parent-child relationships are believed to play a central role in the way that people perceive responsiveness in adulthood. From birth, the infant produces reflex-like behaviors (such as crying and smiling) to which the caregiver responds. According to social learning theory (Rotter, 1966), the contingency between the infant s signal and the caregiver s response is believed to lead the child to experience a cause-effect relationship. Stable patterns in such

PERCEIVED RESPONSIVENESS 4 contingent or cause-effect relationships lead to generalized expectancies; that is, the expectation that the same contingencies will operate in future novel situations or relationships. Inconsistent patterns of contingent responsiveness have been proposed as the basis for learned helplessness (Fincham & Cain, 1986). Responsiveness is also a fundamental aspect of attachment theory because a sensitive mother is aware of her infant s cues, interprets them correctly, and responds promptly and appropriately (Smith & Pederson, 1988, p. 1097). The caregiver s responsiveness gives the infant a sense of control over its environment, thus instilling in the infant a sense of security and, consequently, a secure attachment to the responsive caregiver (Maccoby, 1980). The actual experience of the contingency depends on the responsiveness of the partner. In this regard, the original source of Perceived Responsiveness is the person whose responsiveness is being observed: a partner or target effect. This is a key feature in Reis s (2014) interpersonal model of responsiveness. When one partner demonstrates a willingness to set aside his or her own preferences and instead prioritize the other s needs and interests, an observer is led to the logical inference of caring and concern. These inferences must have some basis in fact. If motivated perception were fundamentally out of touch with social reality, it would not well serve people s goals and motivational purposes (p. 264). Several studies of adult relationships have demonstrated that the actual behavior of the partner is a reliable predictor of how the partner is perceived in adult relationships (Abbey, Andrews, & Halman, 1995; Collins & Feeney, 2000; Coriell & Cohen, 1996; Davis, 1982; Gable, Gonzaga & Strachman, 2006; Laurenceau, Barrett, & Pietromonaco,1998; Simpson, Ickes, & Blackstone, 1995), including studies specific to Perceived Responsiveness (Debrot et al., 2012). Although they include studies of Perceived Responsiveness in adult romantic relationships, prior studies have not investigated Perceived Responsiveness in key nuclear family relationships, such as adolescents perceptions of their parents and their siblings. Hypothesis 1 of

PERCEIVED RESPONSIVENESS 5 the current study is that characteristics of the partner will be a significant factor in Perceived Responsiveness in all nuclear family relationships. As mentioned above, when an infant experiences consistent patterns of contingent responsiveness from the caregiver, generalized expectancies that other people will behave similarly develop (Rotter, 1966; see also Bowlby, 1973). However, once the contingent responsiveness of the caregiver becomes internalized as a generalized expectancy, it has become partially detached from the behavior of any particular partner and influences how other people in general are perceived. Today, most theories of interpersonal perception, including attachment theory (Bowlby, 1973; Davis, 1982; Main, Kaplan, & Cassidy, 1985), propose that characteristics of the perceiver (e.g., their internal working models of relationships) serve as a filter through which the behavior of others is interpreted. Moreover, these filters or schemas can be as important as -- if not more important than target effects in determining how the partner is perceived. People may also attribute their own internalized characteristics to others when they interpret others behaviors (Andersen & Chen, 2002; Holmes, 1968; Lemay, Clark, & Feeney, 2007; Markus, Smith, & Moreland, 1985). For example, via a process of projection, individuals who perceive themselves as responsive to others have a tendency to perceive their partners as responsive (Lemay & Clark, 2008; Lemay, Clark, & Feeney, 2007; Canevello & Crocker, 2010; Debrot, Cook, Perrez, & Horn, 2012). Based on these considerations, Hypothesis 2 of the current study is that characteristics of the perceiver will be a significant factor in the perception of responsiveness in family relationships. Although Perceived Responsiveness may be a function of perceiver and partner characteristics, it can also be a function of the interaction of the perceiver s characteristics and the partner s characteristics, a relationship specific phenomenon (Clark & Mills, 1993; Collins & Feeney, 2000; Davis, 1982; Lakey, Ross, Butler & Bentley, 1996; Branje, S. J. T., van Acken,

