RETURNING MATERIALS:
MSU Place in book drop to remove this

checkout from your record. FINES
”BRAKES will be charged if book is returned
2—- after the date stamped below.

 

 

 

 

 

SOCIAL INTERACTION: SEX DIFFERENCES IN THE EXCHANGE

OF EMOTIONAL SUPPORT

By
Linda A. SuIIivan

A DISSERTATION
Submitted to

Michigan State University

in partial fulfiIIment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Psychology
1989

5L7“ 5'7 25]

ABSTRACT
SOCIAL INTERACTION: SEX DIFFERENCES IN THE EXCHANGE
OF EMOTIONAL SUPPORT

By

Linda A. Sullivan

A study was conducted to examine the exchange of emotional support in
social interaction with a specific focus on sex differences in support
and reciprocity of support. Previous research suggested that females
would receive and provide more emotional support than would males, and
that the exchange of support would be characterized by reciprocity.
Subjects completed a modified version of the Rochester Interaction
Record (RIR) for a period of two weeks on which they described their
social interactions with specific emphasis on social support. They also
completed a mood checklist daily and completed some postrecord—keeping
personality measures. Results supported the hypotheses that females
would need, provide, receive, and perceive a greater need in other’s for
support than would males. It was concluded that females greater level
of negative affect and greater communal orientation in comparison with
males contributed to these results. Results also supported the
hypothesis that support exchange would be characterized by reciprocity.

However, both sexes perceived that they received more support from

persons of the opposite-sex than they provided to them. These results
were interpreted within the empirical framework of Clark and Mills’
(1979) distinction between communal and exchange relationships. Mood
analyses showed positive relationships between supportive behaviors and
both negative and positive affect. It was concluded that the emotional
content of both these indices created the positive relationship between
them. Finally, personality measures (empathy and social skills) were
less effective than sex in predicting differences in reported supportive

behaviors and perceptions.

Copyright by
Linda Ann Sullivan

1989

This dissertation is dedicated to six women:
Ann, Joan, Joyce, LAJ, Riia, and Sally,

for never breaking an appointment with me.

ACKNOWLEDGEMENTS

When you are a Linda you are never alone! I am most grateful to my
chairwoman, LAJ, for her guidance and direction through the years. Her
appropriate sense of mellowness no doubt has increased my life
expectancy. I could not have hoped for a better mentor and role-model.
Thanks are due to my committee members, Lawrence Messe, Ann Baumgartner,
Robert Emmons, and Josefina Fiquiras-McDonough, for their constructive
criticism and encouragement. I am particularly grateful to Bob Emmons
for taking the time to review this work in absentia. I am indebted to
Larry Hesse for his cars, parking stickers and office space, but of
course, also for his scholarly direction throughout my graduate career.
My ”ghost committee member" John Nezlek is due thanks for what I
sincerely hope is obvious.

The people who are the gate keepers of every graduate student’s
sanity tend to be those in the same boat. Thank you Rick, Jan, and
Dave, for your sarcasm and cynicism!! Thank you again Rick for all your
*&*%$@#. For their help from near and afar, from old and new, I am
grateful to Sally, Joyce, Riia, Laurent, and Diane: teachers of
reciprocity and emotion. Diane, I’ll never believe that it could have
been done without you. Thank you for being there when help was most
needed. I am deeply indebted to my family for far too much to mention,

but I must say thank you for the great cheer leading. I was very lucky

vi

to have had a great group of research assistants, Angela Armstrong,
Holly Connor, Lauren Epstein, Marcelle Kendall, Shelly Thibert (Miller),
Karen Mizak, Curtis Myers, Alicia Panamo, Garry Randall, Michelle
Ursulowicz, and Karen Hillins. Hay to Go Team! Thanks also to Phil
Gardner without whom this would never have been printed on time. And
last, but definitely not least, I thank my subjects/coinvestigators for

what you are about to read.

vii

TABLE OF CONTENTS
LIST OF TABLES
CHAPTER I: INTRODUCTION
Social Interaction: Definition and Measurement
Social Support: Definition
Social Support: Measurement
Sex Differences in Social Interaction and Social Support
Reciprocity in Social Interaction and Social Support
Hypotheses
Social Support, Personality, and Positive and Negative Affect
CHAPTER II: METHOD
Subjects
Materials
Procedure
CHAPTER III: RESULTS
Construction of Variables
Reliability of the RIR
Quality of Social Interaction: The Research Hypotheses
Receipt of Support
Provision of Support
Need for Support
Perception of Others’s Need for Support

viii

XI

11
15
24
33
36
40
4o
44
45
45
45
46
43
48
so
51
53

Support Provision and Perceived Need for Support
Provision of Support and Difficulty of Support Provision

Relationship between Support Received and Support
Provision, Intimacy, and Quality of Interaction

Reciprocity of Support
Quality of Social Interactions: Additional Analyses
Difficulty of Provision of Support
Intimacy of Interactions
Quality of Interactions
Support Needed and Support Received
Support Needed and Support Other Needed
Support Provided and Support Other Needed
Relationship Between Variability of Support and Personality
Personality Variables as Mediators of Support Measures
Relationships between Support Indices and Mood
CHAPTER IV: DISCUSSION
Sex Differences in Social Support
Reciprocity of Social Support
Relationships between Supportive Behaviors
Intimacy, Quality, and Support
Personality and Support
Affect and Support
Conclusions
REFERENCES
APPENDIX A: Rochester Interaction Record
APPENDIX 8: Personal Experience Form
APPENDIX C: Instruction Booklet

ix

54
56

57
59
60
6O
61
62
63
65
65

67
71
77
78
80
84
86
87
88
91
94
105
106
108

APPENDIX D:
APPENDIX E:
APPENDIX F:
APPENDIX G:
FOOTNOTES

Verbatim Script
Information to Participants
Post-experimental Interview

Nomenclature of RIRAP Variables

114
116
118
119
122

10.
ll.
12.

13.

14.

15.

LIST OF TABLES

Comparative Quantitative Indices of Interaction Pattern ......... 47

Mean Number of Interactions for Recording Period and
Means and Standard Deviations for Receipt of

Support as a Function of Interaction Category ................. 49
Means and Standard Deviations for Provision of Support

as a Function of Interaction Category ......................... 51
Means and Standard Deviations for Need of Support as a

Function of Interaction Category .............................. 52
Means and Standard Deviations for Perception of Other’s

Need for Support as a Function of Interaction Category ........ 54
Relationship between Support Provision and Support

Other Needed .................................................. 55
Relationship between Support Provided and Difficulty ............ 56
Relationship between Support Received and Intimacy .............. 57
Relationship between Support Provided and Intimacy .............. 57
Relationship between Quality and Support Received ............... 58
Relationship between Quality and Support Provided ............... 59
Relationship between Support Provided and Support

Received ...................................................... 59
Means and Standard Deviations for Intimacy of

Interaction as a Function of Interaction Category ............. 62
Means and Standard Deviations for Quality of Interaction

as a Function of Interaction Category ......................... 63
Relationship between Support Received and Needed ................ 64

xi

16.

17.

18.

19.

20.
21.

22.

23.
24.
25.

Relationship between Support Needed and Support

Other Needed .................................................. 65
Relationship between Support Provided and Support

Other Needed .................................................. 65
Relationship between Variability of Support and

Personality for Females ....................................... 66
Relationship between Variability of Support and

Personality for Males ......................................... 67
Multiple Regression Analyses to Predict Support Measures ........ 68
Multiple Regression Analyses to Predict Support Measures

for Females ................................................... 69
Multiple Regression Analyses to Predict Support Measures

for Males ..................................................... 70
Correlations between Moods and Overall Support Measures ......... 74
Correlations between Moods and Same-Sex Support Measures ........ 75

Correlations between Moods and Opposite-Sex Support
Measures ...................................................... 76

xii

INTRODUCTION

Historically the study of human relationships and social
interaction has not been considered a topic either worthy of, or
appropriate for, scientific investigation (Berscheid & Peplau, 1983).
Recently this viewpoint has been changing as evidenced by the
publication of journals specifically devoted to social relationships
(6-9. W) and by the
development and continuation of conferences focused on interpersonal
relationships (e.g. The Iowa Conference on Personal Relationships).

Recent interest in social interaction as a focus of scientific
inquiry has no doubt been encouraged by the relevance of social
interaction to other social phenomena. Clearly, descriptive analyses of
an individual’s pattern of social interactions are of interest to social
psychologists, while the characteristics of social networks are of
interest to sociologists and ecological psychologists. However, social
interaction assumes a social significance beyond mere academic interest
when one considers its relevance to aspects of an individual’s well-
being or accomplishments. For example, research on loneliness (Berg &
Peplau, 1982; Wheeler, Reis, & Nezlek, 1983) has shown that both the
quantity and quality of social interaction, as well as specific

dimensions of social interaction, are related to feelings of loneliness,

2
a state that affects about one quarter of the population (Peplau,
Russell, & Heim, 1979).

Social interaction has also been found to affect performance,
particularly academic performance. For instance the quality and
quantity of males’ social interaction is negatively related to their
grade point average (Nezlek, Wheeler, & Reis, 1983B; Sullivan, 1986).

At the graduate level, Hall (1969) found that amount of peer interaction
was positively related to academic achievement, specifically on doctoral
comprehensive examinations. Further, Kowalski (1982), after reviewing
the reasons behind the 40% withdrawal rate at American colleges,
concluded that "a positive personal relationship" (p. 47) with a faculty
member can greatly increase the chances of students persisting in
college. Thus social interaction appears to have important effects on
academic performance.

Social interaction has also been related to health. Reis, Hheeler,
Kernis, Spiegel, and Nezlek (1985) found that the quality of females’
social interactions was negatively related to seeking health care.
However, the major component of social interaction relating to health is
social support. Reis et al. (1985) did not specifically measure social
support, so it is not clear to what extent the measure of social
interaction was tapping this phenomenon. Much research has shown that
social support is positively related to health and well-being (see Cohen
& Syme, 1985, for a review). Indeed, the vast majority of social
support research has focused on its relationship to well-being.

Surprisingly, the social support research has not examined support
as a component of social interaction, primarily because ”there is a

tendency among some researchers to equate social interaction with social

3
support” (Rook, 1984, p. 1097). Recently, some authors (e.g., Reis,
1984; Shinn, Lehmann & Wong, 1984) have argued that it is more
appropriate to think of social support as a component of social
interaction (e.g. Reis, 1984). Social interaction may take place with
social support being irrelevant, but social support cannot take place
without social interaction (House, 1981). In the view of the present
author, social interaction should not be viewed as social support, but
rather as the mechanism by which social support can be expressed. To
consider social interaction and social support as identical processes is
to confuse the figure with the ground. Social support is best
conceptualized as a figure against the ground of social interaction.

It would appear then that any investigation of social support in
the context of naturally occurring social interaction should increase
our understanding of the underlying components of social interaction.
Such an investigation should also increase our knowledge of the nature
of support.

The purpose of the present research was to investigate sex
differences in social interaction with the focus being on the exchange
of support. The integration of support with social interaction makes
imperative the investigation of sex differences and exchange. As Reis
(1986) noted, sex is "a fundamental factor that differentiates numerous
aspects of social life" (p. 105). Further, "interaction between persons
is an exchange of goods, material and nonmaterial' (Homans, 1958, p.
597). Thus, by placing support in the context of social interaction,
one must attend to sex differences and the exchange of support, both of
which have been neglected in the research conducted on support to date.

The following sections outline the operationalizations and measurement

4
of social interaction and social support in previous research and in the
present research.

i r : ' i ' M m

Social interaction is considered by most to be a multidimensional
activity whose focus is centered on the behavior of others. For
example, Wheeler and Nezlek (I977) define the activity as any face-to-
face encounter in which person A modifies her or his behavior in
response to person B and vice versa. Social interaction involves
components such as control, quantity, quality, intimacy, satisfaction,
reciprocity, and support (Reis, I984; Wheeler and Nezlek, 1977). In an
attempt to classify social interaction according to characteristics that
would lend themselves to empirical quantification, Forgas (1976) had
subjects record all their interactions for the previous day. The 25
most often mentioned episodes were then presented to other subjects who
placed them into categories according to their similarity. A measure of
relatedness between each pair of stimuli was then calculated and
submitted to multidimensional scaling. Forgas (1976) found that for
students, the main attributes, or dimensions, differentiating social
interactions were intimacy, pleasantness, and self-confidence i.e.,
knowing what to do in a social situation.

In general, social interaction is not a phenomenon that has been
defined by researchers in their study of the topic. This problem has
also been observed to occur in the friendship literature by Winstead and
Derlega (1986). As they point out, the concept of friendship has rarely
been defined in friendship research. One is left to infer the
operational definitions of social interaction from the various measures

used; and these measures have varied widely. Some have used interviews

5
to assess subjects’ supportive and problematic social ties (e.g., Rook,
1984). Others assume that variables such as Greek organization
membership are an index of social activity, considering members to be
"socially prominent" (e.g., Sanford, 1962, p. 150).

Measures of interaction have been more precise in the laboratory.
For example, the conversations of groups have been content-analyzed to
measure themes of interaction (Aries, 1976). More detailed analyses of
group conversations have been made possible with Bales (1950, 1970)
1ntgLactign_£rggg§§_Analysis, and naturalistic studies of interactions
have been made possible with Ickes’ unstrggtgrgg_lntg§agtign_£ar§digm
(See Ickes, 1983, for a discussion of this paradigm). The laboratory,
however, is a rather limited context for studies of social interaction,
because these situations are either highly structured or are confined to
relatively short periods of time. So too are questionnaire and
interview studies limited in their measure of social interaction,
because they are prone to memory limitations, memory distortion, and
potential social desirability effects.

Clearly a naturalistic approach to the study of social interaction
that is not subject to the above shortcomings is most appropriate. One
such approach involves the maintenance of diaries or daily records of
social interaction. Such an approach has been in use for many years
(e.g., Bolton & Kammeyer, 1967); however, one particular diary method,
the Rochester Interaction Record (RIR; Wheeler & Nezlek, 1977), recently
has become popular among researchers interested in collecting data on
person’s social experiences.

The RIR is a standard form that requires subjects to record the

duration of a given interaction, the sex of the participants, the

6
context of the interaction, as well as seven affective responses to the
interaction. Thus the RIR measures the quantity, structure, and
subjective dimensions of the diary keeper’s interaction pattern. Nezlek
and his colleagues (Nezlek, Wheeler, & Reis, 1983A) maintain that this
measure has advantages over other methods. First, the RIR taps a wide
variety of social interactions. Second, it reduces memory limitations
and memory distortions because the diary is maintained daily. Each
recording represents a single distinct social interaction thus
alleviating the need for a subject to attempt to make global responses.
Finally, the measure is flexible insofar as other researchers can
redefine the subjective scales for their own purposes. Subsequent
research has demonstrated that use of the RIR, or variants thereof, has
become popular in investigations of a number of variables associated
with social interaction, including social support (Cutrona, 1986),
health (Reis, Wheeler, Kernis, Spiegel, & Nezlek, 1985), intimacy
(Reis, Senchak, & Solomon, 1985), loneliness (Wheeler, Reis, & Nezlek,
1983), and physical attractiveness (Reis, Wheeler, Spiegel, Kernis,
Nezlek, & Perri, 1982).

For the purposes of this study, social interaction is defined as
behavior that involves the responding of one individual to another
(Wheeler & Nezlek, 1977). The clearest example of interaction would be
a conversation in which one person responds to the other. In contrast,
sitting side-by-side watching television is not an interaction. A
variant of the RIR, to be described in detail later, was used to measure

social interaction.

Soojal Soooort: Dofiojtioo

Delineating the components, or dimensions, of social support has
proved to be a major issue for researchers. The original definition
claimed support to be information that led the individual to believe
that she or he was cared for and loved and held in esteem and value, and
belonged to a network of communication and mutual obligation (Cobb,
1976). It is rather ironic that social support should be referred to as
a ”really fundamental variable” (Cobb, 1979, p. 103) because researchers
have defined this "fundamental variable” in many different ways. Some
have been circular, asserting that social support is essentially support
provided by one’s social network (e.g., Lin, Simeone, Ensel, a Kuo,
1979). Others have been vague, defining social support as simply the
frequency of social interactions (e.g., Goplerud, 1980). Still others
have focused on the intent, defining social support as "an exchange of
resources between two individuals perceived by the provider or the
recipient to be intended to enhance the well-being of the recipient"
(Shumaker & Brownwell, 1984, p. 11). As Wilcox and Vernberg (1980)
note, the question, "What is social support?" has elicited dozens of
answers. In fact, Cohen and Syme (1985) maintain that there are
probably as many definitions of social support as there are papers on
the topic. Heller and Lakey (1985) have commented that "social support
has come to stand for any type of social interaction that is associated
with health and well-being" (p. 296)!

Social support researchers do agree that social support, like
social interaction, is clearly a multidimensional construct (Leavy,
1983; Sarason & Sarason, 1985; Thoits, 1985; Wilcox & Vernberg, 1985).

The issue of defining social support is now centered on what the

8
components, or dimensions, of social support are. These components are
commonly referred to as "functions”. (The reader is referred to
Gottlieb (1978), House (1981), and Thoits (1985) for discussions and
empirical derivations of the various components or functions of social
support).

Although researchers differ in how they conceptualize and label
support, five reasonably-common components can be extracted from the
literature. Emotional support refers to a class of behavior that
includes the conveyance of concern, love, empathy, and trust, such as
confiding in one or comforting one in times of stress. lostromootol
support involves acts that directly help an individual such as paying
bills or helping complete a task. lofotmotioool support refers to
indirect aid or information that the individual can use to cope, such as
information about a job opening or how to complete a task. Aoogaiool
support involves the conveyance of information relevant to self-
evaluation, such as a supervisor informing an employee as to what
constitutes an average performance. Finally, ooooooionoto or social
interaction support refers to actions aimed at diverting or distracting
an individual’s attention away from a problem such as engaging an
individual in athletic activity or playing games. This last function is
also conceived of as simply having someone to do things with, so it is
not clear how the concept is distinct from social interaction in
general.

For a number of reasons, emotional support was the component of
support that was examined in this research. First, the lay person’s
connotation of social support is likely to be what researchers call

emotional support (House, 1981). Thus subjects should find it easier to

9
respond to questions about emotional support. Second, other kinds of
support may not be independent of emotional support (House, 1981;
Thoits, 1985). It is unavoidable that any act of helping may convey
emotional support to an individual. For example, lending someone money
or giving them a ride can express, indirectly, concern and caring.
Further, as House and Kahn (1985) discussed in their review of the
measurement of social support, the functions of social support are
conceptually independent; empirically, however, there is some evidence
that they are interdependent. Measures constructed by researchers
designed to test different functions of social support frequently are
found to be highly correlated. For instance, in one study, reviewed by
House and Kahn (1985) the average correlation between four of the five
measures of support was .90. Some researchers (Brookings & Bolton,
1988; Caldwell, personal communication, 28 April, 1988; McCormick,
Siegert, & Walkey, 1987; Stokes & Wilson, 1984) have conducted factor
analyses of various measures of support. Generally, large correlations
between the subscales suggest an overall general support factor.
However, there also appears to be some unique variance contained in the
subscales. Brookings and Bolton (1988) concluded that use of either the
total score or the subscale scores is empirically defensible for the
measure they investigated. Others do not agree that use of subscale
scores is defensible, unless one is striving for particular sensitivity,
arguing that people who provide emotional support to others tend also to
provide the other kinds of support (c.f., Sarason, Sarason, Shearin, a
Pierce, in press; Sarason, Shearin, Pierce, & Sarason, 1987).

Similarly, House and Kahn (1985) argue that a priority should be to

10
measure emotional support first with other aspects following as
appropriate.

Third, most researchers (e.g., Cobb, 1976; House, 1981; Sarason et
al, 1987; Thoits, 1985) claim that emotional support is the essence of
social support, with other kinds of support playing less salient roles.
Sarason et al (1987) concluded that in varying degrees, each of the
measures they examined assessed feelings of being loved and accepted,
and measures that tap directly these feelings "may provide the most
accurate assessment of the construct"(p. 831). Specifically relating
social support to health, Cobb (1979) claims that emotional support ”is
more important than all the others put together" (p. 94). Also of
pertinence is Thoits’s (1985) discussion of the conclusions reached by
reviewers of social support literature: "Socioemotional support from
significant, or primary, others appears to be the most powerful
predictor of reduced psychological distress or disorder, whether
stressful circumstances are present or absent” (p. 54). Thoits claims
that such a conclusion should be viewed as tentative because researchers
generally do not compare the various functions of support within one
study, but she nevertheless notes that the empirical evidence is in
favor of emotional support as the most important component of support.
Gottlieb (1985) and Levy (1983) reached similar conclusions.