PERCEIVED RESPONSIVENESS 6 M. A. G., & van Lieshout, 2002; Sameroff, 1975). Lemay and Clark (2008) have argued that Perceived Responsiveness is entirely a relationship specific phenomenon, a finding supported by their analysis of adult friendship groups. Cook (2000) found that about 40% of the systematic variance in attachment security in the family members relationships, which theoretically depends on responsiveness, was unique to the specific relationship between the family members. And in a meta-analysis of perceived social support, Lakey (2010; cited in Lakey & Orehek, 2011) found that 63% of the systematic variance in perceived support was relationship-specific. Consequently, we expect a relatively large amount of the variance in Perceived Responsiveness to be relationship-specific (Hypothesis 3). As mentioned earlier, characteristics of the partner play an important role in perceptions of the partner s responsiveness. But relationships are two-sided. Consequently, as one person s responsiveness affects the other person s responsiveness, a pattern of reciprocal responsiveness can develop which maintains and enhances each person s motivation to respond responsively in the future (Reis, 2014, p. 265). Trust develops when each partner feels confident that the other will take his or her best interests into account (Holmes & Rempel, 1989) in other words, that the partner s motives and behavior toward oneself are benevolent and caring. In turn, perceiving this benevolence motivates reciprocal benevolence and caring, setting off the chain-like sequence of mutual cyclical growth. Perceiving a lack of benevolence of course has the opposite effect. Based on the preceding, what is reciprocated is the benevolent behavior attributed to another person (whether actually perceived or the product of projection) that is foundational to the perception of responsiveness. Even though it is the behavior and not the perception that is reciprocated, we will refer to this phenomenon as reciprocity of Perceived Responsiveness to remain consistent with other literature (e.g., Kenny, 1994).

PERCEIVED RESPONSIVENESS 7 Reciprocity can occur at both the individual and dyadic levels of analysis (Kenny, Kashy, & Cook, 2006; Kenny & Nasby, 1980). For our purposes, it occurs at the individual level of analysis to the extent that a person who perceives other family members in general as responsive is generally perceived by other family members as responsive. Branje, van Acken, and van Lieshout (2002) found significant reciprocity of perceived support at the individual level of analysis for adolescent family members, but none for other family roles. Cook (2000) found individual level reciprocity of attachment security for mothers, fathers, older siblings, and younger siblings, when both measures of attachment security (Depend and Anxious) were taken into account. Based on these findings, Hypothesis 4a is that there will be reciprocity of Perceived Responsiveness at the individual level of analysis. There is reciprocity of Perceived Responsiveness at the dyadic level of analysis if a person whose perception of responsiveness is unique to a particular partner is uniquely perceived as responsive by that partner. Branje and colleagues (2002) found reciprocity of Perceived Support at the dyadic level in the marital dyad and the sibling dyad, but not in cross-generational dyads. With one exception (the mother older sibling dyad), Cook (2000) found the same pattern of dyadic reciprocity of attachment security. In the only study to perform an SRM analysis of a specific measure of Perceived Responsiveness (Lemay and Clark, 2008), a dyadlevel factor corresponding to an SRM dyadic reciprocity correlation was the primary source variance in Perceived Responsiveness. Hypothesis 4b is that there will be dyadic reciprocity of Perceived Responsiveness in family dyads consisting of members of the same generation (i.e., the marital and sibling dyads). There is very little theory about variables that might vary at the group or family level, and we are unaware of any hypotheses of group effects specific to Perceived Responsiveness. For family groups, however, there are empirical grounds for proposing family-level effects. In the aforementioned studies of Perceived Support and Attachment Security, small but statistically

PERCEIVED RESPONSIVENESS 8 significant variances in family effects were found (Branje, van Acken, & van Lieshout, 2002: Cook, 2000). In some families, the average person experiences more support and attachment security (viz., less relationship anxiety) than in other families. Based on these findings and the theoretical relatedness of Perceived Support, Attachment Security, and Perceived Responsiveness, Hypothesis 5 is that there will be variance at the family level for Perceived Responsiveness. The Social Relations Model These hypotheses will be tested using the social relations model (SRM: Kenny & LaVoie, 1984). The SRM uses data from a round-robin design, a design in which directed relationship measures are obtained for each person in the group in relationship to each of the other group members. These measures are directed relationship measures because person A s score in relation to person B is not the same as person B s score in relationship to person A. In this study, the measures of Perceived Responsiveness were obtained from family groups consisting of a mother, a father, an older child and a younger child. There are 12 directed relationships in a family of four people (i.e., each of four people are observed in relation to three partners). When the observations are family member s perceptions of the other group members, the SRM can partition each perception into multiple constituent components: perceiver effects; target effects; relationship effects; and group effects. As indicated earlier, Hypothesis 1 is that characteristics of the partner or target of Perceived Responsiveness will be a significant determinant of Perceived Responsiveness. This is important because it indicates that perception of responsiveness is not simply subjective; not simply in the eye of the beholder. The hypothesis can be tested by findings of significant variance in SRM target effects. Target effects are based on similarity in the way a person is perceived by multiple group members, independent of the group effect. This consistency in the perceptions of multiple perceivers implies that target effects are observed at the individual level

PERCEIVED RESPONSIVENESS 9 of analysis. There are four target variances in the family SRM, one for each family role (i.e., mother, father, older child and younger child). Thus, there are four opportunities for Hypothesis 1 to be supported. Hypothesis 2 is that characteristics of the perceiver will partially determine Perceived Responsiveness. Perceiver effects are important because they reflect psychological factors that determine Perceived Responsiveness; factors such as assumed similarity and projection. Perceiver effects are based on similarity in the person s perceptions of multiple partners, independent of the group effect. For example, if someone perceives their three partners as high in responsiveness, but the family as a whole is not perceived as highly responsive, the person s perceiver effect will be high. Because they reflect general characteristics of a person, like target effects, perceiver effects are observed at the individual level of analysis. Hypothesis 2 will be supported if there are significant between-family variances in the perceiver effects. There are four perceiver effects in a family of four, one for each family role. Each of these provides a test of Hypothesis 2. It is rare that a person perceives each of the other family members to be exactly same. If the perceiver did rate everyone as exactly the same, then the perceiver effect would completely explain how that perceiver differed from the average family member. It is also rare for everyone to perceive a particular family member in exactly the same way. If everyone in the family did rate a person in exactly the same way, then that person s target effect would completely explain how that person differed from the average family member. If the way a person is perceived is not explained by the family effect, the perceiver effect, or the target effect, then it is unique to a perceiver-target relationship. This is called a relationship effect. Relationship effects are important because they demonstrate that interpersonal behavior can be relationship-specific and not due to personality traits or other individual differences. They are observed at the dyadic level of analysis, not at the individual level of analysis. Hypothesis 3 will be supported by significant

PERCEIVED RESPONSIVENESS 10 variance in the relationship effects for Perceived Responsiveness. There is a relationship effect for each of the 12 directed relationships in a four-person family. Thus, Hypothesis 3 may be supported in some relationships but not others. People often confuse the SRM effects with the variances in these effects. The mean of each of the SRM effects is important in describing family roles; for example, if mothers are generally perceived as the most responsive person in the family, the mean target effect will be larger for mothers than for fathers or either of the children. The variances describe how much families differ around these mean effects. Variances are important in testing hypotheses about relationships. If there is no variance in a particular effect, that effect does not explain variance in the observed scores. So, if the father perceiver effect does not have significant variance, characteristics of fathers do not explain variance in how much they perceive their wives to be responsive. However, the means for the SRM effects can be used to show how a particular family differs from the average family in a sample; for example, as part of a family assessment (Cook & Kenny, 2004), or to show how the average family in one culture differs from the average family from another culture. In a round-robin design consisting of four family members, each person is the perceiver of three partners and the target of the perceptions of three perceivers. The correlation between how a person perceives other family members in general (i.e., their perceiver effect) and how that person is perceived by other family members in general (i.e., their target effect) measures reciprocity of perception at the individual level of analysis. Often referred to as a generalized reciprocity correlation, the correlation of a person s perceiver and target effects will be estimated for each family role. Generalized reciprocity of Perceived Responsiveness is important because it indicates that the person participates in the creation of their own social environment. For example, by making kind and supportive comments to others, he or she elicits kind and supportive from others. Hypothesis 4a will be tested for 4 separate roles (i.e., mother, father,