Fourth, Gottlieb (1978) found that of all the helping behaviors
reported by his sample of single mothers in response to the three most
stressful events they were currently experiencing, almost half could be
classified as emotional support. Moreover, depending on the problem,
emotional support was particularly salient. For instance an emotional

problem elicited help that was primarily emotional and childcare

11
problems were also most commonly met with emotional support.

Fifth, it is of practical significance to focus on emotional
support in the interaction of "normal" young adults. One would assume
that when sampling a relatively short period of an individual’s daily
life, other kinds of support may be infrequent occurrences. For
example, direct aid, such as helping one move, giving one a ride, or
giving one money, are not daily occurrences. Informational aid such as
giving guidance on how to perform a task or telling one about job
openings are likely only if one is looking for a job or needs guidance.
In contrast, the need or desire for concern or understanding are much
more likely to be daily occurrences.

Finally, asking subjects to furnish a breakdown of the kinds of
support they received, and provided to others, on a daily basis may be
quite a burdensome task whose scope could easily affect the validity of
the data. One must also question the need for such a request in light
of the fact that different dimensions of social support do not appear to
be empirically distinguishable.

For the reasons noted above, the present research focused on
emotional support, operationalized as the extent to which another person
makes one feel loved, accepted, or cared about; or the extent to which
the other makes one feel valued, trusted, respected, or held in esteem
(cf., Cobb, 1976). Emotional support was measured by the RIR because
such an approach permitted a reliable assessment of the exchange of
support in the context of social interaction.

50 Su 0 t° M s rement
To date there are almost as many measures of social support as

there are studies on the topic because each investigator has tended to

12
design his or her own instruments (Cohen and Syme, 1985; House and Kahn,
1985). As a result, the validity and reliability of these measures,
with a few exceptions, are largely unknown (Depner, Wethington, &
Ingersoll-Dayton, 1984; Heitzmann & Kaplan, 1988; Levy, 1983; McFarlane,
Neale, Norman, Roy, & Steiner, 1981). Reviews of the measurement of
social support, such as those by Tardy (1985) and Sarason et al. (1987),
have restricted their discussion to the few measures for which
psychometric properties are known. Of these measures none appears to
have psychometric properties that would support its use in preference to
others (House & Kahn, 1985). Therefore, use of the RIR is justified
because it has already been established as a reliable measure of social
interaction.

Generally, social support has been measured by means of
questionnaires or interviews. These procedures can be classified into
three approaches (House & Kahn, 1985): (1) investigating the existence
of social relationships; (2) conducting social network analyses; and,
(3) examining the functional components of social support.
Investigations of social relationships have been the most simple and
crudest means of assessing social support. This approach has involved
determining the marital status of the subject, their membership in
church organizations or other voluntary organizations, or the number of
friends and relatives they have. The majority of these studies simple
have measured the existence or quantity of relationships. Few have
measured the quality or the frequency of contact within these
relationships (House, 1981; House & Kahn, 1985). Nevertheless, these

measures have been shown to be fairly objective, simple, and reliable

(House & Kahn, 1985).

13

Measures of social networks have been more elaborate and
informative of an individual’s social life. However, the cost-
effectiveness of such an approach is questionable, as is the validity of
the measures (House 8 Kahn, 1985). The emphasis of this approach is on
the structure of an individual’s social relationships. Measures include
network characteristics such as size, sex composition, density or
connectedness (i.e., to what extent one’s friends and/or relatives know
each other), durability, frequency of interaction with network members,
reciprocity, and homogeneity. Most researchers have not examined all
these characteristics in one study but rather examine one, or a few,
particular characteristics. Most popular has been size of the network
followed by its density (House & Kahn, 1985). As with the social
relationship approach there still remains a need to examine the quality
of network variables and the amount and quality of support that network
members extend to the individual (House, 1981).

The measurement of the functional components of social support has
also been diverse. Measures have either tapped a global dimension of
social support or have attempted to measure specific components. Some
measure the perception of support that is available while others measure
the actual amount received. Some investigate the availability of
support and/or the satisfaction with, or the quality of the support that
is available. Moreover, the various measures are not strongly related
to one another (Barrera, 1986). For a more critical analysis of these
measures and others the reader is referred to Barrera (1986), Heitzmann
and Kaplan (1988), House and Kahn (1985), and Tardy (1985).

Regardless of the diversity of measurement most studies share two

characteristics. First, studies have measured oerceived support in

14
contrast to actual support. For example, in Heitzmann and Kaplan’s
(1988) review of twenty-three measures of support they describe only two
scales as measuring the actual receipt of support. Of course, measures
of the amount of support received are still very subjective and may be
conceived of as measuring ”perceived-received" support (Barrera, 1986,
p. 417). As House (1981) notes, this is "appropriate because social
support is likely to be effective only to the extent it is perceived”
(p. 27). This is very much in line with the Lewinian and Symbolic
Interactionist perspectives that the environment or reality is as one
interprets or perceives it to be. House does argue, however, that more
attention needs to be paid to the assessment of objective support,
meaning the amount of support that others perceive they provide to the
individual. Clearly these measures are also subjective but they are at
least independent of the subject’s perception (House, 1981).

Second, and perhaps most consistently, studies conducted to date
have been directed at the receipt of support and not the provision of
support (House, 1981; Tardy, 1985). It is rather interesting that
social support research should focus on receipt to such a large extent,
when one considers the helping or altruism literature in social
psychology which, until recently, has focused entirely on the provision
of help. More than likely the relationship between the receipt of
social support and various well-being measures is responsible for this
approach. Of those studies that have investigated the provision of
support, the investigation has concerned reciprocity, asking questions
such as "would this person come to you to discuss home and family?"
(McFarlane, Neale, Norman, Roy, & Steiner, 1981, p. 91). Others

(Ingersoll-Dayton & Antonucci, 1983) have asked if an individual has a

15

relationship in which confiding was a mutual activity and if there were
people who would take care of them when ill and to whom they would
return the favor. The measure of reciprocity was the number of people
who received support from the individual minus the number who provided
support. These measures clearly do not directly measure the amount of
social support provided, thus little is known about "the other side of
the coin”.

As already noted, there appears to be no outstanding empirical or
conceptual measure of social support. Therefore, the RIR was chosen
because of its psychometric properties. Using the RIR to measure
support is not entirely novel. Cutrona (1986) recently used a variant
of this instrument to measure the number of helping behaviors that were
conveyed to an individual during social interaction. However, she did
not measure the amount of help extended. For instance, an individual
might be offered advice in a given interaction but this advice could
range from a little to a great deal. Furthermore, as Gottlieb (1978)
has already suggested "The task of designing rigorous evaluative
research on natural support systems is complicated by the fact that
informal helping transactions arise spontaneously as the natural by-
products of people’s participation in their social networks and are,
therefore, almost impossible to observe iooxixo ..... it may be feasible
to have the participants themselves complete daily or session-by-session
logs in which they record their social interactions using a
classification scheme" (p. 114).

It would seem most appropriate then, to measure support as an
ongoing interaction process and, as noted earlier, if one conceptualizes

support in this way then the research must be guided by the norms of

16
social interaction. Specifically, attention must be paid to sex
differences and the exchange of resources because of their demonstrated
importance to the issue. The next sections review the evidence on sex
differences and exchange in social interaction and social support.
‘. I : - - 1 7 o .‘o u ._ 4400

Perhaps the most obvious way in which sex exerts its effects on
social life is simply on the frequency of social interaction. Females
have been found to spend more time socializing than have males (e.g.,
Sullivan, 1986; Wheeler and Nezlek, 1977) and to have more close friends
than have males. For instance, middle-aged adult males report an
average of 3.2 close friends while middle-aged adult females report an
average of 4.7 (Bell, 1981). In that study, every female reported
having at least one close friend whereas ten percent of males reported
having no close friend (Bell, 1981).

Both sexes most commonly engage in same-sex contact (Sullivan,
1986; Nezlek, Wheeler, & Reis, 1983A; Reis, 1986) spending between twice
to three times as much time with the same sex as with the opposite sex.
Although females prefer same-sex friendships more than do males (Rose,
1985) and fewer females than males have opposite-sex friends (Booth &
Hess, 1974), Sullivan (1986) did not find any sex difference in the
amount of time that subjects spent socializing with same-sex or with
opposite-sex cointeractants. Nezlek, Wheeler, and Reis (1983A) did find
sex differences in their research on social interaction. Males spent
more time than did females socializing with the same sex, whereas
females spent more time than did males socializing with the opposite
sex. Sullivan’s (1986) more restrictive definition of social

interaction may explain this discrepancy.

17

A very robust and reliable sex difference in social activity is the
finding that females tend to affiliate in dyads while males affiliate in
groups. This difference emerges as early as kindergarten (see Winstead,
1986). Males and females also differ in the style or content of their
interactions. In a review of sex differences in same-sex friendships
Winstead noted that female interaction tends to be characterized by
talking. Frequently the conversations are in-depth talks about feelings
and concerns and involve mutual confiding. Male interaction, on the
other hand, is better characterized by activity of some sort, or
conversations about activities. This differentiation of interaction
content has been documented elsewhere (e.g., Wheeler & Nezlek, 1977) and
is particularly salient in group settings.

Aries (1976) found that the conversations of female groups centered
on the expression of feelings and self-disclosure. The conversations of
male groups centered around story telling and competition. When the
subjects were placed in mixed-sex groups the conversations of males
became more "female-like" by involving more self-disclosure than had
been the case in the all-male groups. Over time the females showed a
preference for the all-female groups while the males showed a preference
for the mixed groups. This last finding has also been reported by Dabbs
and Ruback (1984) who noted in their study of same-sex groups that the
females appeared to be enjoying themselves more than the males were.
They also found that the female groups vocalized more than the male
groups did. Vocalizing was not limited to talking but included other
sounds such as laughing.

Aries (1982) reported similar findings in her study of mixed-sex

groups. Further, even though the sample was comprised of bright,

18
career-oriented, motivated individuals and the females initiated many
interactions they nevertheless maintained a "reactive” stance by
agreeing or disagreeing to suggestions. The males were more likely to
adopt a "proactive" stance by making suggestions and giving opinions.
Other evidence indicates that males are more likely than females to
control interactions when females are present. Males also report more
self-initiation and having had more influence over opposite sex
interactions than do females (Nezlek, Wheeler, & Reis, 1983A).

Perhaps the most stereotypical sex difference in social interaction
is the intimacy of the interaction. For instance, most middle-aged
males have never told their closest friend that they like him (Bell,
1981). Females are more likely to express feelings of affection and
love for same-sex friends than are males (see Winstead, 1986). Females
of all ages are also more prone to engage in self-disclosure (Berg &
McQuinn, 1986; Cozby, 1973; Jourard, 1979), and self-disclosure to a
greater level of intimacy (Grigsby & Weatherby, 1983), than are males.
These findings have been replicated with eight and eleven year old
children (Cohn & Strassberg, 1983). Komarovsky (1976) has speculated
that this sex difference in levels of self-disclosure may disappear in
close heterosexual relationships. Parelman (1983) found this not to be
the case. In marriage, the sex difference was still apparent. Females
are also more prone to disclose their negative traits than are males
(Kopfstein & Kopfstein, 1973).

Overall then, interactions in which females are present are rated
as more intimate or "affectively richer" (Nezlek, Wheeler, & Reis,
1983A, p. 64) or "meaningful" (Reis, 1986, p. 94) than all-male

interactions. Fischer and Norus (1981) report data that show that

19
female-female close relationships are rated highest on intimacy while
male-male close relationships are rated lowest on intimacy of all
possible sex compositions. This suggestion, that all-female
interactions or relationships are more intimate than female-male
interactions or relationships, is generally not supported by other
research. The research by Nezlek and his colleagues (1983A) showed no
significant differences in intimacy betweeh female-male and all-female
interactions in terms of intimacy. Further, Aries (1976) showed that
males in mixed sex groups were more self-disclosing than were males in
all male groups. It would seem that once a female is present in an
interaction the balance swings toward more intimacy, thus the most
marked sex composition effect on intimacy is the all male interaction
which consistently turns out to produce the lowest level of intimacy in
an interaction or relationship. These results suggest that females
bring emotion, or intimacy, into social interactions.

The empirical evidence on intimacy oreemotion supports Deaux’s
(1977) conceptualization of the different self-presentation styles
developed by the sexes. She argues that males adopt a style of
distancing themselves from others and asserting status. Females, in
contrast, develop an affiliative style aimed at reducing distance
between themselves and others. Deaux’s conceptualization very much
mirrors Gilligan’s (1982) distinction between male and female images of
relationships. Gilligan argues that males see a danger in connection or
intimacy with others, females see a danger in separation or distance
from others, as evidenced by an analysis of Thematic Apperception Test
(TAT) imagery. These conceptualizations are not unlike earlier

distinctions made between the sexes such as Parson’s (1951; Parson &

20
Bales, 1955) distinction between the social roles of females and males,
the one fulfilling the role of home and child care labelled the
expressive role, and the other fulfilling the role of economical and
political care-taker, labelled the instrumental role (see also Eagly,
1987).

Elsewhere Spence, Helmreich, and Deaux (1985) have noted that sex
role researchers are in agreement that the sexes differ on personality
traits. These traits form clusters commonly referred to as expressive
and instrumental qualities (Spence et al., 1985) echoing Parson’s
distinction between the roles females and males characteristically
adopt. A similar distinction is made by Bakan (1966) labelling the
female personality cluster as communion and the male personality cluster
as agency. Evidently, expression, emotion, intimacy, or communion are
more dominant in females than in males, particularly in the realm of
relationships or social interaction.

This concept of females’ communal orientation is particularly
interesting if one considers emotional support as a component of social
interaction. It should follow that this characteristic difference
between the sexes should be reflected in the exchange of emotional
support. However, as Thoits (1985) and many others (e.g., Cohen & Syme,
1985; Heitzmann & Kaplan, 1988) have noted, it is difficult to review
the literature on social support, particularly when attempting to review
the work conducted on sex differences as many researchers have used only
women (e.g., Gottlieb, 1978), others have used only men (e.g, Sarason,
Sarason, Potter, & Antoni, 1985), others (e.g., Barrera, Sandler, &
Ramsey, 1981; Cutrona, 1986; Sarason, Levine, Basham, & Sarason, 1983,

study 1) have not reported any analyses testing for sex differences,

21
while still others (e.g., Rook, 1987, study 1) have not specified the
sex of their subjects. As it stands we know very little about sex
differences in social support (Burda, Vaux, & Schill, 1984; Rosenthal,
Gesten, & Schiffman, 1986).

Of those who have examined sex differences—-a variety of measures
have been used. Stokes and Wilson (1984) used the Inventory of Socially
Supportive Behaviors (ISSB; Barrera, Sandler, & Ramsey, 1981) which taps
the frequency of receiving 40 specific supportive behaviors within the
last month. The 1538 measures four kinds of social support--emotional,
instrumental, informational, and socializing (Stokes 4 Wilson, 1984).

On total 1588 scores, males and females did not differ, however females
did report receiving significantly more emotional support than did
males. Similarly, Caldwell, Pearson, and Chin (1987) examined sex
differences on the ISSB and their own measure of social support—-the
Social Dimension Scale (508). The SDS measures the amount and
satisfaction of support received from same sex friends, opposite sex
friends, family, and helping professionals. Their results showed that
females reported receiving more support from same-sex and opposite-sex
friends and from family than did males. Females also scored higher on
the 1858 than did males. Caldwell et al used the total scores from each
measure thus collapsing across the four components of support that each
one measures. This was done because a factor analysis of the measures
revealed that only one factor emerged across the four components
(Caldwell, personal communication, April 28, 1988).

Burda, Vaux and Schill (1984) found that females perceived more
support from friends and family than did males on a global measure of

support. On more specific measures of support, females reported

22
receiving more emotional support than did males but the sexes did not
differ on the receipt of information, instrumental, or companionate
support.

Butler, Giordano, and Neren (1985) asked their subjects to rate the
severity of the most stressful event they had experienced in the
previous year and to rate how well they believed they had coped with the
event. They did not find any sex differenCes on either of these
measures. However, when subjects were asked how much support they
perceived they had been offered females reported receiving more support
from friends and family than did males. Although the measure used by
Butler et al. (1985) tapped five dimensions of support, the authors
collapsed these dimensions to obtain a total support score.

Overall the evidence demonstrates sex differences in social support
with females perceiving that they are given more support than are males.
There are a number of explanations for these findings. The most obvious
is that females ask for more support. Butler et al (1985) found that
females reported requesting more support from family and friends than
did males. Further, females self-disclose more than males do (Cozby,
1973; Jourard, 1979) and express more negative traits about themselves
than males do (Kopfstein & Kopfstein, 1973) which may elicit support
from others. In support of this argument Berg and McQuinn (1986) have
found that females self-disclose more than males in dating relationships
and also report receiving more favors and help from their partners than
do males. Berg and McQuinn (1986) did not report correlations between
these measures, however, so the relationship between the two is not
clear. Also, females have been found to be more depressed than males

(Weissman & Klerman, 1977) and depressed females have been shown to

23
somehow convey their depression to others (Coyne, 1976). These two
factors may combine to elicit support from others although Coyne found
that the conveyance of depression led to negative affect and less
acceptance by others.

Another explanation for females’ greater receipt of support is that
it is more rewarding to support a female than a male. Females report
more satisfaction with overall support (Caldwell, Pearson, & Chin, 1987)
and with emotional support (Rosenthal et al., 1986) than do males. If
they convey their satisfaction to the source of support they may
increase the likelihood of future emotional support being provided.
Overall, the evidence suggests that the social interactions of females
are characterized by emotional support to a greater extent than are the
social interactions of males.

The findings indicating that females request more support than
males coupled with the expressive orientation of females would suggest
that females need or desire more support than males do. Indeed,
Rosenthal et al (1986) found that females reported more of a need for
emotional support than males reported. However, the sex difference was
attributable to an interaction with sex roles. Among the sex role
groups the only difference found was that sex-typed males reported less
of a need for emotional support than did sex-typed females.
Nevertheless, one would expect to replicate the overall sex difference
for emotional support in the context of social interaction.

Up to this point the discussion has been concerned with sex
differences in social interaction and the receipt of social support. As
noted earlier, social exchange or reciprocity is a factor fundamental to

social interaction. Yet the exchange, or reciprocity of support is a

24

neglected topic. The next section reviews the evidence on reciprocity
within social interaction and social support.
R ci ' in l r i l

Berg and Clark (1986) have argued that individuals determine the
norm of exchange in interactions very quickly. In discussing a number
of studies conducted on social exchange they concluded that "the nature
of social exchange was differentiated within the first hour!” (p. 113).
Further, determining whether the relationship lasts can be predicted
from measures of attitudes and social exchange (Berg & McQuinn, 1986).
However, the authors note that social exchange is only one factor in the
evaluation and maintenance of a relationship. Nevertheless, if social
exchange is not perceived as fair or equitable, the individuals in the
relationship will become distressed and attempts will be made to restore
equity, regardless of whether the individuals are being overbenefited or
underbenefited by the exchange (Walster, Berscheid, & Walster, 1973).

Research on social interaction demonstrates that the norm of
reciprocity is particularly salient in the area of self-disclosure. For
instance, Cohn and Strassberg (1983) found that children as young as
eight years of age were able to reciprocate the level of self-disclosure
of another child who was a stranger to them. These authors note that
their results replicate the findings of self-disclosure reciprocity
among adults (e.g., Cunningham a Strassberg, 1981). In discussing the
findings that disclosure reciprocity may be greater among strangers than
among friends or spouses (e.g., Morton, 1978), Berg and Clark (1986)
argue that these findings do not imply that reciprocity of disclosure is
any less among individuals who are well acquainted with each other.

They maintain that strangers may reciprocate self-disclosure because of

25
an immediacy norm, whereas intimates may be less likely to reciprocate
immediately because of the assumption that such an opportunity will
arise in a future interaction. Disclosure reciprocity may operate at
the same level among friends and strangers but friends may choose to
exercise the possibility of "catch you later" that would not be Open to
strangers and obviously would not be measured in a short interaction in
the laboratory.