PERCEIVED RESPONSIVENESS 11 older child, and younger child). The SRM also provides measures of reciprocity that are unique to specific dyads. Dyadic reciprocity is measured by the correlation of the two relationship effects within a type of dyad, for example mother s unique perception of father correlated with father s unique perception of mother. Like generalized reciprocity, dyadic reciprocity indicates that the person is responsible for their own social environment, but in a more restricted, relationship-specific context. There are six dyads in a four-person family: mother-father; mother-older child; mother-younger child; father-older child; father-younger child; and older child-younger child. Hypothesis 4b, that Perceived Responsiveness will be reciprocal at the dyadic level of analysis, will be tested separately for each of these dyads. Based on studies of Social Support (Branje, van Aken, & van Lieshout, 2002) and Attachment Security (Cook, 2000), support for dyadic reciprocity will most likely be found in dyads in which both members are from the same generation (i.e., the marital dyad and the sibling dyad). The group effect captures the degree of similarity between family members; that is, the shared environment or culture of the group. Technically it is based on the average score of all the perceptions in the group, the family mean. For example, in some families everyone may perceive their partners to be higher in responsiveness than they do in other families. Given a sample of families, the SRM tests whether there is reliable variance in the family group effect. If the variance is significant, it means that in every relationship, perceptions of responsiveness are partly determined by differences between families as groups. Hypothesis 5 will be supported if there is reliable variance in the family group effect. Participants Method This study is part of a larger investigation of family relationships conducted between 1994 and 2000 (Cook, 2000, 2001). The sampling and data collection procedures have been described in the earlier publications. The present study utilized data obtained from 207 two-

PERCEIVED RESPONSIVENESS 12 parent two-child families (mother, father, older child and younger child). The age of the parents averaged 46 years (SD = 4.39) for mothers and 48 years (SD = 5.44) for fathers. The average age for the children was 19.66 (SD = 2.19) years for the older child (59% women and girls) and 16.08 years (SD = 2.16) for the younger child (48% women and girls). Most of the parents (95%) were married. Fifty-three percent of the mothers and 69% of the fathers had at least a college level education. The average family income (based on the mother s report) was between $81,453 and $97,744 inflation adjusted U.S. dollars (United States, Department of Labor, Bureau of Labor Statistics, 2016). Design and Procedures In order to perform the SRM analysis, data were collected according to a round-robin family design. In a round-robin family design, observations are obtained on each family member in relation to each of the other participating family members. In a four-person family this produces 12 relationship-specific obervations: four individuals rating their relationship with each of the other three family members. Each family member received a packet of questionnaires along with a postage-paid return addressed envelope. The two children were paid $15.00 (US) each if all family members completed their questionnaires. Each family member signed an informed consent form. The study was approved by the IRBs of the University of Texas at Austin and the Maine Medical Center in Portland, Maine. Measures The observations used in the SRM analysis were ratings of the other family members responsiveness toward the respondent from a version of Losoff s (1989) Adolescent Responsiveness Scale that we adapted for use in family relationships. We will refer to this scale as Losoff s Responsiveness Scale. In evaluating the face validity of the scale items, we used the following definition: Perceived Responsiveness is the perception that a partner comprehends and values one s internal state and point of view, and will act supportively and empathically to one s

PERCEIVED RESPONSIVENESS 13 desires, needs, goals, values, and preferences (Reis, Clark & Holmes, 2004). The items align quite well with the three foci of this definition: A. Comprehends one s internal state and point of view 1. knows what makes you feel happy or sad 2. knows what s on your mind 6. knows your favorite activities B. Values one s internal state and point of view 5. tries to figure out what you are concerned about 7. tries to pick up how you are feeling 8. acts interested in what you have to say C. Will act supportively and empathically to one s desires, needs, goals, values, and preferences 3. changes what he/she says and does to fit your moods 4. does the kinds of things you like to do In the general instructions preceding the packet of questionnaires, participants were informed that they would be rating their relationship with other family members and that the identity of the person would be indicated at the beginning of each section. Whenever you see, please think of that person. Then, for example, at the top of the section of Perceived Responsiveness items, the instructions were, Now think about the relationship between and YOU. CIRCLE the number corresponding to the most accurate answer. Project staff wrote the name of the target person in the blank line at the beginning of each section, so it was very clear who the target was for each set of items. Because the items were originally developed to measure responsiveness in the friendships of adolescents, exploratory factor analyses were performed to test the unidimensionality of the 8- item scale in each of the 12 family relationships. An oblique rotation was used so the correlation

PERCEIVED RESPONSIVENESS 14 between factors would be measured if a two-dimensional structure emerged. A unidimensional structure was observed in seven of the 12 relationships, but not in the mother-older child, motheryounger child, father-older child, father-younger child, and younger child-older child relationships. The manner in which the items were associated with the two factors can be found online in Supplemental Table 1. Correlations between the factors ranged from.41 to.59. Given the expected unidimensional structure of the scale and the substantial correlations of the factors when two dimensions emerged, use of all eight items in all 12 relationships seemed justified. The coefficient alpha reliabilities were also adequate, ranging from.77 to.93, mean =.85 for the 12 measures. All but one of the coefficient alphas was greater than.80. The reliabilities for each family relationship are presented online in Supplemental Table 2. We also tested the validity of the ARS when used in family relationships (see online Supplemental Table 2). We tested correlations between Responsiveness and two dimensions of relationship specific Attachment Security (see Cook, 2000): (1) Comfort depending on the target (Depend) and (2) anxiety about being avoided or rejected by the target (Anxious). The responsiveness of the target should be positively associated with the perceiver s comfort depending on the target and negatively correlated with the perceiver s anxiety about the relationship with the target. The correlations were as expected in all 12 family relationships, averaging r =.539 for Depend and r = -.412 for Anxious. We also tested correlations of Perceived Responsiveness with two dimensions of Perceived Control, Effectance and Acquiescence (see Cook, 1993). Effectance, which measures a sense of being able to influence the target person, should be positively correlated with the target s responsiveness. This expectation was supported in all 12 family relationships (average r =.350). Acquiescence measures the perceiver s sense of being influenced by the target, and it was not expected to be correlated highly with the target s responsiveness. Although there were statistically significant correlations in six of the 12 family relationships, the correlations were small (M =.151). Consistent with research on child compliance

PERCEIVED RESPONSIVENESS 15 (Lay, Waters & Park, 1989; Parpal & Maccoby, 1985), there appears to be a tendency to be more acquiescent with family members who are perceived to be responsive. Analysis The specifics on how to perform a family SRM have been described elsewhere (Kenny, Kashy & Cook, 2006, Chapter 9) and will not be reiterated here. Although typically analyzed using structural equation modeling, the model can also be estimated using multilevel modeling (Rasbash, Jenkins, O Conner et al., 2011). The present analysis was conducted using fsrm (Stas, Schönbrodt, & Loeys, 2015), an R package (R Development Core Team) for the family SRM. The fsrm package facilitates the use of the family SRM by making automatic the complex structural equation modeling specifications for (1) the latent variables which estimate the variances of the perceiver, target, relationship, and family/group components of the SRM, (2) the reciprocity correlations at the individual and dyadic levels of analysis, and (3) the latent variable means for the SRM components. It also provides a variance partitioning of the observed scores for each of the relationships in the family; that is, the percentage of variance in each relational measure that is explained by the family, perceiver, target, relationship, and error components, respectively. The fsrm can be used with 3-person or 4-person families, and it accepts data that use either a single measure of each family relationship or multiple measures of each family relationship. The use of multiple measures (usually 2) of each relationship is needed to separate the variance of relationship effects from residual or error variance (Kenny, Kashy, & Cook, 2006). Items from Losoff s Responsiveness Scale were partitioned into two 4-item parcels for this purpose. Since both parcels measure the same construct, an arbitrary assignment of items to parcels is acceptable. In this study items were assigned in a manner that minimized the difference in the means of the two parcels. This prevents problems of model fit due to chance factors arising from other methods of arbitrary item-to-parcel assignment rules (e.g., assignment of even