One would expect the same norm of reciprocity to apply to support
as a component of social interaction yet such investigations have been
rare. Ironically, the original definition of social support (Cobb,
1976) emphasizes reciprocity as a defining feature of support in
referring to "mutual obligation” (p. 300). The research that has been
conducted on reciprocity has tended to be on samples selected from
abnormal populations with a focus on reciprocal Lolotiooohtos rather
than reciprocal SUDEQLI- Further, reciprocity was not the main focus of
these studies but rather a post hoc consideration. Reports that those
to whom support was given were also likely to be sources of support does
not give any information about the omooot of support exchanged. An
individual may provide a great deal of support to another and only
receive a small amount in return. This would be classified as a
reciprocal relationship. However, it cannot be classified as reciprocal
support, in the sense of being equitable. For instance, in their
research on the development and evaluation of the Social Relationship
Scale, McFarlane and his colleagues (McFarlane et al, 1981) used a
sample of married couples involved in a parent therapist program,
couples in treatment because of family problems, community college

students, and graduate students. Their measure of reciprocity was the

26

number of people who would come to the respondent to discuss various
problems. When describing the number of people to whom they felt they
could turn, to discuss various categories of potential life stress, the
mean number of people that the respondent mentioned three times or more,
was 2.59. Of these people a mean number of 1.62 was considered to be
reciprocal, i.e. these people would also turn to them. When asked about
focal individuals who could be turned to in a major crisis--eighty-seven
percent of these individuals were people who would also turn to the
respondent. Thus a sizable proportion of relationships were perceived
to be reciprocal, but the actual amount of reciprocity was not measured.

Ingersoll-Dayton and Antonucci’s (1983) study of the reciprocity of
support with adult females and males appears to be the only study
specifically focused on the reciprocity of relationships. The aspects
of relationships examined were confiding in another and providing care
for another during illness. The measure of reciprocity was the number
of individuals from whom such support was received minus the number to
whom such support was provided. Among spouses, perceived reciprocity
was quite high with more than 80% reporting mutual sick-care and
confiding. The rate of reciprocity was somewhat lower for friends (63%
and 56% respectively) and children (65% and 51%). Those who reported
non-reciprocal relationships tended to provide support to a greater
number of others than provided to them. Again it appears that there is
a high level of reciprocal relationships but the amount of support is
not being measured in this paradigm.

Tolsdorf (1976) compared the reciprocity of support among two
groups of males hospitalized for psychiatric problems or medical

problems. Results showed that the medical group demonstrated reciprocal

27
support by giving as much support, advice, and feedback to members of
their social network as they received. The psychiatric group was found
to receive more support than they gave. Tolsdorf maintained that the
imbalance was due to the reported unwillingness of the psychiatric
subjects to return the support. However, he also reported that the
networks of these subjects were more heavily dominated by family members
than were the networks of the medical subjects. It may be the case that
individuals with problems are provided with more support than they want,
or need, and therefore they are unwilling to return the “favor“. Either
way, caution is warranted in making conclusions about reciprocity when
studying abnormal individuals. Another atypical sample, elderly
individuals who resided in single-room occupancy hotels, was
investigated by Cohen and Sokolovsky (1979). They classified the
support relationships of these people as either reciprocal,
instrumental--provision of support greater than receipt, or dependent—-
receipt of support greater than provision. The authors do not report
the breakdown of these different kinds of relationships except to note
that the older subjects had more dependent relationships than the
younger ones.

Taken together these studies tell us little other than the fact
that individuals tend to function in reciprocal relationships of
support. They reveal nothing about the actual amount of support being
reciprocated or about sex differences in the provision or reciprocity of
social support. A notable exception is the finding of Burda, Vaux, and
Schill (1984) that females report a larger percentage of reciprocal
support relationships than do males. Obviously the provision of support

is half the picture of reciprocity and it has already been demonstrated

28
that there are sex differences in the other half of the picture with
females receiving more emotional support than males. In his review of
sex differences in social support Vaux (1985) does not mention any
studies on the provision of social support. He does hypothesize though
that females may be more likely than males to provide support. There is
other tangential evidence to support his hypothesis. Results from a
study of middle-aged and older ethnic groups showed that social ties
were negatively related to psychological adjustment for first and second
generation Italian and Polish women (Cohler & Lieberman, 1980). The
authors concluded that this relationship was perhaps due to the
responsibility of caring for more people suggesting that females are
more likely to be support providers than males.

Bernard (1981) also argues, in her review of female friendship
research, that in the nineteenth century female friendships provided the
emotional support that females did not receive from males. A century
later it is still the same with males reporting that their closest
confidant is their wife, and females reporting that their closest
confidant is another female, with this effect being more salient in blue
collar marriages. These results suggest that both males and females
turn to females for the provision of emotional support.

Unfortunately, many of the studies of sex differences in social
support have not fully crossed the sex of subject with the sex of the
target. For instance, Berg’s (1984) research on the development of
friendship between pairs of roommates showed that females received more
help with problems and projects from their roommate than did males.
Because the investigation obviously had to be concerned with same-sex

pairs it is not clear if the sex difference was due to a sex of subject

29
effect or a sex of target effect, although the evidence discussed above
would suggest a sex of subject effect.

Other research has involved dating or married couples as subjects,
again preventing target sex by subject sex analyses. Berg and McQuinn
(1986) looked at the social exchange between continuing and non-
continuing dating couples over two time periods. They found that with
couples who continued to date, the females initially gave more resources
(did things to help or give pleasure to the partner) than did the males,
but with couples who broke up, the males reported more initial giving
than the females did. Perhaps, females only give when they feel that
the relationship may continue or when they have some feelings for their
partner, or males who appeared too attentive were unattractive to these
females.

Others have argued that females provide more social support than do
males because of the demands of their social role, e.g., motherhood.
Fischer (1982) reports that females with children were particularly
likely to feel demands (presumably requests for support, implicit or
explicit) from their family. "Children demand, and women respond to
these demands, as well as to the demands of others" (p. 136). Belle
(1982) argues along similar lines, maintaining that females provide most
of the support in society because of their sense of connectedness to
others. She argues that females are likely to find themselves involved
in non-reciprocal support relationships such as the mother-child
relationship. These arguments are quite valid, however, they appear to
stress the social role of the female as being responsible for the excess
support she supplies. This would suggest that if males were to occupy

the same social roles they would be as supportive. The research

30
discussed above on roommates and dating partners casts doubt on this
argument. These are identical social roles for males and females but
there are sex differences. Of course, even when the role is the same
for females and males, each sex may enact the role differently, based on
stereotypical role requirements. Thus, college life may not erase sex-
role requirements, however, the role requirement for females and males
are as equitable as they are likely to be in the life-course and do not
blatantly elicit sexist norms, particularly those pertaining to
helpfulness. I would argue then, that if females provide more support
than males it is not because of their social role requirements but
because of some other aspect of their make-up. The evidence discussed
above is not as “empirically pure” as one would like when addressing sex
differences in the provision of social support. As noted, target sex
has not been crossed with subject sex to produce an experimental design
to answer the question directly. Rather, one is left to infer from the
cross sex studies and the same sex studies that females provide more
support than males.

Of interest here is the fact that the most closely related area of
social psychology--helping and altruism--has travelled the same route as
the social support literature but in reverse. It was remarked earlier
that social support research is directed towards the receipt of support,
while the helping research is directed toward the provision of help.
Another major difference is that social support is concerned with
"everyday" helping behaviors from friends and relatives over certain
time periods while the helping studies have attended to brief
interactions among strangers and research accomplices in an emergency

situation. It is hard to believe, that essentially, these two fields

31
are measuring the same behavior: expressing concern and acting to
alleviate another’s distress. They do share one thing in common,
however: a rare attention to sex differences.

In a recent review of the helping literature Piliavin and Unger
(1985) summarized by stating that no firm conclusions could be reached
regarding the effects of sex on the helping process because 'no serious
attention has yet been paid to the dynamics of the role of sex and
gender in the helping relationship" (p. 181). The authors call
attention to a problem mentioned above, namely the need to examine the
interactive effects of sex of recipient and sex of helper. Up to this
point, the physical attacks and the falls have been played by female
accomplices, whereas the heart attacks and robberies have been played by
male accomplices. If sex of subject is examined at all it is usually
entered into the analyses post hoc--it is quite unlikely to be a major
independent variable (Piliavin & Unger, 1985).

More recently, Eagly (1987) conducted a meta-analysis of sex
differences in helping behavior. She reached much the same conclusion
as Piliavin and Unger (1985) which is that no conclusion can be reached
about sex differences in helping relationships. Eagly’s treatment of
her review was based on a social roles analysis. Although she found
that males are more helpful than females she argued that such a result
is due to the specialized situations of the research which hold dangers
and difficulties for females to a greater extent than males. Indeed,
when she had subjects rate the situations she was analyzing on a number
of dimensions she found that males, in contrast to females, rated
themselves as more comfortable, competent, and likely to help. Eagly

went on to argue that the studies have examined the type of help that

32
males are more likely to provide and have ignored communal forms of
caring which is the kind of help than women are likely to provide.

At this point there appears to be no clear cut evidence that
females provide more help, of any kind, than do males. However, the
research on social support discussed earlier does suggest that females
provide more support than do males. There are also other reasons to
support this expectation. First, the female stereotype of nurturance,
connectedness, and kindness is very conducive to helping behavior of a
communal or emotional kind. Second, females have been shown to be more
empathetic than males which is a trait characterized by emotional
matching and/or sympathetic responding (Eisenberg & Lennon, 1983) and is
related with helping behavior (Fultz, Batson, Fortenbach, McCarthy, &
Varney, 1986). Moreover, females have been shown to be more in tune to
the needs and desires of their loved ones, to be more affected by the
stresses of their loved ones, and to say that people are more important
to them than have males (see Kessler, McLeod, & Wethington, 1985, for a
review of these studies). If females are so affected personally then
they should be more likely to perceive a need for emotional support and
be more responsive than males by intervening and giving support when
they perceive the need for it. The empathetic and interpersonal
orientation of females would suggest not only that females provide more
emotional support than do males but that they find it easier to do so.
Earlier, it was noted that Eagly (1987) found that males rated
themselves as more comfortable, competent, and likely to help than
females, in situations conducive to "male-helping". The reverse should
hold true: that females feel more comfortable and competent providing

help of a communal or emotional kind than do males.

33

In conclusion then, the purpose of the present research was to
examine sex differences in social interaction among college-aged adults
with the main focus being on the exchange of emotional support. The
following hypotheses were developed.
Hyootheseg
1. Research that has investigated sex differences in social support has
consistently shown that females report receiving more support than do
males (e.g. Caldwell, Pearson, & Chin, 1987; Stokes a Wilson, 1984). It
was expected that this finding would be replicated, therefore, females
will report receiving more emotional support than will males.
2. Many arguments have been presented to suggest that females provide
more support than do males. These arguments center around females’
greater comfort with interpersonal issues, their greater intimacy and
empathy in interaction, and their greater sense of connectedness to
others, in contrast to males (Belle, 1982; Bernard, 1981; Eagly, 1987; &
Fischer, 1982). In line with these arguments it was hypothesized that
females will report providing more emotional support than will males.
3. Rosenthal et al (1986) have found that females report a greater need
for support than do males. Females also request more support from
others than do males (Butler et al, 1985). It was hypothesized,
therefore, that females will report a greater need or desire for
emotional support than will males.
4. Given that females have been shown to be more empathetic (Eisenberg
& Lennon, 1983), and more in tune to the needs and desires of their
loved ones (Kessler et al, 1985) than have males, it was hypothesized

that females will perceive a greater need for support on the part of

others than will males.

34
5. Because females have been shown to be more empathetic (emotional
matching; Eisenberg & Lennon, 1983), and more affected by the stresses
of their loved ones (Kessler et al, 1985) than have males, and given the
female stereotype of nurturance and kindness, it was hypothesized that
females will be more responsive than will males such that there will be
a stronger relationship, for females than for males, between the other’s
perceived need for support and the provision of support.
6. As noted above, females are more affected by the stresses of others
than are males and they are more empathetic (sympathetic responding;
Eisenberg & Lennon, 1983). Eagly (1987) has also suggested that females
may be more comfortable, than would males, in providing help of a
communal kind. These arguments suggest that there will be no
relationship for females between the provision of support and the
difficulty of providing support. A positive relationship between these
variables is hypothesized for males.

Earlier it was stated that among the main attributes distinguishing
social interaction for college students were intimacy and quality
(Forgas, 1976). One must question how the exchange of support relates
to these other attributes of social interaction. Quality has been
defined (Wheeler & Nezlek, 1977) as the degree to which the interaction
is pleasant. Intimacy has been defined as having components such as
meaningfulness (Nezlek & Wheeler, 1984; Reis, Sanchek, & Solomon, 1985),
self-disclosure, cohesiveness, support, and sexuality (see Clark & Reis,
1988). For present purposes, intimacy was considered to be
cohesiveness, or closeness. Obviously, the pleasantness of an
interaction is entirely a subjective judgment, but one would assume that

a conversation about the weather would not be rated by many as a close

35
interaction and would therefore be low on intimacy. A conversation
about one’s personal fears and values, on the other hand, would probably
be rated as rather close and thus high on intimacy. One would expect
that the exchange of emotional support, earlier defined as the
conveyance, or receipt, of feelings of being cared for or held in
esteem, would be rated as quite close and thus high on intimacy. There
is no reason to expect that the sexes would differ in their
interpretation of such an interaction. One would expect sex
differences, however, on the relationship between the exchange of
emotional support and overall quality of the interaction. As discussed
previously, the conversations of males tend to be less personally
focused than those of females. Further, male interactions tend to be
more task-orientated than those of females suggesting that emotional
support is not necessarily seen as an enjoyable component of male
interaction. On the other hand the conversational content of female
interaction suggests that for them emotional support is an enjoyable
pursuit. One would expect that emotional support is not related to the
quality of interactions for males but positively related to the quality
of the interaction for females. The following hypotheses were
developed:
7 a. There will be positive relationships for both sexes between the
intimacy of interactions and support received.
7 b. There will be positive relationships for both sexes between the
intimacy of interactions and support provided.
7 c. There will be a positive relationship for females between the

quality of interactions and support received. There will be no

relationship for males.

36
7 d. There will be a positive relationship for females between the
quality of interactions and support provided. There will be no
relationship for males.

Previous research has shown that support relationships tend to be
reciprocal, that is, people to whom help is provided tend to return this
help (Ingersoll-Dayton & Antonucci, 1983). Further, the self-disclosure
literature shows that self-disclosure tends to be reciprocated, in terms
of amount and intimacy (e.g. Cunningham 8 Strassberg, 1981). 8. It was
hypothesized then, that emotional support would also be characterized by
equitable reciprocity.

Social Sgooort, Parsonality, ano Poaitiva and Negative Affaot

Two additional relationships were investigated; the relationship
between social support and mood, and the relationship between social
support and some personality indices. Many studies have shown
consistent relationships between mood and helping. Because the present
study concerned helping behavior, it was considered pertinent to
investigate its relationship to mood. Analyses were conducted on the
relationship between sex, the support measures, and the personality
measures, as sex might, or might not, be the principal individual
difference variable distinguishing between levels of support.

The author is unaware of any studies conducted on the relationship
between support and mood, although much research has documented the
positive relationship between the receipt of support and various
measures of well-being (see Cohen & Wills, 1985; Leavy, 1983; for
reviews) of which mood is a component (Diener, 1984). Numerous studies
have been conducted, however, concerning the effects of mood on helping

behavior from a social psychological perspective. For instance, Isen

37
and Levin (1972) found that good mood induced by receiving cookies or a
dime led to more helping behavior than did a neutral mood. This
finding, that a good mood promotes helping behavior, has been replicated
in much research (see Carlson, Charlin, & Miller, 1988 for a meta-
analytic review). Also positively associated with helping behavior is
negative mood (Cialdini & Kenrick, 1976; Kenrick, Baumann, & Cialdini,
1979) but only when helping is perceived as a means to alleviate the
negative mood. Moreover, mood has been associated with social activity
in general. For instance, Stone (1987) found that desirable social
events were positively related to positive mood and negatively related
to negative mood. In a study conducted with Japanese students Clark and
Watson (1988) found that the amount of important social events reported
over a period of 90 consecutive days was positively related to positive
mood and generally unrelated to negative mood. Social events that
specifically involved helping behaviors were positively related to
positive mood although not significantly so. They were unrelated to
negative mood.

The above research suggests that the provision of support will be
positively related to positive mood or affect. It was expected that the
previous findings, although based on more temporally direct analyses,
would generalize to the present study even though the measures in the
present study were averages over a two-week period. Clear predictions
for support provision and negative mood cannot be made. Experimentally
induced negative mood sometimes leads to increased helping, however, if
one considers helping to be a positive social event then the
correlational studies show both a negative relation and no relation

between provision of support and negative mood. Neither does previous

38

research suggest hypotheses pertaining to the relationship between other
support variables and mood. One would expect though that the receipt of
support is generally a positive event and would be related to positive
affect and that need for support is a negative event and would be
related to negative affect.

The relationship between personality and support was also explored.
The variable that is perhaps most conceptually and empirically (e.g.
Barnett, Howard, King, & Dino, 1981; Krebs, 1975; Schaller & Cialdini,
1988) linked with helping behavior is empathy, defined as the ability to
role-play and thus vicariously experience the distress of another.
Mehrabian and Epstein (1972) found that empathy was "a primary
personality attribute for predicting helping behavior" being a better
predictor than similarity to, or liking for, the person in need of help.

Previous research has shown that certain personality traits
discriminate different amounts of support received. For instance,
Dunkel-Scheffer, Folkman, and Lazarus (1987) have shown that the higher
subjects are in self-esteem, religiosity, and value for authority, the
more emotional support they report receiving. Sarason, Sarason, Hacker,
and Basham (1985) report that social skills assessed by subjects’
partners in a dyadic interaction, the experimenter, and themselves were
related to higher levels of social support receipt. Further, in a
longitudinal study Cohen, Sherrod, and Clark (1986) have found that
increases in social skills were positively related to increases in
friendships and perceived support.

The personality measures examined in the present study were empathy
(Mehrabian & Epstein,1972) and social skills (Riggio & Throckmorton,

1986). These variables were selected to determine if potential sex

39
differences in supportive behaviors were being mediated by empathy and
social skills. Females have generally been shown to be more socially
skilled than have males (e.g. Riggio, 1986), and more empathetic
(Eisenberg & Lennon, 1983) at least on self-report measures. The
research question posed is whether sex, empathy, or social skill is the
better predictor of support provision, difficulty of provision, and the
perception of others’ need, and whether social skill or sex is the

better predictor of support receipt.

CHAPTER 11
METHOD

Subjects

Eighty-six females and seventy-four males were recruited for a
study titled "Psychological Record Keeping". All subjects were freshmen
at Michigan State University who participated in exchange for credit
towards their Introductory Psychology class. Their mean age was 18
years, 89% of the sample were white, 8% were black, and 3% were
classified as other. After attending the introductory meeting three
males and two females declined to participate. One male discontinued
participation during the recording period. An additional 16 males and
14 females were excluded because of inaccuracy in their records and
failure to understand the instructions. 1 Thus the final sample
consisted of 70 females and 54 males (78% of those who attended the
initial session).
Matarials

Booheater interaotioo Record (RIR). A modified version of the RIR
was used to measure social interaction. The measure was modified to
include subjective measures of social support in place of the previous
measures of disclosure, satisfaction, initiation, and influence. The

RIR can be found in Appendix A.

40

41
l io P . This form consisted of 24 items
pertaining to positive and negative affect (Diener & Emmons, 1985;
Emmons, 1986). Subjects rated each item on a 1-7 point scale (1 - not
at all, 7 - extremely much). Included in the 24 items were additional
items to measure feelings of helpfulness and sociability. The Personal
Reaction Form can be found in Appendix B.
i l ills Inven or S . The $51 (Riggio & Throckmorton,
1986; Riggio, 1986) was used to assess social skills. The inventory
includes six sub-scales, each containing 15 items. Ratings are made on
9 point scales (1 - not at all true of me, 9 - very true of me). The
Emotional Expressivity subscale measures the ability to send emotional
messages including the nonverbal expression of attitudes. The Emotional
Sensitivity dimension assesses skill at receiving and decoding the
emotional and nonverbal communication of others. The Emotional Control
dimension measures skills of controlling emotional and nonverbal
expression. The Social Expressivity dimension assesses skills of verbal
expression and the ability to engage others in social discourse. The
Social Sensitivity dimension measures skills of receiving verbal
messages and an understanding of norms of appropriate social behavior.
Finally, the Social Control dimension assesses skills of self-
presentation.
Empathy Scale. A 33 item measure of emotional empathy was used
(Mehrabian & Epstein, 1972). All items were rated on a 9 point scale (1
. disagree strongly, 9 = Agree strongly). Sample items include "It

makes me sad to see a lonely stranger in a group", and "Seeing people

cry upsets me".