PERCEIVED RESPONSIVENESS 16 numbered items to one parcel and odd numbered items to the other). The fsrm provides three widely accepted measures of the adequacy of the family SRM analysis: the chi-square value for goodness of fit: the RMSEA (Root Mean-Square Error of Approximation): and the Comparative Fit Index (CFI). Missing Data Missing data is a particularly difficult problem in the context of a family round-robin SRM study. There may be data missing at the item level (e.g., one of the 6 items in mother s report of father s responsiveness toward her); there may be data missing on all the items for one of the 12 relationships in the family (e.g., the younger child provides no ratings of responsiveness from the father); all the data may be missing for one of the four family members (e.g., if the older child did not return the questionnaires); and all the data from a family that consented to participate may be missing. It is unclear whether it would be appropriate to use either multiple imputation or full information maximum likelihood (FIML) methods for any level of missing data above the item level. We excluded families if all the data (perceptions of three targets) from one individual were missing (e.g., father with mother, father with older child, and father with younger child) and we used FIML to adjust for missing data at the relationship level (e.g., a family missing only mother s rating in relation to father). Earlier publications from this project indicate that families with missing data tend to function more poorly than those from which complete data was obtained. Consequently, the results of this analysis probably reflect patterns characteristic of relatively high functioning middle-class Caucasian families. Results The family SRM provided a good fit with the data. Although the chi-square goodness of fit test was statistically significant, χ2 (df = 125, N = 207) = 189.553, p <.001, the CFI was 0.975 and the RMSEA was 0.05, both well within the acceptable range. This indicates the

PERCEIVED RESPONSIVENESS 17 appropriateness of the SRM in accounting for perceptions of responsiveness in family relationships. SRM Factor Means and Variances for Responsiveness The SRM factor means and variances are presented in Table 1. The factor means describe the nature of Perceived Responsiveness in the relationships of the average family, or at least the average family in this sample. Given questionnaire response options ranging from 1 (never true) to 5 (always true), the family mean effect of 3.444 suggests that overall, family members generally perceived each other as moderately responsive. The variances indicate whether an SRM factor is a reliable predictor of the Perceived Responsiveness scores (i.e., the observed scores) of which it is a component. The family factor is a component in all 12 family relationships, but its between-family variance was not significant. As can be seen in Table 2, only about 2% of the total variance (2.6% of the systematic variance) in family members perceptions of each other s responsiveness was due to characteristics of the family as a group. The means for the perceiver factors describe the degree to which individual family members perceptions of the responsiveness differ from the mean of the family factor. Of the four perceiver effects, two differed significantly from the family mean effect. The typical mother experienced other family members as more responsive than the average family member, (M =.104), and the typical younger child experienced other family members as less responsive than the average family member (M = -.121). On the other hand, the variances of all four perceiver effects were significant. This means that regardless of the person s role in the family, their perception regarding the responsiveness of the other family members is partially determined by their own characteristics (e.g., a generalized expectancy). Some individuals perceive other family members as more responsive than do other individuals. As can be seen in the bottom row of Table 2, perceiver characteristics account for approximately 19% of the total variance and 25.7% of the systematic variance in Perceived Responsiveness in family relationships.

PERCEIVED RESPONSIVENESS 18 The means for the target factors measure the degree to which the average person in each of the four roles (mother, father, etc.) is perceived as responsive, relative to the family mean. As can be seen in Table 1, mothers were, on average, perceived as significantly more responsive than the average member of the family (M =.375), whereas fathers (M = -.077), the older child (M = -.088), and the younger child (M = -.210) were perceived as significantly less responsive than the average member of their family. The variances of all four target factors were also statistically significant. Within each of the four roles, there were between-family differences in how much the occupant of the role was experienced as responsive. Perceived responsiveness was significantly determined by the characteristics of the target, and not just the perceiver s tendency to see other family members as responsive (i.e., perceiver effects). Based on a target effect variance of.126, fathers varied about twice as much as other family members in how responsive they were perceived to be. As can be seen in the bottom row of Table 2, averaged across roles and relationships, characteristics of the partner or target explained 16% of the total variance in Perceived Responsiveness (or 23% of the systematic variance), almost as much as was explained by perceiver effects. The means for the relationship factors describe the degree to which family members perceptions of responsiveness reflect the unique relationship of the perceiver to the target. Whereas a perceiver effect indicates how much a perceiver sees others in general as responsive, relationship effects can be thought of as deviations from this general perceptual style. These deviations are estimated while controlling for characteristics of the family, the perceiver, and the target. For mothers, only the relationship with the younger child deviated significantly from their perception of other family members in general. They experienced the younger child as uniquely less responsive (M = -.047). Fathers, in contrast, experienced the mother as uniquely less responsive (M = -.059). For the older and younger children, the relationship effects did not differ from their general tendency to see other family members as responsive. Although few of the