42

mm:

1otrodootory_moottog. The subjects met in mixed-sex groups of 12
to 16. At this meeting the nature of the study was explained in detail.
Subjects were told that the study was designed to uncover some facts
about people’s patterns of social interaction. An interaction was
defined as any situation involving two or more people in which the
behavior of each person is in response to the behavior of the other
person. Subjects were given some examples of interactions (e.g. a
conversation, dancing). Two examples of what would not constitute an
interaction also were provided (e.g. sitting side by side in class and
not talking, sitting beside someone while eating dinner and not
conversing with them). The subjects were informed that they would be
required to record every social interaction they had lasting 10 minutes
or longer for a period of two weeks. The researcher emphasized that the
success of the study depended on their honesty and full cooperation. 2

The subjects were given a full description and explanation of the
interaction record. As the scales on the RIR were being described the
nature of support was explained. Support was defined as a feeling of
being loved, accepted or cared about; or a feeling that one is valued,
trusted, respected, or held in esteem. Examples of supportive
interactions were provided. The examples were generated from pilot test
undergraduates who were asked for examples of their social interactions
in which support was either provided or received. The subjects were
then informed that they would be required to complete the Personal

Reaction Form at the end of each day, indicating their feelings at the
time of completing the form.

43

Subjects were instructed to read the booklet of instructions at
this time and were provided with a copy to take with them. This
booklet, presented in Appendix C, reiterated what was said at the
meeting, provided examples of social interactions, and outlined in
detail how to complete the interaction records. When subjects had
finished reading these instructions it was stressed that they had to
complete the records at least once each day. If a day was forgotten or
missed it was to be skipped. They were not to attempt to remember
interactions from a previous day. They were once again assured of their
confidentiality in completing the records and told that anyone who
wished could decline to participate at any time.

The subjects were then provided with a packet that contained 14
Personal Reaction Forms; 28 sheets of paper, each one containing 4
interaction records; a booklet of instructions; and a scratch sheet, to
be used to jot down the initials of interactants and time of
interactions if the subject did not have their records with them during
the day. Finally, subjects were told that they would be telephoned
during the first week of recording to check that everything was
proceeding as planned, and that they would also be telephoned during the
second week to arrange a time to return with the completed records for a
post-experimental session. A copy of the verbatim script for this
meeting is presented in Appendix D.

Recoroing period. All subjects completed their recording during
the last week of October and the first three weeks of November of Fall
Term 1987. This time period was chosen so as not to conflict with
midterm examination times or holiday periods. All subjects began their

recording early to mid-week. The purpose of this was to ensure that a

44
subject did not begin recording on a weekend day which would be an
atypical start to their recording. Within three to five days of
recording each subject was telephoned and asked if there were any
unforeseen problems. No subject reported having difficulty with the
procedure. A week later each subject was called to arrange a time for
the post-record-keeping session.

- i - riod ' . Subjects met in groups 14 to 17 days
after the initial meeting. They returned their materials and completed
a post-recording-period questionnaire. This questionnaire contained
items pertaining to reactions to the study, along with the Empathy scale
and the Social Skills Inventory. Subjects were informed that some of
them would be randomly selected for a short interview a few days later.
When the materials were completed subjects were provided with some
background information to the study (See Appendix E) and address labels,
to be completed if they wished to receive the results of the study.

They were then dismissed.

A few days after this session 20 females and 20 males were randomly
chosen, telephoned, and asked to participate in a short interview. All
subjects agreed and were asked questions about their experience with the
record keeping. A copy of the format for this interview is presented in
Appendix F.

The completed data packages for each subject were examined for
accuracy and ambiguity. Any participant whose record appeared to
contain inaccurate data was telephoned and asked questions pertaining to

sources of difficulty or inaccuracy.

CHAPTER III
RESULTS

99W

The Rochester Interaction Record Analysis Package (RIRAP; Nezlek &
Wheeler, 1984) was used to transform the data into summary indices.
RIRAP calculates three different sets of summary variables. One set of
variables represents all the interactions a subject recorded, producing
overall averages for the various dimensions. The second set of
variables represents a breakdown of interactions by sex of interactant,
constructing variables in which either the same-sex or the opposite-sex
were represented. The third set of variables classifies interactions by
degree of closeness according to sex, constructing variables such as
those pertaining to interactions in which the same-sex best friend was
present or in which the second-best opposite-sex friend was present.
Closeness was defined by the number of times a set of initials appeared
in the subjects’ record (see Wheeler & Nezlek, 1977, for empirical
support for this operationalization of friendship).

Seven measures of the subjective aspects of the interaction were
calculated by averaging across all interactions that fell into a
particular category (e.g., same-sex, opposite-sex best friend). These
measures were: support received, support needed, intimacy of

interaction, quality of interaction, support other needed, support

45

46
provided, and difficulty of providing support (e.g., the average
intimacy of same-sex interactions, the average amount of support
provided during interactions in which the opposite-sex best friend was
present). Measures of the quantitative aspects of the interactions
included the following: potoaot calculations of all interactions falling
into a certain category (e.g., the percentage of interactions that were
with the same-sex, the percentage of interactions that were with the
opposite-sex); loooth calculations of the average length of interactions
in a certain category (e.g., the average length of interactions overall,
the average length of interactions with groups); and por_day
calculations of the average number of interactions in a certain category
recorded per day (e.g., the average number of interactions per day with
the best friend, the average number of interactions per day overall).

In addition, the general nature of the interaction, §ooializing,
ralaxing, wont, life necassitx, and othag, was represented by a
percentage breakdown within each category of interaction. All the above
measures were calculated separately for each subject, corrected by the
number of days of record keeping. See Appendix G for the nomenclature
and a more detailed description of all variables calculated by RIRAP. 3
Reliability of the RIR

Subjects maintained the records for a mean number of 13.74 days.
The reliability of the RIR measures could not be assessed formally, thus
the replicability was assessed by comparing the results with those of
similar studies using the RIR (see Wheeler, 1988, Wheeler & Nezlek,

1977; & Nezlek, Wheeler, & Reis, 1983A). These comparisons can be seen

in Table l.

47
Comparing the present findings with those of past research for the
quantitative variables reveals a clear pattern. Interactions with the
same-sex were more common than interactions with the opposite-sex or
with mixed sex interactants. Most interactions involved only one other
person, followed by interactions with two others, three others, and

groups. The present subjects differed somewhat from previous subjects

Table 1
Conparatiye Qnantitativo Indioes of Interaction Pattarn§

 

 

Interaction Present Study Previous Study
With Same-sex 51% 56%
With Opposite—sex 26% 19%
With Mixed—sex 23% 25%
With Groups 16% 22%
With One Person 53% 48%
With Two People 18% 19%
With Three People 13% 11%
Number Per Day 4.9 7.4
Length Per Day 311 363
Average Length 65 50

 

in that they spent less time interacting, and interacted with fewer
different people over the two week recording period (present: 28,
previous: 39).

On the subjective measures, the data from present and previous
research are comparable on two measures; the quality and intimacy of
interactions. Repeated measures analyses, used to examine the quality
and intimacy of interactions among the different sex composition
categories revealed main effects for composition on both quality,

E(1’122)'13o31, 2 < .01, and intimacy, £(1,122)=74.45, Q < .01. As will

48
be presented more fully later, opposite-sex interactions were rated as
more intimate and pleasant than same—sex interactions, findings that
correspond well with those of Wheeler and Nezlek (1977).
Quality of Social 111W

One-tailed t-tests and Pearson correlations were used to examine

 

the research hypotheses. To further explore the subjective aspects of
interactions a number of repeated measures analyses of variance were
conducted. The first repeated measures analysis investigated the sex
composition of the interaction comparing same-sex and opposite-sex
interactions by way of a 2 (sex of subject) X 2 (interaction
composition; same, opposite) analysis with sex of subject as the between
subjects variable and interaction composition as the within subjects
variable. The second set of analyses were 2(sex of subject) X 2(sex of
friend) X 2(closeness of interaction; best friend present, second-best
friend present) repeated measures analyses with sex of subject as the
between subjects variable, and sex of friend and closeness as the within
subjects variable. It was expected that interactions with second-best
friends would be qualitatively "less rich" than interactions with best
friends. It was also expected that potential sex effects of composition
would remain when friends only were investigated. 4

Receipt of Support. In support of Hypothesis 1, females reported
receiving more support, overall (M=2.83), than did males (M-2.35),
t(122)-2.67, p < .01. A 2 (sex of subject) X 2 (interaction
composition: same-sex, opposite-sex) repeated measures analyses revealed
a main effect for composition £(1,120)=66.89, p < .01. More support was

received from opposite-sex interactants (M=3.17) than from same-sex

interactants (3:2.44).

49

The 2 (sex of subject) X 2 (sex of friend) X 2 (closeness) repeated
measures analysis revealed a main effect for sex of friend,
[(1,117)-52.01, p < .01. More support was received from the opposite-
sex friends (M-3.02) than from the same-sex friends (M-2.36).

There was also a main effect for closeness, [(1,117)-6.58, p < .05.
More support was received during interactions in which the best friend
was present (M=2.81) than during interactions in which the second best
friend (fi-2.57) was present. The interaction between sex of friend and
closeness was significant E(1,117)-4.43, p < .05. The interaction
indicated that the effect for closeness occurred among opposite-sex
friends. The means for these analyses are displayed in Table 2, along
with the average number of interactions occurring within each category

of interaction.

 

 

Table 2

“'1‘ i-|p- o ' .. o o i: 0 {1‘4 '- no 1.41.21 .10 .14.. o

I‘ 0| 0 10 of ‘04: .‘ .. n u of I ‘ .g 01 ._ “'0

Female Male

Interaction N X SQ N X SD

Same-sex 40 2.69 1.03 30 2.13 0.84

Opposite-sex 17 3.35 1.41 17 2.95 1.12

Same Best 19 2.55 1.13 15 2.17 1.00

Same Second 11 2.47 1.20 8 2.12 1.05
Best

Opposite Best 10 3.43 1.67 9 2.94 1.30

Opposite Second 5 2.86 1.58 5 2.75 1.28

Best

50
Overall, the results for receipt of support show that females
received more support than did males and that more support was received
from the opposite-sex. This effect was also found when only
interactions that involved the best friends were examined. Closeness of
friendship distinguished between the amount of support received but only
for opposite-sex friends. The closer the friend, the more support

received from him or her.

£noyioion_ot_§nppont. In support of Hypothesis 2, females provided
more support, overall (fl=2.69), than did males (fi22.33), t(122)-2.00, p
< .05. The repeated measures analysis conducted with sex composition
(same-sex, opposite-sex) as the within subject variable revealed a main
effect for composition [(1,120)=44.16, p < .01, and an interaction with
sex, £(1,120)-4.04, p < .05. More support was provided to opposite-sex
interactants (U - 2.97) than to same-sex interactants (fl - 2.45). The
interaction was due to the significant sex difference in providing
support to same-sex others, fi(1,120)=8.97, p < .01. Males provided less
support to same-sex others (5:2.14) than did females (522.70).

The 2 (sex of subject) X 2 (sex of friend) X 2 (closeness) repeated
measures analysis revealed a main effect for sex of friend,
£(1,116)-29.28, p < .01. More support was provided to opposite-sex best
friends (Ma2.85) than to same-sex best friends (M=2.37). The main effect
for closeness was also significant, fi(1,116)=8.25, p < .01. The best
friends (Ms2.72) were provided with more support than the second best
friends (M=2.49). The interaction between sex of friend and closeness
was also significant £(1,116)=4.67, p < .05. Again, the effect for

closeness was due to the difference between opposite-sex best friends

51
and opposite-sex second-best friends. See Table 3 for the means for
these analyses.

In sum, the results for provision of support evidenced a pattern
similar to that for receipt of support. Females provided more support
than males, more support was provided to the opposite-sex, and males
provided less support to same-sex cointeractants than did females.
Again, the opposite-sex friends were provided with more support than
their same-sex counterparts and closeness of friendship affected the
amount of support provided, but only for opposite-sex friends. The best

friend was provided with more support than the second-best friend.

Table 3

Moansnand Standard Deviations for Provision of Snpport a; a
anotion of Interaction Catogory

 

 

 

Interaction Female Male
21 5.11 X 5.12
Same 2.70 1.09 2.14 0.89
Opposite 3.08 1.28 2.83 1.02
Same Best 2.59 1.26 2.16 1.01
Same Second 2.56 1.35 2.05 0.93
Best
Opposite Best 3.16 1.55 2.89 1.26
Opposite Second 2.66 1.53 2.63 1.14
Best

 

Neao for Support. In support of Hypothesis 3, females reported
needing more support (M=2.99), than did males (M=2.37), t(122)=3.29, p <

52
.01. The 2 (sex) X 2 (composition: same-sex, opposite-sex) repeated
measures analyses demonstrated a main effect for composition
£(1,122)=56.99, p < .01. More support was needed from opposite-sex
interactants (M a 3.19) than from same-sex interactants (M - 2.54).

The 2 (sex) X 2 (sex of friend) X 2 (closeness) repeated measures
analysis revealed a main effect for sex of friend, [(1,117) - 52.50, p <
.01. More support was needed from the opposite-sex friends (3.3.05)
than from the same-sex friends (fl-2.46). The main effect for closeness
was also significant, fi(1,117) . 6.39, p < .05. More support was needed
from the best friends (M=2.88) than from the second-best friends

(M=2.63). These means are presented in Table 4.

Table 4
Maans and Standard Deviation; for Neao of Support aa a Ennotion of

Interaction Catogory

 

 

Interaction Female Male

3 so X 50
Same 2.83 1.19 2.17 0.91
Opposite 3.40 1.41 2.91 1.08
Same Best 2.78 1.26 2.22 0.98
Same Second 2.62 1.42 2.06 0.95

Best

Opposite Best 3.56 1.67 2.76 1.11
Opposite Second 3.03 1.72 2.69 1.33

Best

 

53

In sum, females reported needing more support than did males and
more support was needed from the opposite-sex than from the same-sex.
The finding that more support was needed from the opposite-sex held for
the analyses involving only the best friends of each sex. Closeness of
friendship also distinguished between the amount of support needed for
both same and opposite-sex friends. The closer the friend, the more
support was needed.

r f e ’ r . In support of Hypothesis
4, females (Ma3.67) perceived that others needed more support than did
males (fl=3.09), t(122)=2.91, p < .01. A 2 (sex) X 2 (composition: same-
sex, opposite-sex) repeated measures analyses produced a main effect for
composition £(1,122)=32.76, p < .01. More need was perceived during
interactions with the opposite-sex (M-2.73) than during interactions
with the same-sex (fl=2.25). The significant interaction (£(1,122)-6.86,
p < .01), was due to the sex difference found for same-sex interactions,
fi(1,122)=13.97, p < .01. Males (Ms1.89) perceived less need for support
on the part of same-sex cointeractants than did females (M-2.54).

The 2 (sex) X 2 (sex of friend) X 2 (closeness) repeated measures
analysis revealed a main effect for sex of friend, [(1,117) - 31.75, p <
.01. Subjects perceived more need of support when interacting with
opposite-sex friends (Ma2.68) than they did when interacting with the
same-sex friends (M-2.18). There was also a main effect for closeness
of friendship, £(1,117)=6.12, p < .05. Best friends (M-2.54) were
perceived as needing more support than second best friends (M-2.32).

The means for these analyses are presented in Table 5.
The results for perceived others’ need for support show that

females perceived more need than did males, more need was perceived on

54
the part of the opposite-sex, and males perceived less need on the part
of same-sex cointeractants than did females. The finding that more need
was perceived for the opposite-sex held across the analyses for best
friends and again, closeness of friendship distinguished between the
perceived need for support. Best friends were perceived as being more in

need than were second-best friends.

Table 5
Means and Standard Deviations for Perception of Qtnor’s Hood for Support

aa_a,Function of Interaction Category

 

 

Interaction Female Male
X 5.0 X 512
Same 2.54 1.09 1.89 0.78
Opposite 2.81 1.26 2.62 0.99
Same Best 2.50 1.29 1.88 0.93
Same Second 2.35 1.23 1.83 0.91
Best
Opposite Best 3.03 1.56 2.61 1.21
Opposite Second 2.54 1.44 2.50 1.10
Best

 

Support Provision and Peroaived Need for Support. Hypothesis 5

stated that females would be more responsive to the perceived needs of
others than would males such that there would be a stronger relationship
between support provided and the perceived need for support, for females
than for males. This hypothesis was not supported. Correlational

analyses showed that both sexes were equally responsive to the perceived

55
need of support in others (females r(69)-.78, p < .01; males r(53)-.81,
p < .01). Table 6 shows the correlates between support provision and
perceived need of support across the categories of interactions. As can
be seen in this table, composition of the interaction had minimal

effects on responsiveness.

Table 6
Rel tion hi betw en Su rt rovi i n and S r Ot Nee

 

 

Same Opposite S1 S2 01 02
Female .77 .80 .75 .78 .87 .83
Male .76 .74 .77 .78 .72 .57

 

Nota. SI - Same-sex best friend. 01 = Opposite-sex best friend. 52 =
Same-sex second-best friend. 02 - Opposite-sex second-best friend. For
females n - 70, for males n = 54. All p’s = .000

Two (sex of subject) X 2 (support: support provided, support other
needed) repeated measures analyses were conducted with support as the
within subject variable to determine the direction of the difference
between these two variables. The repeated measures were conducted for
sex composition (same-sex and opposite-sex) and best friends (same-sex
best friend, opposite-sex best friend, same-sex second-best friend, and
opposite-sex second-best friend). For example, the first repeated
measure analysis would have support provided to the same-sex and
perceived need for support for the same-sex as the two levels of the
{within subject variable. A main effect for support would indicate a
significant difference between support provision and perceived need for

support. Results showed that with the exception of the analysis for the

56

second best opposite-sex friend there was more support provided than was
perceived as being needed. The reader is referred to Tables 3 and 5 to
compare the means within each category of interaction.

i i n f r if lt f
Hypothesis 6 stated that there would no be relationship between support
provided and the difficulty of providing that support for females, but
there would be a positive relationship between these variables for
males. There was support for this hypothesis. The provision of support
and the difficulty of such provision was positively correlated for
males, r(53)=.39, p < .01. These variables were only marginally related
for females, r(69)-.18, p < .08. Table 7 shows the correlates for these
variables for the other categories of interactions. What is of interest
is the pattern that emerged between same and opposite-sex interactions.
Generally, it appears that both females and males had difficulty when
they supported same-sex interactants but difficulty was not related to

the support they gave to opposite-sex interactants.

 

 

Table 7
i n ' b w en or vi ed i It
Same Opposite $1 $2 01 02
Female .35 .16 .20 .27 .13 .19
Male .41 .14 .29 .48 -.19 -.03

 

Note. For females n . 70; the .05 level of significance
TS r - |.21|. For males n - 54; the .05 level of
significance is r . |.23|.

57
t 0 hi tw en r Re n 5

im n lit f t r . In support of hypotheses 7 a. and
7 b. there were significant positive relationships for both sexes
between the intimacy of the interaction and the support received
(females r(69)-.42, p < .01; males r(53)=.29, p < .05), and the intimacy
of the interaction and the support provided (females r(69)-.30, p < .01,
males r(53)=.25, p < .05). In both cases it can be seen (see Tables 8

and 9) that the relationships between support and intimacy were not as

Table 8
Relationghip betweon Support Received and Intimacy

 

Same Opposite $1 $2 01 OZ

 

 

 

 

Females .32 .61 .38 .54 .78 .57
Males .13 .35 .24 .44 .50 .36
Table 9
hi e ween S t vided a im c
Same Opposite $1 S2 01 02
Females .30 .46 .30 .44 .70 .38
Males .15 .25 .29 .17 .38 .34

 

Note. For females n . 70; the .05 level of significance
is r - |.21|. For males n - 54; the .05 level of
significance is r = |.23l.

strong for males as they were for females. Also apparent is that the

strongest relationship between these variables occurs for the best

58
friend of the opposite-sex, for both sexes. Further, the results
suggest that males’ same-sex cointeractants need to be close friends
before intimacy and support are related.

Hypothesis 7 c. stated that there would be a positive relationship
between the quality of interactions and support received for females,
and no relationship between these variables for males. In partial
support of this hypothesis, results showed positive relationships
between the quality of interactions and support received for ootn sexes
(females r(69)=.34, p < .01, males r(53)=.30, p < .01). A similar
pattern of results partially supported hypothesis 7 d. which stated that
there would be a positive relationship between the quality of
interactions and support provided for females, and no relationship
between these variables for males. Again positive relationships were
found for both sexes between quality and support provided (females
r(69)=.36, p < .01, males r(53)=.30, p < .05). The correlations between
these variables for the various categories of interactions are presented
in Tables 10-11. Overall then, for both females and males, quality,

intimacy, and support show moderate relationships with each other.

Table 10
Relationship between Quality and Support Received

 

Same Opposite $1 $2 01 OZ

 

Females .25 .53 .22 .33 .47 .47
Males .29 .37 .32 .05 .62 .39

 

59

 

 

Table 11
l on hi ween i n ort Provided
Same Opposite $1 SZ 01 02
Females .33 .47 .33 .24 .43 .54
Males .33 .35 .35 .00 .46 .47

 

Note. For females n - 70; the .05 level of significance
is r - |.21|. For males n - 54; the .05 level of
significance is r = |.23l.

Baoiprooity of Support. To test the final hypothesis, that support
exchange is characterized by reciprocity, the overall amount of support
received was correlated with the overall amount of support provided.

The results for both females (r(69)=.84, p < .01) and males (r(53)-.88,
p < .01) demonstrate that support exchange is reciprocal in nature.
Amount of support received, and amount provided was correlated for same
and opposite-sex interactions, and other interactions characterized by
degree of friendship. These correlates are shown in Table 12 and also

demonstrate a high degree of reciprocity.

Table 12
t n i be we Su rt Pr vi ed and Su t Re

 

 

Same Opposite S1 52 Ol 02
Females .77 .81 .66 .84 .85 .65
Males .80 .76 .82 .55 .82 .61

 

Note. For females n = 70, for males n . 54. All p’s < .01.

60

Repeated measures analyses conducted on support exchange (amount of
support received and amount provided) supported the correlational
analyses. The absence of a main effect on the within subject variable
(support exchange) would indicate reciprocity. There were main effects
for support exchange on interactions involving the opposite-sex
£(1,122)-7.74, p < .01, and the opposite-sex best friend £(1,122)-5.03,
p < .05. Both these main effects were due to support having been
received to a greater degree than it was provided.

Overall, the correlational analyses support Hypothesis 8,
demonstrating that support was characterized by a high degree of
reciprocity. In those cases where there was a difference between
support received and support provided, the imbalance was due to more
support having been received from the opposite sex than was provided to
them. The means for both variables can be compared by referring to
Tables 2 and 3.

l o i t i n ° i i n l l s
'f f r . No hypothesis was formulated
regarding the difficulty of providing support, and indeed the t-test
showed that females (M-2.22) did not differ from males (fi-2.34), p -
n.s., in the difficulty they reported with providing support.

The same series of repeated measures analyses was conducted on this
variable as was conducted on the other subjective variables. Few effects
were found. The 2 (sex) X 2 (sex of friend) X 2 (closeness) repeated
measures revealed a significant interaction between sex of friend and
sex of subject, £(1,82)=9.26, p < .05. It appeared that both sexes had

more difficulty providing support to their female friends (females

61
M-2.48; males M-2.66) than to their male friends (females M-2.24; males
M-2.30).

n ti . Overall, females (M=3.67) reported more
intimacy in their interactions than did males (M-3.09), t(122)=2.91, p <
.01. Repeated measures analyses [2 (sex) X 2 (composition: same-sex,
opposite-sex)] revealed a main effect for composition, E(1,122)-74.45, p
< .01, and an interaction with sex of subject, E(1,lZZ)-5.56, p < .05.
More intimacy was reported with opposite-sex interactants (M - 4.09)
than with same-sex interactants (M = 3.15). The significant interaction
was due to the greater sex difference found for same-sex interactions.
Males (M-2.70) rated these interactions as less intimate than did
females (M=3.50; £(l,122)=12.51, p < .01).

The 2 (sex) X 2 (sex of friend) X 2 (closeness) repeated measures
revealed a significant main effect for sex of friend f(1,117)-111.41, p
< .01. Interactions with the opposite-sex friends (M-4.08) were more
intimate than interactions with the same-sex friends (M-3.12). There
was also a marginal effect for closeness, £(1,117)-6.03, p < .06.
Interactions with best friends present (M=3.72) were rated as more
intimate than were interactions with the second-best friends (M-3.48).

These analyses also produced some significant interactions. The
interaction between sex of friend and sex of subject, E(l,117)-5.78, p <
.05, showed that males evidenced a greater difference in ratings of
interactions with opposite-sex friends and same-sex friends than did
females. The interaction between sex of friend and closeness,
£(1,117)=9.49, p < .05, showed that the effect for closeness was due to

ratings for the opposite-sex. Closeness did not affect the intimacy of

62
same-sex interactions. The means for these analyses are presented in
Table 13.

In summary, the results for intimacy show that females had more
intimate interactions than did males, opposite-sex interactions were
more intimate than were same-sex interactions, and the intimacy of
males’ interactions with friends were affected to a greater extent by
the sex of the friend than were the interactions of females. Finally,
closeness of friendship also affected the intimacy of interactions with
opposite-sex friends, but not same-sex friends. The closer the friend,

the more intimacy reported.

Table 13
Means and Standard Deviations for Intimacy of Interaotion as a Eunotion
of Intaraction Category

 

 

Interaction Female Male
3 SD X 5.0
Same 3.50 1.33 2.70 1.14
Opposite 4.20 1.41 3.93 1.02
Same Best 3.42 1.45 2.68 1.17
Same Second 3.37 1.50 2.82 1.32
Best
Opposite Best 4.41 1.56 4.13 1.42
Opposite Second 3.82 1.64 3.80 1.55
Best

 

Quality of Interactions. Overall, there was no difference between

females and males regarding the quality or pleasantness of their

63

interactions. However, the 2 (sex) X 2 (composition: same-sex, opposite-
sex) repeated measures analysis revealed a main effect for sex
composition. Opposite-sex interactions (Ms4.86) were rated as more
pleasant than same-sex interactions (M=4.53, £(1,122)-13.81, p < .01).

The 2 (sex) X 2 (sex of friend) X 2 (closeness) repeated measures
revealed a significant main effect for sex of friend £(1,117)-19.67, p <
.01. Interactions with the opposite-sex friends (My4.93) were more
pleasant than interactions with the same-sex friends (M-4.56). The
means for these analyses are presented in Table 14. Overall then, the
results for quality show that opposite-sex interactions were more

pleasurable than were same-sex interactions.

Table 14

Means and Standard Deviations for Quality of Interaction as a Eunotion
of Interaction Catagory

 

 

 

 

Interaction Female Male
X 5.0 X 5.0
Same 4.58 0.90 4.45 0.87
Opposite 4.86 1.35 4.86 0.90
Same Best 4.48 1.06 4.49 0.97
Same Second 4.68 1.08 4.59 1.04
Best
Opposite Best 5.07 1.28 4.90 1.08
Opposite Second 4.84 1.21 4.92 1.16
Best
rt Ne de n Su Received. The relationship between

support needed and support received was examined to determine if

64
individuals generally received the support that they reported needing.
The relationship between these variables was examined with correlational
analyses and repeated measures ANOVA’s [2 (sex) X 2 (support: support
needed, support received)]. For both males and females, overall receipt
and need was strongly correlated (females r(69)=.92, p < .01; males
r(53)-.88, p < .01). As the correlates in Table 15 demonstrate, there
was a strong relationship between the variables across all categories of
interaction indicating that individuals tended to receive the amount of
support they reported needing. However, the direction of differences
between support needed and support received indicated that less support
was being received than was needed. Significant main effects derived

from the repeated measures showed that imbalances between support

 

 

Table 15
i h' w e eived nd Needed
Same Opposite 51 S2 01 02
Females .93 .88 .87 .92 .86 .91
Males .84 .75 .91 .78 .80 .63

 

Moto. For females n . 70, for males n . 54. All p’s - .000

receipt and support need occurred during interactions with the same-sex
£(1,122)=5.51, p < .05, and with the same-sex best friend fi(1,122)=8.27,
p < .01. In both cases the main effect was due to less support having
been received than was needed. The means for these variables can be

compared by referring to Tables 2 and 4.

65
Suoport_Mooooo_ano_Support_Qtnor_Moooao. The relationship between
support needed and support other needed was examined to determine if
subjects tended to perceive the others’ need for support as similar to
their own. As can be seen in Table 16, the generally moderate to high
correlations suggest that one’s perception of one’s own need for support

and one’s perception of another’s need for support were positively

 

 

related.
Table 16
el i n hi etween Su ort e e nd u 0 r
Overall Same Opposite SI $2 01 02
Females .69 .60 .60 .49 .65 .65 .53
Males .69 .63 .42 .61 .65 .59 .55

 

Noto. For females n - 70, for males n = 54. All p’s < .05.

r vi e nd ther Need . The relationship
between support provided and support other needed was examined to
determine if subjects tended to respond to others at a level similar to

their own need for support. The high correlations presented in Table 17

 

 

Table 17
twe n u rt Nee nd 0 t r e
Overall Same Opposite $1 $2 01 02
Females .81 .75 .77 .55 .76 .79 .64
Males .78 .75 .58 .76 .75 .65 .39

 

Nota. For females n = 70, for males n = 54. All p’s < .05.

66

show that one’s provision of support was related to one’s own need for
support.

R- 1 . 51'. betw-:n . 19' . ,... .,. '- o1. . The final
set of additional analyses concerned the relationship between the
various personality measures and variability in support. The research
question posed was whether personality was significantly related to

support variability. Table 18 shows the correlations between the

 

 

 

Table 18
Re ti nshi be w en ari bilitv of SupoortIand_Eersonaljty_£or_£emalas
Personality
Empathy Esens Econt Eexpr Ssens Scont Sexpr
SDSSR .12 .07 -.05 .14 -.11 .20 .21
SDSSP .12 .20 -.10 .20 -.10 .31 .30
SDSSN .01 .06 .06 .03 -.O3 .14 .15
SDSON .13 .14 -.06 .24 .03 .18 .20
SDDIFF -.06 .17 .13 .02 -.02 .08 .07

 

Nota. For females n = 70; the .05 level of significance

is r - |.21|. SDSSRsstandard deviation for support receipt.
SDSSP-support provision. SDSSNasupport need. SDSON-support other need.
SDDIFF-support difficulty. Esens-emotional sensitivity. Econt-emotional
control. Eexpraemotional expressivity. Ssens=social sensitivity.
Scont-social control. Sexpr-social expressivity.

various personality measures and the standard deviations for the various
support measures, for females. Table 19 shows the corresponding
correlations for males. As can be seen in these tables, personality did

not appear to be related to variability in the supportive behaviors and

67
perceptions of females. However, these variables were positively

related for males.

 

 

Table 19
R' 1 '01 110 19 e-1 Va 49 0 .!!'r .00 " 0 a 0 u: ‘
Personality

Empathy Esens Econt Eexpr Ssens Scont Sexpr
SDSSR .30 .34 .22 .21 .13 .40 .35
SDSSP .26 .36 .33 .20 .13 .36 .32
SDSSN .46 .35 .07 .09 .34 .11 .18
SDSON .23 .40 .25 .09 .00 .29 .20
SDDIFF .34 .26 -.14 .03 .20 .12 .09

 

Moto. For males n = 54; the .05 level of significance is r - [.23].
See Table 18 for nomenclature of variables.

Parsonality Variables as Maoiators of Support Measuras

The $51 is a new measure of social skills, however, the subscale
means, intercorrelations, and Cronbach’s alphas for this sample compare
very well with those reported by Riggio (1986). The subscale
intercorrelations ranged from -.47 to .83. The means and Cronbach’s
alphas for each of the subscales ranged from 4.82 - 6.29, and .72 - .87,
respectively. The mean rating for the Empathy scale was 6.19. The
standardized Cronbach’s alpha was .85.

To examine the effects of social skills and empathy on the overall
support measures stepwise multiple regression analyses were performed
for all subjects. Because the focus of the study was on sex

differences, multiple regressions were calculated separately for females

68
and males, with the different support measures as the criterion, and
with sex of subject and the subscales of the two personality measures as
predictors. Interactive terms were also entered as predictors (e.g.,
sex X empathy).
The results for the overall regression analyses are presented in

Table 20. Emotional sensitivity predicted the amount of support

 

 

Table 20
l i l e r ion n l s ' t e
Criterion Predictors Beta F-value R R2
Provide Emotsens .29 11.06** .29 .08
Receive Emotsens .24 7.28** .24 .06
Sex -.18 5.80** .30 .09
Need Sex -.29 10.62** .29 .08

Difficulty No significant predictors
Other need Sex -.25 08.06** .25 .06

Moto. *p< .05, **p< .01.

 

received and provided. The more emotional sensitive an individual was
the more support they provided and received. Sex was also a predictor
of the receipt of support; being female was associated with receiving
support. Sex was the only predictor of the amount of support needed and
the amount of support perceived as being needed by others. Again, being
female was associated with needing support and perceiving that need in
others. There were no significant predictors of the difficulty of

providing support. Overall then, being female was the primary predictor

69
of support indices with emotional sensitivity being the stronger
predictor of support receipt and provision.

When the regressions were performed separately for the sexes an
interesting pattern emerged. For females, emotional expressivity
predicted the receipt of support: the more emotionally expressive the
more support females reported receiving. Empathy predicted the
difficulty of providing support: greater empathy was associated with
less difficulty in providing support. There were no significant
predictors for the provision of support, the need for support, or the
perception of others’ need. The results of these analyses can be seen

in Table 21.

 

 

Table 21

ol- Re-re i- An. e . ' -. .... 1;. - - . 'n< -
Criterion Predictors Beta F-value R R2
Provide No significant predictors
Receive Emotexpr .27 5.22* .27 .07
Need No significant predictors
Difficulty Empathy -.24 4.07* .24 .06

Other need No significant predictors

Motg. * p < .05, ** p < .01.

 

As can be seen in Table 22 emotional sensitivity was the only
significant predictor of all the support measures for males. In each
case the relationship between the criterion and the predictor was

positive.

7O

 

 

Table 22

1. pl- 'e-r- sion A .l e . P -.' .opor u-t - - . .1 :
Criterion Predictors Beta F-value R R2
Provide Emotsens .40 9.68** .40 .16
Receive Emotsens .31 5.19* .31 .09
Need Emotsens .32 5.59* .32 .10
Difficulty Emotsens .44 11.87** .44 .19
Other need Emotsens .32 5.59* .32 .10

 

Moto. * p < .05, ** p < .01.

It should be noted that there were sex differences on the various
personality measures. Females were more empathetic (t(120)-6.04, p <
.01; females M - 6.50, males M - 5.72); more emotionally expressive
(t(120)-1.91, p < .06; females M - 5.62, males M - 5.28); more
emotionally sensitive (t(120)-3.05, p < .01; females M - 6.51, males M =
5.99); and more socially sensitive (t(120)=3.65, p < .01; females M -
6.42, males M - 5.69). Males (M a 5.29) were more emotionally
controlled than females (M - 4.51), t(120)=4.48, p . .01. There was no
sex difference for social expressiveness (females M = 5.95, males M -
5.87) or social control (females M = 5.63, males M . 6.01).

It did not appear that restriction of range attenuated the
correlations between personality and support, as females and males did
not differ on variability on these measures. That is, the differences

between the sexes on the standard deviations of the personality measures

were small.

 

71
ti w o i e M

Items from the Personal Reaction Form were totaled and divided by
the number of days the mood checklist was maintained. Thus each score
represented the average daily rating for that particular feeling. 5
Factor analyses by the principal components method followed by varimax
rotation revealed two factors, labelled positive and negative affect,
accounting for 50% and 23% of the variance, respectively (the other
factors had eigenvalues less than one). The first factor had all 13
positive affect items loading above .67 (range .68 - .89, eigenvalue =
9.44) and none of the negative affect items loading above .56. The
negative affect factor had all 6 items loading above .62 (range .62 -
.89, eigenvalue = 4.45) with none of the positive affect items loading
above .46.

Because the principal variables of interest pertained to support,
items indicative of support or helpfulness were included on the mood
measure. These items loaded on the positive affect factor (range .75 -
.89). However, conceptually they were considered to be distinct from
positive affect and were labeled collectively as the altruism factor. A
reliability analysis was then conducted on the positive affect, negative
affect, and altruism factors. See Appendix B for the items contained in
each of the three factors.

Cronbach’s alpha for the 9 items on the positive affect factor was
.95 (M . 3.24). The standardized Cronbach’s alpha for the 6 items on
the negative affect factor was .68 (M = 2.62). The standardized

Cronbach’s alpha for the 4 items on the altruism factor was .95 (M =

3.08).

.g—a»

72

The intercorrelations between the three factors showed that, for
females, positive and negative affect were independent, r(68) - .10, p =
n.s. Positive affect and altruism were strongly related, r(68)-.69, p <
.01, with negative affect and altruism showing a moderate relationship
with each other, r(68) - .36, p < .01. For males, positive and negative
affect were moderately related, r(52) . .31, p < .05. Positive affect
and altruism were strongly related, r(52) . .83, p < .01, as was
negative affect and altruism, r(52) - .54, p < .01. For both sexes
then, these feelings were positively related to each other. Females and
males differed only on the negative affect factor t(1,120) - 2.72, p <
.01. Females reported more negative affect (M - 2.80) than did males (M
= 2.37).

The three factors were correlated with the various support measures
separately for overall support, same-sex support, and opposite-sex
support. The correlates for these analyses are presented in Tables 23
through 25. It can be seen that across the various categories of
interaction the three mood factors were moderately related to most of
the support measures. Because these relationships were all positive it
was suspected that intensity of feelings might be mediating these
relationships, such that individuals high on feeling intensity might
simply be rating themselves high on all moods and support measures, and
those low on feeling intensity might be rating themselves low on all
moods and support measures. Auxiliary data collected from Larsen’s
Affect Intensity Measure (AIM, 1984) provided a measure of emotional
intensity. The correlational analyses were repeated with emotional
intensity partialled out. As the partial correlates in Tables 19-21

indicate, emotional intensity had minimal effects on the relationships

73
between the variables. Further, positive affect was partialled out of
subsequent correlations with negative affect, and negative affect was
partialled out of subsequent correlations with positive affect. Again,
the partials had minimal effects on the pattern of correlates. Although
this pattern is quite strong across the different categories of
interaction, there are two salient deviations from the pattern. First,
if one looks at the correlates of need for support, it is quite clear
that needing support was associated with feeling good for females.
There was no relationship between these variables for males. The second
salient deviation from the pattern of results is the pattern for the
difficulty of providing support. Feelings had no relationship with
difficulty for females. For males, it appeared that difficulty was
associated with feeling bad, however, emotional intensity mediated these
relationships. When emotional intensity was taken into account there

was generally no relationship between difficulty and feelings.

Table 23

 

 

 

r la io betw e v ll Su r Me
Females Males
Support POSAFF NEGAFF ALTRUISM POSAFF NEGAFF ALTRUISM
Receipt .48 .48 .51 .35 .49 .39
(.43) (.45) (.47) (.33) (.44) (.35)
Provision .40 .37 .50 .47 .51 .52
(.41) (.37) (.50) (.44) (.42) (.45)
Need .34 .60 .43 .17 .50 .29
(.29) (.58) (.40) (.15) (.48) (.27)
Other Need .27 .34 .33 .36 .37 .36
(.27) (.34) (.33) (.35) (.33) (.33)
Difficulty -.06 .10 -.06 .17 .35 .15
(-.06) (.10) (-.06) (.00) (.18) (-.05)
Moto. For females n - 69 and r=|.20|, p < .05. For males

n - 53 and r-|.23|, p < .05.
the corresponding correlates controlling for emotional

Values in parentheses are

intensity: for females n - 66, for males n . 48.