PERCEIVED RESPONSIVENESS 19 mean relationship effects were statistically different from zero, the variances of all the relationship factors were statistically significant, independent of variance due to errors of measurement. This means that in all 12 types of nuclear family relationships, the unique relationship of the perceiver to the target influenced the observed score for Perceived Responsiveness. Table 2 shows that on average 37% of the total variance in Perceived Responsiveness was due to the unique perceiver-target relationship (48.2% of the systematic variance), about twice as much as either the perceiver or target effects explained. Reciprocity of Perceived Responsiveness Four statistically significant correlations measuring reciprocity of Perceived Responsiveness were found. At the individual level of analysis, there was substantial reciprocity of Perceived Responsiveness for both the older child (r =.41, p <.05) and the younger child (r =.41, p <.05). The more one of the children views other family members as responsive, the more that child is viewed by other family members as responsive. That the correlations were the same is coincidental. At the dyadic level of analysis, there were two large and significant reciprocity correlations. There was reciprocal Perceived Responsiveness in the mother-father dyads (r =.538, p <.001), the older-child younger-child dyads (r =.733, p <.001). Thus, if a person is responsive toward a family member from one s own generation, that family member will likely be responsive in return. Discussion Target Effects for Perceived Responsiveness Hypothesis 1 addressed the question Do characteristics of the target (or partner) influence perceptions of that target s responsiveness? If characteristics of the target influence perceptions of responsiveness, it is reasonable to entertain the proposition that the perceptions are in some sense objective or reality based. This is important because so much theory today proposes that interpersonal perceptions are in the head of the perceiver (e.g., schemas or

PERCEIVED RESPONSIVENESS 20 internal working models). But in attachment theory (Bowlby, 1973), it is not the imagined responsiveness of the caregiver that produces a securely attached infant, it is a caregiver who is actually responsive; one who successfully meets the infant s needs (e.g., Bogenschneider & Pallock, 2008). Likewise, some grounding in reality should anchor perceptions of responsiveness in the relationships of adolescents and adults, especially in their family relationships. Significant variance in the target effects for Perceived Responsiveness in families supports Hypothesis 1. In fact, there was statistically significant variance in the target effects of each role; mother, father, older child, and younger child. As indicated in Table 2, characteristics of the target accounted for about 16% of the total variance and 23% of the systematic variance in the Perceived Responsiveness. This is much more than the partner variance for Perceived Support in family relationships (Branje, van Aken, van Lieshout, 2002), which we calculated to account for about 5% of the systematic variance. Our findings are closer to the amount of systematic variance in perceptions of the Big Five personality traits accounted for by the target effects of long-acquainted others, averaging about 28% (Kenny, Albright, Malloy, & Kashy, 1994). Other studies have found that enacted responsiveness predicts Perceived Responsiveness in adult romantic relationships (Debrot et al., 2012), but the present study is the first to show that it occurs in parent-child and sibling relationships as well. Interestingly, ranking the roles by the size of their target effects (i.e., the factor means) produces what you might expect given stereotypes of the amount of responsibility for the care of other family members traditionally accepted by people in these roles. Mothers were perceived as most responsive, fathers were next, followed by the older child. Younger children, typically adolescents, were perceived as least responsive. The position of mothers at the top of this hierarchy of care was not remotely challenged by other family members. It is perhaps encouraging that fathers were at least second. Perceiver Effects for Perceived Responsiveness