Table 24

 

 

 

r i e ween
Females Males
Support POSAFF NEGAFF ALTRUISM POSAFF NEGAFF ALTRUISM
Receipt .43 .54 .49 .24 .36 .27
(.38) (.52) (.46) (.30) (.31) (.30)
Provision .35 .33 .45 .42 .42 .47
(.38) (.34) (.46) (.43) (.32) (.46)
Need .30 .60 .41 .13 .34 .18
(.26) (.58) (.38) (.16) (.30) (.20)
Other Need .21 .29 .26 .27 .29 .30
(.21) (.29) (.27) (.27) (.19) (.28)
Difficulty .09 .06 .00 .26 .36 .25
(.12) (.08) (.02) (.16) (.26) (.11)
Note. For females n - 69 and r-|.20|, p < .05. For males

n - 53 and r=|.23|, p < .05.
the corresponding correlates controlling for emotional

Values in parentheses are

intensity: for females n - 66, for males n - 45.

Table 25
e t ion be w--1

 

Females

Males

 

 

 

Support POSAFF NEGAFF ALTRUISM POSAFF NEGAFF ALTRUISM
Receipt .38 .44 .32 .35 .29
(.34) (.41) (.31) (.33) (.26)
Provision .30 .46 .50 .45 .47
(.29) (.46) (.47) (.38) (.39)
Need .49 .39 .17 .45 .29
(.46) (.35) (.14) (.41) (.23)
Other Need .17 .31 .44 .31 .31
(.05) (.31) (.44) (.29) (.30)
Difficulty .09 .05 -.06 .19 .04
(.08) (.04) (-.17) (-.02) (-.13)
Mots. For females n - 67 and r-|.20|, p < .05. For males

n - 53 and r-|.23|, p < .05.
the corresponding correlates controlling for emotional

Values in parentheses are

intensity: for females n = 63, for males n - 48.

CHAPTER IV
DISCUSSION

It was argued that researchers need to conceptualize social support
as a component of social interaction. Such a conceptualization makes
imperative the investigation of sex differences and the norms of
reciprocity. The significant patterns of results for both these
variables (to be discussed in more detail later) support the use of such
a conceptualization. Putting social support into this framework guides
the direction of research and delineates the boundaries of support in
its relationships to other variables. As Rook points out, the nature of
support has continued to expand, thus masking the essence of what the
variable is (Rook, 1983, cited in Shumaker & Brownell, 1984). The
significant correlations of intimacy and quality with support indicate
that these variables are related, but the magnitude of the correlates
demonstrate that they are not the same variables. Further, the mean
ratings of support fall somewhere between two and three on a seven point
rating scale. These ratings suggest that support is obviously a
component of relevance to social interaction but it is not an overriding
one. Neither is it as focal as other components, such as quality and
intimacy (see also Forgas, 1976). Thus social support is distinct form
other components of social interaction, and these components are just as

relevant, if not more so, to an interaction.

77

78
Sax uifforances in Socia1_Support

The first major pattern of findings to be discussed are the overall
sex differences in support exchange. The results showed support for all
hypotheses regarding sex differences. Females received and provided
more support than did males. They also needed more support and saw that
need in others to a greater extent than did males. The finding that
females received more support than did males is in agreement with other
research (Burda, Vaux, & Schill, 1984; Butler, Giordano, & Neren, 1985;
Caldwell, Pearson, & Chin, 1987; Dunkel-Schetter, Folkman, a Lazarus,
1987; Stokes & Wilson, 1984; see also Sarason, Sarason, Hacker, &
Basham, 1985) as is the finding that they needed more support than did
males (Rosenthal et al, 1986).

The finding of greater female provision of support substantiates
the conceptual arguments of Belle (1982), Bernard (1981), Cohler and
Lieberman (1980), Fischer (1982), and Vaux (1985). These authors argued
that females are more interpersonally focused and responsive than males,
and assume a stereotypical presentation of being the confidente in
interpersonal situations. This result does not agree with Eagly’s
(1987) finding that males are the more helpful sex. However, as she
notes, her review concerned behaviors that males were more comfortable
about performing than were females (e.g. intervening in a dangerous
situation such as a potential drowning incident). The present study
provides a nice complement to Eagly’s (1987) review in that here, the
help investigated was presumably more comfortable for females to provide

than it was for males to provide (e.g. interpersonal comfort).

79

Why do females provide more support than do males? One possibility
is that females do not actually provide more support than males but
perceive that they do. Another possibility is that they do provide more
support than males because they perceived a greater need for support
than do males, and these variables were strongly related. Surprisingly
the regression analyses showed that neither empathy nor social skills
predicted this perception of others’ need for females. It is possible
that females are more skilled than are males at perceiving others’ need
and the results are indicative of a skill other than those measured in
the present study. It is also possible that females, having a greater
desire or motivation to be communal, may want to perceive a need for, or
provide, support in order to be responsive and execute their communal
orientation towards others. It is important to note that there is no
measure of accuracy of perception in these data. We do not know how
much the other interactant(s) needed support but simply that females
perceived the need to be greater than did males.

The second question about these results is why did females report
receiving more support than did males? Females may simply perceive that
they receive more support than do males. If females actually did receive
more support than did males, the more pertinent question is--how?
Receiving support is not a behavior but rather the elicitation of
behavior from another. The only significant predictor of support
receipt for females was emotional expressiveness (a sample item being "I
show my emotions or feelings"). In other words--if females want support
they engage in behaviors conducive to receiving it. This concords with

previous research showing that females do indeed ask for more support

80
(Butler et al, 1985) and self-disclose (e.g. Cozby, 1973) more than do
males.

The reason that females request more support (as shown in previous
research) appears to be due to the finding that they need more support
than did males. Females need for more support is not predicted by any
of the personality measures. Perhaps their greater need for support is
due to their greater communal nature, in comparison with males. Thus,
females may need the caring and support of others because of their
orientation towards interrelatedness. On the other hand, females
reported a greater level of negative affect than did males and negative
affect was strongly associated with needing support. If females were
more depressed/unhappy than were males, then they should have needed the
good feelings of others to a greater extent than did males.

Reciprocity of Social Support

One of the main focuses of this research was to examine the
reciprocity of support exchange. In agreement with social exchange
theorists (e.g. Homans, 1958; Walster, Berscheid, & Walster, 1973) and
in concordance with hypothesis eight support exchange was characterized
by reciprocity. The overall amount of support that individuals reported
receiving was strongly correlated with the overall amount of support
they reported providing. This result is not surprising and extends the
previous research on support reciprocity. That research (Ingersoll-
Dayton & Antonucci, 1983; McFarlane et al, 1981; Tolsdorf, 1976) showed
that rolationships tended to be reciprocal, i.e. the people one helps
tend to help one in return. The present results showed that the rate of

support exchanged during interactions is perceived as being highly

81
reciprocal, that is, the more one helps another, the more they return
the help. Results specifically addressing the perceived amount of
support exchanged showed that where there was an imbalance between the
receipt and provision of support it was due to more support being
received than was provided. Moreover, this imbalance only occurred
during interactions with the opposite-sex, for both sexes.

There are a number of explanations for this finding. The first is
that there is a reporting bias such that both sexes like to portray
themselves as the receiver rather than the provider, but only to the
opposite-sex. This is not what one would expect from a socially
desirable bias in reporting. Further, other results indicated that
subjects reported needing more support than they received, but only from
the same-sex. If these subjects are responding with a socially
desirable bias then they are doing so in a strange and unsystematic way.

Another potential explanation is that there is no bias and that the
females are reporting about a different group of males than those who
participated, and likewise for the males (clearly very different
groups)! It was the observation of the investigator that a number of
subjects attended the study sessions accompanied by the opposite-sex.

It seems quite likely that these individuals appeared in each others
records which does not add credence to this explanation.

The results point more towards a perceptual bias. Obviously, one
can never measure the actual amount of support an individual receives or
provides because it is an entirely subjective judgment, but both sexes
show some desire or motivation to perceive themselves as receiving more

from the opposite sex then they provide in return. Perhaps the best

82
explanation is to draw on Clark and Mills (1979) distinction between
exchange and communal relationships. In a nutshell, Clark and Mills
(1979) argue that perceptions of reciprocity depend on the type of
relationship one is engaged in, or desires to be engaged in. In a
communal relationship benefits and costs are not part of an exchange
process. If this is an issue, then the relationship is specifically one
of exchange. The defining feature of a communal relationship is that
members are concerned about the welfare of the other and respond to the
others’ need, regardless of previous exchange. A family is the most
typical example of a communal relationship, or relationships that fall
under the rubric of the model of altruistic love. Clark and Mills
(1979) found that attraction for a partner decreased when a benefit was
directly reciprocated and a communal relationship was desired,
suggesting that individuals do not want "tit for tat" reciprocity when
the development of such a relationship is their intention.

If one accepts that freshmen are in the process of heterosocial
orientation then the opposite sex may be considered a potential pool of
communal relationships. One may be motivated, at this age and in this
social milieu, to perceive exchanges with opposite-sex interactants as
communal and not exchange-based. One is also motivated to believe that
the other is responding to the one’s needs unselfishly (cf Kelley, 1983)
hence the perception that one receives more support from the opposite-
sex than one provides to them. Other results support this argument. On
the other support measures--provision, receipt, need, and other’s need--
the opposite-sex elicited higher ratings than did the same-sex. That

is, more support was provided to the opposite-sex and received from

83
them, than from the same-sex. More support was needed from them and
they were perceived as needing more support than were the same-sex.
Recall that support was defined as the expression of caring, love,
esteem, or respect for another. By definition, it appeared that
subjects believed that their interactions with the opposite-sex were
characterized by caring to a greater degree than their interactions with
the same-sex.

Closeness of friendship was also a factor in the results,
indicating that the more time one spent with another, the more
supportive those interactions were perceived to be. Most likely the
results are due to knowing a person better and understanding their needs
more, and also turning to them to satisfy one’s own needs. Some results
showed that closeness interacted with sex of friend, such that closeness
of the friendship distinguished only between opposite-sex friends. This
implies that same—sex friends achieve a certain level of support
exchange but go no further, perhaps because the romantic or communal
element is not present to distinguish these relationships on that level
of closeness.

One final note is in order regarding the exchange of support as it
depends on the sex of the participants in the interaction. House and
Kahn assert that "obviously, the issues of reciprocity and sex
composition are related in that women may benefit more from
relationships with other women not because such relationships are less
demanding, but because they are more reciprocal" (1985, p. 93). These
data support House and Kahn. Results showed that same-sex interactions

were more reciprocal than opposite-sex interactions. However, the

84
effects hold for both sexes. Males’ interactions with other males were
also more reciprocal than their interactions with females. Of course,
how one defines oonotit and whether reciprocity is more beneficial than

non-reciprocity depends on tghe relationship and what is sought by its

members.
Ralationships between Supportiye Behaviors

Intercorrelations between the subscales of the RIR generally
supported the hypotheses and produced some surprising results. It was
hypothesized that females would be more responsive than would males, in
terms of providing support where they perceived a need for it. In fact
both sexes were very responsive to others’ need for support and
regardless of interaction sex composition, subjects perceived that they
provided more support than was needed. It is not clear why subjects
would provide more support than was needed unless one agrees with Batson
and his colleagues (1988) that individuals do act to the benefit of
others, and perhaps there is a belief that when expressing support, if
one simply matches the amount of support the other needs then the
expression of support is not genuine. This idea is somewhat consistent
with Blau’s notion that ”social exchange then, is an intermediate case
between pure calculation of advantage and pure expression of love”
(1964, p. 112).

Of interest is the finding that subjects did not receive the
support they needed from the same-sex. And this included interactions
with their same-sex best friend. It was not the case that subjects
needed more support from same-sex friends than they needed from

opposite-sex friends. Quite the opposite. But perhaps they oxpacted

85
more support from same-sex friends in terms of understanding their
feelings about relationships and class or whatever, and judged their
receipt of support against that standard. In support of this
explanation is the finding that subjects reported difficulty with
providing support to same-sex others. There was a hypothesized sex
difference here in that overall, there would be no relationship between
these variables for females but that males would have difficulty with
providing support. This hypothesis was supported. Nevertheless, for
both sexes, difficulty was associated with provision of support to the
same-sex but was unrelated to the provision of support to the opposite-
sex. Perhaps this reported difficulty is due to an expectation that one
should be supportive to the same-sex and such support may be seen in
light of expectation and exchange, whereas, supporting the opposite-sex
is seen as more indicative of genuine caring and not expectation.
Further, individuals spend a great deal more time in the company of the
same-sex. Thus more same-sex interactions may be inescapable than may
opposite-sex interactions, for example at bedtime or early morning, etc.
Interactions with the opposite-sex are more likely to be freely chosen,
for example dates or lunch-breaks, etc. Indeed the higher quality
reported for opposite-sex interactions would support this argument.
This argument is in line with Thibaut and Kelley’s (1959) assertion that
dyadic interaction is facilitated by interactants being able to provide
each other with high rewards (support) at low cost (difficulty) to
themselves. There may be a ”vicious cycle" operating, for individuals
at this stage of development, such that if expectation and exchange

norms cease to operate, one freely gives certain social resources

86
without difficulty. Because of the lack of difficulty one becomes more
attached and because of the attachment one gives more freely and so on
it goes.
im i

Much has already been said about the intimacy and quality of same-
sex and opposite-sex interactions and relationships, and sex differences
therein (e.g Blyth G Foster-Clark, 1987; Clark G Reis, 1988; Davis,
1978; Fischer G Narus, 1981; Lewis, 1978; Nezlek, Wheeler, G Reis,
1983A; Reis, Sanchak, G Soloman, 1985; Reis G Shaver, 1988; Shields,
1987). Those findings and arguments will not be reiterated here, except
to note that the present results replicate very well those of previous
research, with females reporting more intimacy in interaction than did
males, and the male-male interaction being lower on intimacy than were
other sex compositions. The purpose of including the measures of
intimacy and quality in the present research was to examine their
relationship with support. It was hypothesized that the exchange of
support would be related to feelings of intimacy, for both sexes. It is
not surprising that such feelings would be related to each other. The
pattern of correlates suggest that these feelings tended to be more
closely related for females and for interactions with best friends of
the opposite sex. However, a larger sample size is needed to test these
differences for significance. A tentative conclusion is that the
relationship between intimacy and support may be somewhat dependent on
an interpersonal orientation to interaction, such that closeness and
expression of feelings are related. It is probable that a more task-

focused orientation to interaction, as is characteristic of males

87
(Aries, 1976; Sullivan, 1986), would show that perceived closeness may
be related to factors other than the expression of feelings, such as a
ball-game with others.

It was expected that the quality or pleasantness of interactions
would not be related to support for males because of their greater focus
on task-oriented interaction. It was found that quality and support
were similarly related for males and females indicating that regardless
of sex, supportive exchanges are pleasant events. Overall then, quality
and intimacy are moderately to strongly related with support depending
on the sex composition of the interactions.

P r n u r

Sex was the better predictor of supportive behavior that the
personality measures used in this study. Emotional sensitivity was the
only predictor of the provision of support. Regardless of sex, the more
emotionally sensitive a person was the more support they provided. The
sex difference on provision of support then, is apparently due in part
to the greater emotional sensitivity of females, as compared with males.

Both sex and emotional sensitivity were predictive of support
receipt with emotional sensitivity being the better predictor. Sex did
account for additional variance suggesting that there is something about
being female above and beyond the characteristic of emotional
sensitivity that is predictive of support receipt. The regression
results for females suggest that this characteristic is emotional
expressiveness. It appears that sensitive people elicit support from
others, but others are also more likely to respond if they interact with

someone who clearly expresses their feelings. The pattern of results

88

for males is quite interesting and demonstrates that supportive behavior
in males is characterized by emotional sensitivity. A tentative
conclusion is that the emotional interactions of males are dependent on
their emotional make-up with their social orientation playing a minor
role. The emotional interactions of females appear to depend on
characteristics other than those used in this study. The correlational
analyses conducted on the standard deviations of support measures and
the personality variables enrichened the regression results. Recall
that there was no relationship between these variables for females.
Taken together, the two sets of results indicate that the amount and
variability of females’ supportive behaviors and perceptions are
generally not due to empathy and social skills. Female support then, is
either more dependent on the social situation or personality variables
other than social skills. Male support, on the other hand, appears to
be due, at least in part, to personality variables such as empathy and
social skills.
Affaot and Support

Before discussing the relationships between affect and social
support a couple of issues need to be addressed. These are the
interrelations between the different dimensions of affect and sex
differences on these dimensions. Females differed from males on the
negative affect factor, reporting a greater level of negative affect.
Diener and Emmons (1985) and Watson, Clark, and Teliegren (1988) report
no sex differences in their studies, although Diener (1984) reports
elsewhere that women report more negative and positive affect than do

males but the sex difference is never great.

89

There is considerable debate in the field on the relationship
between positive and negative affect. That debate will not be discussed
here, except to note the work of Diener and Emmons (1985) which is most
comparable to the present method. They report between-subject
correlations of positive and negative affect that were recorded on a
daily basis for six weeks. It is important to recognize that between-
subject correlations refer to the correlations computed for mean
positive affect and mean negative affect summed across the recording
period. Their results showed that the correlates (-.23 to .26) did not
reach significance indicating independence of the two kinds of affect.
Recall that in the present study these variable were related for males,
however, the size of the correlate corresponds well with the range found
by Diener and Emmons (1985) and the larger sample size of the present
study may have created the significant effect. Interestingly, altruism
and negative affect were related for both sexes. This is surprising
because one would expect that feeling helpful would not make one feel
bad, unless one considers the alternative--feeling bad makes one feel
helpful (e.g. Cialdini G Kenrick, 1976). More than likely, subjects who
felt bad helped others in order to alleviate their distress and
subsequently rated themselves as altruistic (Cialdini G Kenrick, 1976).

The hypothesized relationships between mood and social support were
in general supported. The findings did not support Clark and Watson
(1988) or Stone (1987) who showed that negative affect was not related
to social engagement. The measures in the present study more
specifically measured social support. Perhaps the difference in

measurement accounts for this discrepancy. Overall, the results showed

90
a very consistent pattern of positive correlations. One could attempt
to explain each result individually, but a more parsimonious explanation
is in order given the consistent pattern. By definition, support
exchange is an emotional experience and it is unlikely that such
behavior is considered to be affectively neutral. Thus it really is not
surprising that emotion expressed during social interactions was related
to emotion felt at the end of the day. Another way to view the results
is to consider supportive interactions as an exchange of feelings, thus
one engages in supportive interactions when one is feeling "up” or
"down” or is engaged by others who feel emotional. One must also keep
in mind that both measures (support and affect) were indices averaged
over a two-week period so these are representative summaries of
individuals’ level of affect and support as opposed to precise temporal
relationships. A more temporal-specific analysis may show different
relationships between these variables.

The positive correlations held constant across interaction
composition, thus, sex of cointeractant appeared to have minimal effects
on the relationship between the variables. Even though the pattern was
quite consistent there were two noteworthy differences. For males,
positive affect and need for support were not related. This finding may
be due to males’ greater orientation towards autonomy. Needing support
is incompatible with independence and it follows that such need would
not be related to feeling good for males. One may argue that it should
be negatively correlated. However, it is a very human quality to need
others and be affiliative, so for males needing support might sometimes

make them feel good and sometimes not.

91

The other deviation from the pattern of results was the generally
null relationship for difficulty and affect. Difficulty is perhaps seen
as a skill more so than an emotion and is a skill irrelevant to the
general emotional character of an interaction. Given that interactions
are not comprised of support alone then any difficulty with support
provision is not necessarily a social skills deficit and thus is
affectively neutral.

Overall, the findings suggest that experienced emotions are related
to emotional interactions, regardless of one’s sex or level of emotional
intensity, regardless of the emotional components of the interaction,
and regardless of the valence of the emotion.

Conclusions

The main objectives in this research were to examine sex
differences in the exchange of emotional support and to examine the
reciprocity of support exchange. These objectives came about by
conceptualizing social support as a component of social interaction. As
the results clearly show, both sex and reciprocity are major factors in
the exchange of support. These findings are of interest to research
psychologists and students of social interaction. However, these
results take on a greater significance when one considers how they
should direct future research.

Much substantive work has been done on the relationship between
social interaction/social support and well-being (see reviews by Cohen G
Wills, 1985; and Levy, 1983). The argument here is that our research
paradigms must be broadened in scope if we are to fully understand the

complex relationship between these variables. The research has shown

92
that the receipt of support is positively related to well-being (e.g.
Cohen G Wills, 1985), however, some (e.g. Antonucci, 1985) have
cautioned about the costs of receiving help and the costs associated
with other aspects of social interaction, such as intimacy (Clark G
Reis, 1988; Hatfield, 1984). Moreover, a recent article in Psychology
Today (Luks, 1988) discusses the benefits of helping others, while
others have discussed the costs of helping others (Kessler, McLeod, G
Wethington, 1985). The costs of social interaction and the provision of
support have been neglected areas in research on both social interaction
and social support. The present results showed that provision of
support is just as relevant to a social interaction as is the receipt of
support. Moreover, there were many sex differences in the present
findings, and relationships between social interaction measures and some
indices of well-being are differentiated by sex (e.g. Sarason, Shearin,
Pierce, G Sarason, 1987). In sum, sex and exchange are essential
factors in social interaction and support and they should be included as
factors in future research to more clearly understand how social
interaction affects well-being. For example, females have consistently
been shown to receive more support than have males, and support has
consistently been shown to be positively related to well-being, yet
females are less mentally healthy than are males, especially married
females (e.g. Gove, 1973; Radlofff, 1975; Striegel-Moore, Silberstein, G
Rodin, 1986; Weissman G Klerman, 1977). Longitudinal work or cross-
sectional work may explain how health and support are related, and how
this relationship may differ for males and females. For instance, it

may be the case that older females, who function in a different social

93
network than the present subjects, provide more support than they
receive (to children and aged parents) and the excess provision of
support is associated with poorer mental health.

The present work can be used to stimulate more applied work, as
mentioned above, and clearly some experimental laboratory work needs to
be conducted to examine the potential perceptual and expectancy issues
which arose in these results. Obviously the present work has some
shortcomings, the greatest of which is the sample. The results apply to
young incoming freshmen and may not be comparable with research
investigating the social exchange among older settled individuals, given
that the social network would undergo dramatic changes over time. This
research requested subjects to describe instances of supportive
interactions. As a result we do not know what specific behaviors are
contained in those interactions. Also, the data were corrected for the
number of interactions recorded, and it was these average measures that
were correlated with each other and other variables. A different
approach, such as calculating day by day correlations between social
interaction and mood would provide more temporal-specific results and
may increase our knowledge of how these variables are related in other
ways.

The present project produced one hundred and fifty-four folders
describing students social lives. The knowledge contained therein
stands to challenge social support researchers to consider sex and

exchange as factors affecting the results of their work.

LIST OF REFERENCES

LIST OF REFERENCES

Antonucci, T. C. (1985). Social support: Theoretical advances, recent
findings, and pressing issues. In I. G. Sarason, G B. R. Sarason
(Eds.). Social S ° T R r h A Ti i
Martinus Nijhoff pub., Dordrecht, Boston, Lancaster.

Aries, E. (1976). Interaction patterns and themes of male, female, and
mixed groups. Small Group Sahayior, 52, 155-162.

Aries, E. J. (1982). Verbal and nonverbal behavior in single-sex and
mixed—sex groups: Are traditional sex roles changing? Esyonologioal
Reports. 51. 127-134.

Bakan, D. (1966). h li o h m xisten . Chicago: Rand
McNally.

Bales, R. F. (1950). Interaction process analysis: Arnethod for the
study of small groups. Reading, Mass: Addison-Wesley.

Bales, R. F. (1970). Personality and interpersonal behavior.
New York: Holt, Rinehart and Winston.

Barnett, M. A., Howard, J. A., King, L. M. G Dino, G. A. (1981).
Helping behavior and the transfer of empathy. Journal of
Sooial Psychology, 115, 125-132.

Barrera, M. (1986). Distinctions between social support concepts,

measures, and models. Anarioan Journallof Community Psychology,
15, 413-445.

Barrera, M., Sandler, I. N., G Ramsey, T. B. (1981). Preliminary
development of a scale of social support: Studies on college

students. American Journal of Community Psychology, 1981, S,
435-447.

Batson, C. D., Dyck, J. L., Randall Brandt, J., Batson, J. G., Powell,
A. L., McMaster, M. R., G Griffitt, C. (1988). Five studies testing
two new egoistic alternatives to the empathy-altruism hypothesis.
Journal of Personalityrand Social Psychology, SS, 52-77.

Bell, R. R. (1981). Worlds of friendship. Beverly Hills, CA: Sage.

94

95

Belle, D. (1982). The stress of caring: Women as providers of social
support. In L. Goldberger G S. Breznitz (Eds.), Manoooor_ot_stress:

Theoretical and clinical aspects (pp. 496-505). New York: Free
Press.

Berg, J. H. (1984). Development of friendship between roommates.
Journal of Personality ano_Sooial Psychology, i§(2). 346-356.

Berg, J. H., G Clark, M. S. (1986). Differences in social exchange
between intimate and other relationships: Gradually evolving or
quickly apparent? In V. J. Derlega and B. A. Winstead (Eds.)

Friendship and social interaotion. New York: Springer-Verlag.

Berg, J. H., G McQuinn, R. D. (1986). Attraction and exchange
in continuing and noncontinuing dating relationships. Journal

f P on lit an i . 59(5). 942-952.

Berg, J. H., Peplau, L. A. (1982). Loneliness: The relationship

of self-disclosure and androgyny. Personality and Sooial
Psychology Bulletin, S, 624-630.

 

Bernard, J. (1981). The female world. New York: MacMillan.

Berscheid, E., G Peplau, L. A. (1983). The emerging science of
relationships. In H. H. Kelley, A. Christensen, H. H. Harvey,
T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, and
D. R. Peterson (Eds.) Close relationships. New York:
W. H. Freeman and Company.

Blau, P. M. (1964). Exohange and power in social life. New York:
John Wiley G Sons.

Blyth, D. A., G Foster-Clark, F. S. (1987). Gender differences in
perceived intimacy with different members of adolescents’ social
networks. Sex Roles, 11, 689-718.

Bolton, C. D., G Kammeyer, K. C. W. (1967). The university stuoent.

Conn: University Press.

Booth, A., G Hess, E. (1974). Cross-sex friendship. Journal of
Marriageland the Family, 2, 38-47.

Brookings, J. 8., G Bolton, 8. (1988). Confirmatory factor analyses of
the interpersonal support evaluation list. American Journal of
Community Psyohology, I§(l), 137-147.

Burda, P. C. Jr., Vaux, A., Schill, T. (1984). Social support
resources: Variations across sex and sex role. Perso it and

Sooial Psychology Bulletin, 19(1), 119—126.

96

Butler, T. Giordano, S. G Neren, S. (1985). Gender and sex-role
attributes as predictors of utilization of natural support systems
during personal stress events. Sex Roles, 1S(9/10), 515-524.

Caldwell, R. A., Pearson, J. L., G Chin, R. J. (1987). Stress-
moderating effects: Social support in the context of gender and

locus of control. P rs n ' 1 Ps h l . lfiill,
5-17.

Carlson, M., Charlin, U., G Miller, N. (1988). Positive mood and

helping behavior: A test of six hypotheses. Journal of Personality
and Sooial Psychology, SS, 211-229.

Cialdini, R. B., G Kenrick, D. T. (1976). Altruism as hedonism: A
social development perspective on the relationship of negative mood

state and helping. Journal of Personality and Social_Psyonology, S1,
907—914.

Clark, L. A., G Watson, 0. (1988). Mood and the mundane: Relations
between daily life events and self-reported mood. Journal of

Personality and Social Psychology, 55, 290-308.

Clark, M. S., G Mills, J. (1979). Interpersonal attraction in
exchange and communal relationships. Journal of Personality and
Sooial Psychology, 31, 12-24.

Clark, M. S., G Reis, H. T. (1988). Interpersonal processes in close
relationships. Annual Review of Psychology, SS, 609-672.

Cobb, S. (1976). Social Support as a moderator of life stress.
Psychosomatic Medicine, SS, 300-314.

Cobb, S. (1979). Social support and health through the life course

(pp 93-106). Aqing from birth to death: Interdisciplinary
perspectives. In M. L. Riley (Ed.). Washington, DC: American
Association for the Advancement of Science.

Cohen, C. I., G Sokolovsky, J. (1979). Health-seeking behavior and
social networks of the aged living in single-room occupancy hotels.
Journal of the Americaaneriatrios Society, 21, 270-278.

Cohen, S., Sherrod, D. R., G Clark, M. S. (1986). Social skills and the
stress-protective role of social support. Journal of Personality ano
Social Psychology, SQ(5), 963-973.

Cohen, S., G Syme, S. L. (1985) (Eds.). Sooial support and health.
Orlando, FL: Academic Press.

Cohen, S., G Wills, T. A. (1985). Stress, social support, G
the buffering hypothesis. Psychological Bulletin, 2S, 310-357.

97

Cohler, B., G Lieberman, M. (1980). Social relationships and mental

health among three European ethnic groups. Besearon_on_Aging, 2,
445-469.

Cohn, N. B., G Strassberg, D. S. (1983). Self-disclosure reciprocity
among pre- adolescents. P r t nd i l P ch ll '
9, 97- 102.

Coyne, J. C. (1976). Depression and the response of others. Journal of
Abnormal Psychology, SS, 186-193.

Cozby, P. C. (1973). Self—disclosure: A literature. Psyonologioal
Sulletin, ZS, 73-91.

Cunningham, J, G Strassberg, D. S. (1981). Neuroticism and disclosure

reciprocity. Journal of Qoun seli 1ng Psyohology, ZS, 455-458.
Cutrona, C. E. (1986). Behavioral manifestations of social support.
A microanalytic investigation. rn l f P r n l i

Ps yohology, 51(1), 201- 208.

Dabbs, J. M., G Ruback, R. B. (1984). Vocal patterns in male and female
groups. Personality ano Social Psychology Bulletin, 19, 518-525.

Davis, J. D. (1978). When boy meets girl: Sex roles and the
negotiation of intimacy in an acquaintance exercise. Journal of
Personality and Social Psychology, SS, 684-692.

 

 

Deaux, K. (1977). Sex differences in social behavior. In T. Blass

(Ed.). Personality variables in sooial behavior. (pp. 357-378).
Hillsdale, N.J.: Erlbaum.

Diener, E. (1984). Subjective well-being. Psychologioal Sulletin,
SS, 542-575.

Diener, E., G Emmons, R. A. (1985). The independence of positive and

ggggtive affect. Journal of Personality and Social Psyonology, 51,
-1117.

Depner, C. E., Wethington, E., G Ingersoll-Dayton, B. (1984). Social
support: Methodological issues in design and measurement. Journal
of Socialllssues, 40(4), 37-54.

Dunkel-Schetter, C., Folkman, S., G Lazarus, R. S. (1987). Correlates
of social support receipt. Journal of Personality and Social
Psyohology, SS(1), 71-80.

Eagly, A. H. (1987). Sex differences in social behavior: A

social-role interpretation. Hillsdale, N.J.: Lawrence Erlbaum.

Eisenberg, N. G Lennon, R. (1983). Sex differences in empathy and
related capacities. Psychological Bulletin, 25, 100-131.

98

Emmons, R. A. (1986). Personal strivings: An approach to personality

and subjective well-being. Journal of Personality and Sooial
Psychology, SI, 1058-1068.

Fischer, C. S. (1982). To dwell among frienos; Personal netuorks
in town and oity (pp. 125-138). Chicago: University of Chicago
Press.

Fischer, J. L., G Norus, L. R. (1981). Sex roles and intimacy in

same-sex and other sex relationships. Psyono1ogy_otrflomen_uuarterly,
S, 444-455.

Forgas, J.P. (1976). The perception of social episodes: Categorical
and dimensional representations in two different social milieus.

Journal of Personality and Sooial Psychology, SA, 199-209.

Fultz, J., Batson, C. D., Fortenbach, V. A., McCarthy, P., G Varney,
L. L. (1986). Social evaluation and the empathy altruism hypothesis.

Journal of Personality and Sooial Psychology, SQ, 761-769.

Gilligan, C. (1982). In a different voice: Women’s conceptions of self
and morality. Harvard Eduoational Review, 51, 481- 517.

Goplerud, E. N. (1980). Social support and stress during the first year
of graduate school. Professional Psychology, A, 283-290.

Gottlieb, B. H. (1978). The development and application of a
classification scheme of informal helping behaviors. ganaoian Journal
of Behavioral Soience, 19, 105-115.

Gottlieb, B. H. (1985). Social support and community mental health. In
S. Cohen, G S. L. Syme, (Eds.). Sooial Support and Health (pp.303

-325). Orlando, Florida: Academic Press.

Gove, W. R. (1973). Sex, marital status, and mortality. American
Journal of Sociology, 12, 45-67.

Grigsby, J. P., G Weatherby, D. (1983). Gender and sex-role differences
in intimacy of self-disclosure. Psychological Reports, SS, 891-897.

Hall, D. T. (1969). The impact of peer interaction during an academic
role transition. Sooiology of Eduoation, AZ, 118-140.

Hatfield, E. (1984). The dangers of intimacy. In V. J. Derlega (Ed.).

tommunioation. intimacy. and close relationships. Orlando, FL:
Academic Press.

Heitzmann, C. A., G Kaplan, R. M. (1988). Assessment of methods for
measuring social support. Health Psychology, 1(1), 75-109.

99

Heller, K., G Lakey, B. (1985). Perceived support and social
interaction among friends and confidents. In I. G. Sarason, G B. R.
Sarason (Eds.). Social Support: Theory, Pesearon ano Applioations.
Martinus Nijhoff pub., Dordrecht, Boston, Lancaster. (pp. 287-302).

Homans, G. (1958). Social behavior as exchange. e rn of
, SS, 597-606.

House, J. S. (1981). r c or . Reading, MA:
Addison-Wesley.

House, J. S., G Kahn, R. L. (1985). Measures and concepts of social
support. In S. Cohen G S. L. Syme (Eds.) Soo1a1_Support_ano_Mealtn
(pp 83-108). NY: Academic Press.

Ickes, W. (1983). A basic paradigm for the study of unstructured
dyadic interaction. In H. T. Reis (Ed.) Naturalistic approaohes to
studying social interaotion. New Direotions for Methodology of
Social and Behavioral Soience, No. 15. isan Francisco: Jossey-Bass.

Ingersoll-Dayton, B., G Antonucci, T. C. (1983). Reciprocal and
non-reciprocal social support: Contrasting sides of intimate
relationships. Paper presented at the Gerontological Society of
America Meetings, San Fransisco, CA.

Isen, A. M. G Levin, P. F. (1972). Effect of feeling good on helping:

Cookies and kindness. o of P rs lit n o i l c o o ,
11, 384-388.
Jourard, S. M. (1979). Self-disolosure: An experimental analysis of

the transparent self. Huntington, NY: Krieger.

Kelley, H. H. (1983). Love and commitment. In H. H. Kelley,
A. Christensen, H. H. Harvey, T. L. Huston, G. Levinger, E.
McClintock, L. A. Peplau, and D. R. Peterson (Eds.) Slose
relationships. New York: W. H. Freeman and Company.

Kenrick, D. T., Baumann, D. J., G Cialdini, R. B. (1979). A step in the
socialization of altruism as hedonism: Effects of negative mood on
children’s generosity under public and private conditions. Journal
of Personality and Social Psychology, S1, 747-755.

Kessler, R. C., McLeod, J. D., G Wethington, E. (1985). The costs of
caring: A perspective on the relationship between sex and
psychological distress. In I. G. Sarason, G B. R. Sarason (Eds.).

Social Support: Theory, Research and Applications. Martinus Nijhoff
pub., Dordrecht, Boston, Lancaster.

Komarovsky, M. (1976). Dilemma’s of masculinity: A study of college
you h. New York: Worton.

100

Kopfstein, J. H., G Kopfstein, D. (1973). Correlates of self-disclosure
in college students. rn l i lo ,
A1, 163.

Kowalski, C. J. (1982). College dropouts: Some research findings.
Psyehology, A Quarterly Journal of Human Sehavior, 12, 45-49.

Krebs, D. (1975). Empathy and altruism.
ano Social Psychology, S1(6), 1134-1146.

Larsen, R. J. (1984). Theory and measurement of affect intensity as an

individual difference characteristic. Sissertation_Aostraots
International, SA, 22112, Ann Arbor, Michigan.

Leavy, R. C. (1983). Social support and psychological disorder: A
review. Journal of Community Psychology, 11, 3-21.

Levy, R. L. (1983). Social support and compliance: A selective
review and critique of treatment integrity and outcome measurement.
Social Science Medicine, 11(18), 1329-1338.

Lewis, R. A. (1978). Emotional intimacy among men. Journal of Sooial
Issues, SA, 108-121.

Lin, N., Simeone, R. L., Ensel, W. M., G Kuo, W. (1979). Social
support, stressful life events and illness: A model and an empirical

test. Journal of Health and Sooial Behayior, 29, 108-119.
Luks, A. (1988, October). Helper’s high. Psychology Today. Pp. 39-61.

McCormick, I. A., Siegert, R. J., G Walkey, F. H. (1987). Dimensions of
social support: A factorial confirmation. American Journal of
Community Psychology, 1S(1), 73-77.

McFarlane, A. H., Neale, K. A., Norman, G. R., Roy, R. G., G Streiner,
D. L. (1981). Methodological issues in developing a scale to measure

social support. Sonizopnrenia Sulletin, 1(1), 90-100.

Mehrabian, A., G Epstein, N. (1972). A measure of emotional empathy.
Journal of Personality, AS, 525-543.

 

 

Morton, T. L. (1978). Intimacy and reciprocity of exchange: A
comparison of spouses and strangers. Journal of Personality
and Social Psychology, SS(1), 72-81.

Nezlek, J. B., G Wheeler, L. B. (1984). RIRAP: Rochester interaction

Eggggd analysis package. Psychological Documents, 1A, 6. (Ms. No.

101
Nezlek, J. B., Wheeler, L. G Reis, H. T. (1983AL Studies in social

participation. In H. T. Reis (Ed. ) Naturalistio appr roaohes to
stuoy1n no social interaotion. w i ti for

WW
Social and Behavioral Science, No. 1S, (pp. 57- -75), San Francisco.

Jossey- -Bass.

Nezlek, J. 8., Wheeler, L. G cReis, H. T. (19838). SMQSTor_hootuorm:

 

The rela ationship b w m for
Unpublished manuscript, College of William and Mary, Virginia.
Parelman, A. (1983). Emotional intimaoy in marr riage: A_sex;ro1es

perspeotive. Ann Arbor, MI: h University of Michigan
Research Press.

Parsons, T. (1951). Th 0 ' l tem. New York: Free Press.

Parsons, T., G Bales, R. F. (1955). Family, socialization and
interaction process. Glencoe, IL: Free Press.

Peplau, L. A., Russell, D., G Heim, M. (1979). The experience of
loneliness. In I. H. Frieze, D. Bar- Tal, G J. S. Carroll (Eds. ),

New approaches to social problems: Applioations of attribution
theory (pp. 53-78). San Francisco, CA. Jossey- Bass.

Piliavin, J. A., G Unger, R. K. (1985). The helpful but helpless
female: Myth or reality. In V. E. O’Leary, G R. K. Unger, Wallston,
B. S. nomencagender ano_soo1a1_psyoho1ogy. Hillsdale, N.J.: Lawrence

Erlbaum Association.

 

 

Radloff. L. (1975). Sex differences in depression: The effects of
occupation and marital status. Ser roles, 1, 249-265.

Reis, H. T. (1984). Social interaction and well-being. In S. W. Duck

(Eds.), Personal Relationships S: RepairingPersonal Relationships.
New York: Academic Press.

 

Reis, H. T. (1986). Gender effects in social participation: Intimacy,
loneliness and the conduct of social interaction. In Ihe_£merg1ng
Field of Personal Relationships, ed. R. Gilmour, S. Duck, pp. 91-105.
Hillsdale, NJ: Erlbaum.

Reis, H. T., Senchak, M., G Solomon, B. (1985). Sex differences in the
intimacy of social interaction: further exploration of potential

ggglaggt;ons. Journal of Personality and Social Psychology, AS,

Reis, H. T., Shaver, P. (1988). Intimacy as an interpersonal process.
In Handbook of Personal Relationships: Theory. Relationships and
Interventions, ed. S. Duck,. Chichester: Wiley.

102

Reis, H. T., Wheeler, L., Kernis, M. H., Speigel, N., G Nezlek, J.
(1985). On specificity in the impact of social participation on

physical and psychological health. Journal of Personality ano Sooial
Psychology, AS, 456-471.

Reis, H. T., Wheeler, L. Spiegel, N. Kernis, M. H, Nezlek, G J.
Perri, M. (1982). Physical attractiveness in social interactions 11.