PERCEIVED RESPONSIVENESS 21 Hypothesis 2 addressed the question, Do the characteristics of the perceiver influence the perception of responsiveness in other family members? Social Learning Theory (Rotter, 1966) and Attachment Theory (Bowlby, 1973) both predict that the early experience of a responsive caregiver should lead to a general tendency to expect other people to be responsive, or if the caregiver was not responsive, a corresponding expectation of unresponsiveness from others. The generalized expectation of responsiveness from other people independent of their actual responsiveness -- is a perceiver effect. The projection of one s own level of responsiveness onto others (Lemay, Clark, & Feeney, 2007), or assumed similarity (Kenny, 1994), also constitutes a perceiver effect. The present findings add to prior evidence specific to adult romantic relationships (Debrot et al., 2012; Lemay, Clark, & Feeney, 2007) that characteristics of the perceiver reliably predict Perceived Responsiveness, supporting Hypothesis 2. However, this study extends the finding to all nuclear family relationships. Moreover, it does so using a method that explicitly controls for the effect of all target characteristics, not just those selected for a particular investigation. Perceiver effects accounted for about 19% of the total variance and 25.7% of the systematic variance in Perceived Responsiveness. This is only slightly more than the amount that was explained by target effects, but it is considerably less than 44.5% perceiver variance we calculated for Perceived Support based on the Branje, van Aken, and van Lieshout (2002) results. Interestingly, the mean perceiver effects for mothers, fathers, and the older children in our sample did not differ from the family mean. On average, however, the younger child perceived less responsiveness than did the average family member. Relationship Effects for Perceived Responsiveness Hypothesis 3 addressed the question, Is the perception of responsiveness in families unique to the perceiver-target relationship? In the current study about 48% of the systematic variance in perceived responsiveness in family relationships was unique to the specific relationship, supporting Hypothesis 3, independent of the general characteristics of the perceiver,

PERCEIVED RESPONSIVENESS 22 characteristics of the target, and errors of measurement. This approximates the amount of relationship variance estimated by Lakey (2010) in his meta-analysis of perceived social support (63%) and by Cook s (2000) study of attachment security in families (40%). It is over twice the amount of variance accounted for by the more general dispositions of the perceivers and targets. Consequently, even though perceiver and target characteristics can explain non-trivial amounts of the variance, theoretical considerations of the causes of Perceived Responsiveness have more to gain by a focus on relationship specific factors. Two of the factor means for relationship effects are also of interest. Fathers experience less responsiveness from their wives than they do from their average family relationship, and mothers experience less responsiveness from their younger child than they do from their other family relationships. The slightly negative father mother relationship effect probably does not constitute a serious difficulty in the typical marital relationship, given that it is added to the large positive target effect of the typical mother. Mothers are highly responsive to all family members, and only slightly less so toward their husbands. On the other hand, the target effect of the younger child is the lowest target effect in the average family. When the significantly negative mother-child relationship effect is added to the child target effect, the additive effect is even more negative. This highlights the relative difficulty of maternal relationships with adolescent children in the typical family. From a clinical perspective, the mother s experience of the younger child s responsiveness would have to be very low indeed to differ significantly from what is seen in a typical family. Reciprocity of Perceived Responsiveness For readers interested in the application of systems theory to the study of families, the measurement of reciprocity may be the most interesting feature of this study. Of key concern in systems theory (Bowen, 1978; Bowlby, 1949) is the self-organizing capacity of systems, the production of reliable patterns of interpersonal behavior determined by processes of positive and

PERCEIVED RESPONSIVENESS 23 negative feedback. The SRM reciprocity correlations measure the strength of positive and negative feedback loops. Positive feedback, suggesting that one gets what one gives is measured by positive reciprocity correlations at the individual or dyadic levels of analysis. It is associated with deviation amplifying processes (Hoffman, 1971). Negative feedback, indicating that the more one gives the less one gets is measured by negative reciprocity correlations. Negative feedback is associated with deviation dampening processes (Hoffman, 1971). One will observe negative feedback loops in an interpersonal conflict if one person acts to de-escalate the conflict when the other escalates. But in some theories of family systems (e.g., Bowen, 1978), the attempt of one family member to change (e.g., due to therapeutic progress) is countered by resistance from within the family, which would also constitute a negative feedback loop. In this study, there were no negative reciprocity correlations for Perceived Responsiveness, but there were reliable positive reciprocity correlations. At the individual level of analysis, the responsiveness experienced by both the younger child and the older child was significantly and positively correlated with the amount responsiveness these children were perceived to give to others (i.e., target effects). So why is there generalized reciprocity for the children but not the adults in the family? It may help to present the finding in somewhat different language. Said differently, children who provide extra caring and support to other family members receive extra caring and support from other family members. Parents, on the other hand, do not receive extra caring and support from other family members by giving extra. Why? Possibly because both the parents and the children expect the parents to provide caring and support. It is part of the job description for parenting. But neither the parents nor the children expect the children to reciprocate the parents caring and support; that is not part of a child s job description, at least not until much later in their development. One could say that in relationships with dependent others, such as children, one should not expect the dependent to reciprocate responsiveness. In relationships among equals, however, the norm of reciprocity