Why does appearance affect social experience? Journal of Personality
' l Ps chol . AS, 979-996. -

Riggio, R. E. (1986). Assessment of basic social skills. Journal of
Per rsonai ty an no Sooial Psyohology, .S1(3), 649- 660.

Riggio, R. E., G Throckmorton, B. (1986). Manual for the Sooial Skills
Inventory (SSI). Unpublished manuscript. California State College,
Fullerton.

Rook, K. S. (1984). Research on social support, loneliness, and social

isolation. In Shaver P. (Ed.). Review of Social ano Personality
Psychology, Vol 5. Beverly Hills, CA: Sage.

 

Rook, K. S. (1987). Social support versus companionship. Effects on
life stress, lonliness, and evaluations by others. Journal of

Personalitynand Socia1_Psyoho1ogy, S1(6), 1132- 1147.

Rose, S. M. (1985). Same- and cross-sex friendships and the psychology
of homosociality. Sex Poles, 12, 63-75.

Rosenthal, K. R. Gesten, E. L. G Shiffman, S. (1986L Gender and sex
role differences in the perception of social support. Sey_3o1es,
1A(9/10), 481- 699.

Sanford, N. (Ed. L (1962). [he Amer rioan oollege: A_psyoho1ogtoa1_ano
soo ial interpretat tion of higher learning. New York: Wiley and Sons.

Sarason, I. G., Levine, H. M., Basham, R. B., G Sarason, B. R. (1983).
Assessing social support: The social support questionnaire. Journal
of Personality and Sooial Psyohology, 44(1), 127-139.

Sarason, 1. G. G Sarason, B. R. (1985). Social Support: Theory,
Research an nd Applications. Martinus Nijhoff pub., Dordrecht,
Boston, Lancaster.

 

 

Sarason, B. R. Sarason, I. G, Hacker, T. A. G Basham, R. B. (1985).
Concomitants of social support. Social skills, physical
attractiveness, and gender. Journal of Personality an no Social
Psyohology, 52(2), 469-480.

Sarason, I. G, Sarason, B. R. Potter, E. HH G Antoni, M. H. (1985).

Life events, social support, and illness. Psychosomatio Meoicine,
A1(2), 156- 163.

103

Sarason, I. G., Sarason, B. R., Shearin, E. N., G Pierce, G. R. (In
Press). A brief measure of social support: Practical and

theoretical implications. Journal of Sooial and Personal
Relationships.

Sarason, B. R., Shearin, E. N., Pierce, G. R., G Sarason, I. G. (1987).
Interrelations of social support measures: Theoretical and practical
implications. Journal of Personality and Soojal Psyghulggl, 52(4),
813-832.

Schaller, M., G Cialdini, R. B. (1988). The economics of empathic
helping: Support for a mood management motive. Journal of

Experimental Socjal Psyonology, 2A, 163-181.
Shields, S. A. (1987). Women, men, and the dilemma of emotion. In P.
Shaver, G C. Hendrick (Eds.). v w P 1

Psychology, Vol 17, Sex and genoer. Beverly Hills: Sage.

Shinn, M., Lehmann, S., G Wong, N. (1984). Social interaction and
social support. Journal of Social Issues, AQ(4), 55-76.

Shumaker, S. A., G Brownell, A. (1984). Toward a theory of social

support: Closing conceptual gaps. Journal of Soojal Issues,
AQ(4), 11-36.

Spence, J. T., Deaux, K., G Helmreich, R. (1985). Sex roles in
contemporary American society. In G. Lindzey, G E. Aronson,

(Eds.). he handbo k f s i l 5 ch lo 3r yol, 2. New
York: Random House.

Stokes, J. P., G Wilson, 0. G. (1984). The Inventory of Socially
Supportive Behaviors: Dimensionality, prediction, and gender

differences. Ameriean Journal of Sonnunjty Psygthng. 12(1),
53-69.

Stone, A. A. (1987). Event content in a daily survey is differentially
associated with concurrent mood. o r f P r i cial

Esxcbelegx. 52(1). 55-58.

Striegel-Moore, R. H., Silberstein, L. R., G Rodin, J. (1986). Towards

an understanding of risk factors for bulimia. American Psychologist,
Al, 246-263.

Sullivan, L. (1986). Sex differences in the relationship between
academic performance, social behavior, and planned activities.

Unpublished Master’s Thesis. The College of William and Mary,
Virginia.

Tardy, C. H. (1985). Social support measurement. American Journal of
Community Psychology, 1;, 182-202.

104

Thibaut, J. W., G Kelley, H. H. (1959). Ine_soo1al_osyenology_of
groups. New York: John Wiley G Sons.

Thoits, P. A. (1985). Social support and psychological well-being:
Theoretical possibilities. In I. G. Sarason, G B. R. Sarason (Eds.).
Social S ort: Theor rch a d Applioations. Martinus Nijhoff
pub., Dordrecht, Boston, Lancaster.

Tolsdorf, C. C. (1976). Social networks, support and coping. family
process, IS, 407-418.

Vaux, A. (1985). Variations in social support associated with gender,

ethnicity, and age. Journal of Social Issues, A1(1), 89-110.

Walster, E., Berscheid, E., G Walster, G. W. (1973). New directions in
equity research. ourn f P on li ' 1
151-176.

Watson, D., Clark, L. A., G Telleger, A. (1988). Development and
validation of brief measures of positive and negative affect: The

PANAS scales. Journal of Personality and Social_Psyonolooy, SA,
1063-1070.

Weissman, M. M. G Klerman, G. L. (1977). Sex differences and the
epidemiology of depression. Archives of General Psychiatry, SA,
98-111.

Wheeler, L. (1988). My year in Hong Kong: Some observations about

social behavior. Personality and Social Psyonology_flullet1n, 1A,
410-420.

Wheeler, L., G Nezlek, J. B. (1977). Sex differences in social

participation. Journal of Personality and Soeial Psyonology, SS,
742-754.

Wheeler, L., Reis, H., G Nezlek, J. (1983). Loneliness, social
interaction, and sex roles. n P

Psyghology, A5, 943-953.

Wilcox, B. L., G Vernberg, E. M. (1985). Conceptual and theoretical
dilemma’s facing social support reseaarchers. In I. G. Sarason,
G B. R. Sarason (Eds.). Social Support: TheoryLsResearch ano
Applications. Martinus Nijhoff pub., Dordrecht, Boston,
Lancaster.(pp. 3-20).

Winstead, B. A. (1986). Sex differences in same-sex friendships. In V.
J. Derlega, G B. A. Winstead (Eds.). Friendship and social
interaction. NY: Springer-Verlag

Winstead, B. A. G Derlega, V. J. (1986). Friendship and social
interaction: An introduction. In V. J. Derlega and B. A. Winstead

(Eds.) Friendship and social interaction. New York: Springer
-Verlag. (pp. 1-7).

APPENDICES

105

APPENDIX A

Rochester Interaction Record (RIR)

 

Date: Day: Time:
Length: hrs mins
Initials:

Sex:

 

Support needed

or desired: none
Support received: none
Intimacy: not at all
Quality: unpleasant
Support provided: none
Difficulty: very easy
Support other
needed/desired: none
Nature: Work/Study

l

Socializing

am

pm_

If more than 3 others:

NNNNNN
000003000300
##«bcfih-fi
UIU'IUIU'IU'IU'I
030105050305

2 3 4 5 6
Relaxing

#

=&

7

Life necessity

of females ___

of males

a lot
a lot
very intimate
very pleasant
a lot

very difficult

a lot
Other

106

APPENDIX 8

Personal Experience Form

Date: Day:

 

 

Please fill in the circle for the number that best represents how you
are feeling right now on the scale next to each item.

* Happy 1 2 3 4 5 6 7
# Depressed l 2 3 4 5 6 7
1 - Not at all * Joyful l 2 3 4 5 6 7
2 . Very slight # Unhappy 1 2 3 4 5 6 7
3 . Somewhat * Pleased 1 2 3 4 5 6 7
4 = Moderate amount * Aroused/Active 1 2 3 4 5 6 7
5 . Much Sociable/Friendly l 2 3 4 5 6 7
6 - Very much @ Helpful 1 2 3 4 5 6 7
7 = Extremely much @ Kind 1 2 3 4 5 6 7
Selfish l 2 3 4 5 6 7
Lonely 1 2 3 4 5 6 7
* Dominant l 2 3 4 5 6 7
* Enjoyment/Fun 1 2 3 4 5 6 7
# Frustrated 1 2 3 4 5 6 7
* Satisfied l 2 3 4 5 6 7
# Worried/Anxious l 2 3 4 5 6 7

* Anticipating/
Looking Forward tol 2 3 4 5 6 7

# Angry/Hostile l 2 3 4 5 6 7
* Self-confident 1 2 3 4 5 6 7

um.

107

APPENDIX
Dependent
@ Giving
@ Sympathetic
# Needy
Intimate

* . Positive affect items.
@ . Altruistic items.

8 (CONT’)
1234567
1234567
1234567
1234567
1234567

# . Negative affect items.

108
APPENDIX C

Instruction Booklet
Dear Student and Co-investigator:

Surprisingly, psychologists know very little about the nature
of social activity. The purpose of this study is to uncover some
facts about this area. One way to find out about people’s social
lives would be to sit down and ask them a series of questions about
how active they are, who they see, etc. However, we don’t believe
that anyone could give a very accurate answer to those types of
questions; it is simply too difficult to remember everything at
once.

Therefore, in order to find out something about how people
spend time with other people we have devised a simple record
keeping technique, a diary of sorts. By obtaining information via
this technique here and at other places in the country and abroad
(so far there have been four other studies like this in the 0.5.,
one in Canada, and one in Hong Kong), we hope to discover something
about social needs and behaviors.

We must emphasize that the study can succeed only with your
cooperation. The only other way to obtain the type of information
this study will provide would be to follow a group of people 24
hours a day and write down all they did and then ask them questions
about how they felt about what they did. Not to mention the
impracticality of this approach, it is not hard to imagine the
troubles this would causeil This is why we addressed you both as a
student and a co-investigator because we need you to investigate
social interaction with us. You are a co-investigator in the sense
that you are the creator, observer, and recorder of the data we
need. This is a very unusual way to collect psychological

information, but with your honesty and full cooperation, we know it
will work.

So What go We Mean by Social Interaction?

By social interaction we mean a situation in which two or
more people are responding to one another. A conversation is the
clearest example of an interaction. Person A says something,
Person 8 responds, Person A responds to that response, and so
forth. Dancing and love-making are also interactions. In other
words you do not have to speak to the other person(s) to interact
with them. Sitting side-by-side and watching television is not an
interaction. Listening to a lecture is not an interaction, even if
you occasionally ask a question. We include the reading and
writing of letters as instances of interactions.

109

APPENDIX C (CONT’)

Based on previous research, we have found that we can get
accurate and informative data about people’s social lives by asking
them to describe the social contacts they have had each day which
last ten minutes or longer. Ten minutes is a good minimum for it
gives us a reasonable summary of the meaningful social events in a
person’s day and does not cause the record keeping to be
particularly difficult or troublesome. If we were to ask people to
record every time they spoke with anybody their records would be
cluttered with a good deal of nonsense and also probably be
somewhat inaccurate.

Let’s take a difficult example of a social interaction.
Suppose you are dining at the cafeteria in the presence of a group
of people. You are listening to their conversation but seldom say
anything yourself. Do you record this as a social interaction? We
think it is a social interaction if you are following the
conversation and you could enter into it if you wished. If you are
not following the conversation or if it would be inappropriate for
you to enter it, it would not be counted as a social interaction.

Now let’s take some more difficult examples. For instance,
you are with a friend Debbie for 25 minutes, walking around campus.
You are joined by a friend John who talks to you both for 15
minutes and then leaves. You and Debbie continue walking and
talking for another 20 minutes. This sequence of events should be
recorded as three separate interactions - one with Debbie for 25
minutes; a second with Debbie and John for 15 minutes and a third
with Debbie for 20 minutes. If you had not met John you probably
would have recorded the event as one interaction of an hour in
length. Another example might involve you and a friend Nike who go
out to eat together and then decide to go and play some pool. 0n
the way home you decide to stop by the ice-cream store and spend
half an hour in there. If the quality of your interaction with
Mike did not change over the course of the evening then you may
record this event as one interaction. 0n the other hand, if Mike
was upset about something and discussed this with you over dinner,
but cheered up by the time you started playing pool, you may record
two separate interactions. Moreover, if the
quality of the interaction changed when you went to get ice-cream,
you may record three interactions.

We should point out that you don’t have to know the people
you are with to have a social interaction. As an example, if you
talk to a stranger at a bus stop for ten minutes or more, that
would be a social interaction.

110
APPENDIX C (CONT’)

At this point you have probably gathered that we have set no
hard and fast rules about what is, and is not, an interaction and
when an interaction changes substance and when it doesn’t. We are
interested in your impressions of your lives and therefore how you
describe it is precisely the way we want it described. Many
ambiguities will arise as you record what goes on around you.
Strive to be objective, but when in doubt follow your own
impressions, however subjective you think they may be.

At the end of the record keeping period we’ll request some
additional background information. All the information that you
provide us will be held in the strictest confidence. At no point
in time will your name ever be used in a data analysis. We ask for
your name only so that we can more easily sort your data and keep
things organized. All the data will be analyzed anonymously via
the computer. At any time during the study you may ask us any
questions you desire and we will attempt to answer them as
completely as we can. At the completion of the study you will be
given the option of receiving a copy of the results if you so wish,
however, in keeping with our policy of confidentiality, we cannot
release individual results.

Thank you for your cooperation.

If any difficulties or questions arise you can call any time.
Linda Sullivan 353-3935 (office) 339-9775 (home)
436 Baker Hall

Experimental assistant:

GUIDELINES FOR SOCIAL INTERACTION

We would like you to record every interaction of ten minutes
or more in length, for a period of two weeks. An interaction is
defined as any situation involving 2 or more people in which the
behavior of each person is in response to the behavior of the other
person. A conversation is the clearest example of an interaction,
but merely being in the presence of another, like watching TV and
not talking to or touching the person next to you, is not an
interaction. Whether you are on campus or not, whatever you are
doing, keep the record. Make sure to record any phone calls
lasting longer than 10 minutes. You do not have to be face-to-face
to interact with someone. The more consistent and reliable your
recording is the more valid our inferences about the data become.
It is most important that you keep the record everyday, all the
ti e. The only exception to this is class time. Do not record
your class meetings as interactions.

111

APPENDIX C (CONT’)

Participants in other studies have found it useful to update
the record in the afternoon, sometime in the evening, and before
they retire. Regardless, you need to update your record every day.
The entire study depends on your cooperation in keeping these
records. Even if you feel that a certain day was completely
routine with nothing out of the ordinary, record your interactions.
If you have lunch with the same people every day, record it every
day. If you forget to complete your records on a certain day,
don’t try to remember, simply skip that day.

We understand, however, that there may be times when you
cannot carry the forms with you or you may not have the opportunity
to fill out the interaction records immediately. In this case, we
have designed a single ”scratch" sheet for you to note the initials
of the people with whom you have interacted. The sheet will serve
as a memory jog when you get the chance to fill in the detailed log
later.

Do not be dissuaded by the lengthy description of the diary -
the diary itself takes approximately 1 minute to complete. We have
included the following guidelines in case any questions arise.

Read them over, make sure you understand them and go to it!!

DATE: Always record the month and day (e.g., Nov 12 - 11/12) when
the interaction occurred. Do not put down the date only on every
other interaction or only on the first interaction of the day.

DAY: Write down the day of the week.

TIME: Write down the time the interaction started and check a.m.
or p.m.

LENGTH: Record how long the interaction lasted in hours and
minutes.

INITIALS: Record the initials of the other people in the
interaction. If two people have the same initials distinguish them
with a middle initial (if you know it) or the second letter of
their last name. As an example, if you list Jack Kramer as J.K.
then James Kennedy would be J.K.E. The most important thing is to
be consistent. Once you describe James Kennedy as JKE always do
that, otherwise we will have no idea whom certain initials
represent. If you do not know a certain person’s name, put two
question marks for their initials. If you know one initial but not
the other, put down the one you know and put a question mark for
the other. If you happen to find out a person’s name after you
have represented them with question marks, change the "?'s. Be
certain to change only those question marks which you are certain
represent the person whose name you just found out.

 

 

112
APPENDIX C (CONT’)

SEX: In the space provided on the record sheet, write F (for
female) or M (for male) to describe the sex of the person(s) with
whom you interacted, under their initials.

IF MORE THAN 3 OTHERS: There are only spaces for three sets of
initials. If you are interacting with more than three others, do
not put down any initials. Instead, indicate the number of males
and the number of females in the appropriate blank. Do not, of
course, count yourself in this measure. If you are unsure of the
number of pe0ple, or if it varied, give the best estimate you can.

GUIDELINES FOR THE PERSONAL EXPERIENCE FORM

Before you retire for the evening put the day and date on the
Personal Experiences Form and then use the rating scale to report
how you are feeling right at that moment.

For all of the following, circle only one number.

SUPPORT NEEDED OR DESIRED: Circle a number to indicate the extent
to which you needed or desired to feel loved, accepted, or cared
about: or the extent to which you needed or desired to feel that
you were valued, trusted, respected, or held in esteem. This could
be conceptualized as needing or desiring to feel good or feel
better (if you feel down). The need refers to the time of the
interaction not a feeling that you had in retrospect.

SUPPORT RECEIVED: Circle a number to indicate the extent to which
you received support as defined above.

INTIMACY: Circle a number to indicate the degree of closeness you
felt in the interaction. Intimacy may be physical or sexual, but
it does not have to be. For example, circle a 1 if the interaction
was not at all close e.g., you chatted about the weather. Circle a
7 if the interaction was very close e.g you spoke about something
very personal. Circle a 4 if the interaction was close, e.g.,
talking to your roommate about a family matter.

QUALITY: Circle a number to indicate how pleasant the interaction
was. Circle a 1 if the interaction was unpleasant or unenjoyable.

Circle a 2 if the interaction was slightly unpleasant.

Circle a 3 if the interaction was pleasant.

Circle a 4 if the interaction was between pleasant and quite
pleasant.

Circle a 5 if the interaction was quite pleasant.

Circle a 6 if the interaction was between quite pleasant and very
pleasant.

Circle a 7 if the interaction was very pleasant.

113

APPENDIX C (CONT’)

SUPPORT PROVIDED: Circle a number to represent the extent to which
you provided support, as defined above, to the other(s).

DIFFICULTY: If you provided support to the other(s) circle a
number to indicate the extent to which you felt uncomfortable,
tense, or unconfident in any way, in providing support. If you did
not provide any support then leave this item blank.

Circle a 1 if it was very easy.

Circle a 2 if it was somewhat easy.

Circle a 3 if it was slightly easy.

Circle a 4 if it was neither easy nor difficult.
Circle a 5 if it was slightly difficult.

Circle a 6 if it was somewhat difficult.

Circle a 7 if it was very difficult.

SUPPORT OTHER NEEDED OR DESIRED: Circle a number to indicate the
extent to which you felt the other(s) needed support, as defined
above.

NATURE: Circle a word that indicates the primary nature of what
you did during the interaction. Circle one and only one word.
Some interactions are difficult to classify, but try to base your
decision on the predominate theme of the interaction, and be
consistent.

I. WORK/STUDY: any obligatory activity for which you are paid or
activity such as student volunteer on committees, and study time.

2. SOCIALIZING: any activity in which the main focus is to be
with other people such as talking or having dinner together.

3. RELAXATION: any activity in which the main focus is on doing a
task, such as playing sports or spending time at a hobby.

4. LIFE NECESSITY: any activity that you have to do. For

instance going to pick up mail, running errands, eating lunch or
dinner and so on.

5. OTHER: none of the above.

114

APPENDIX D

Verbatim Script

The purpose of this meeting is to describe what your
participation will involve should you decide to take part in this
study. The study is designed to uncover some facts about people’s
social lives. Surprisingly, psychologists know very little about
the nature of social interaction. You might be wondering what we
mean by social interaction. A social interaction is any situation
in which two or more people are responding to one another. A
conversation is the clearest example of an interaction. Person A
says something, Person B responds, Person A responds to that
response and so on. Sitting side by side watching TV is not an
interaction, nor is listening to a lecture.

In order to gain a greater understanding of social
interaction we could approach the topic from a few directions. For
instance, we could interview you and ask you questions about your
social activity. However, it is unlikely that you could accurately
describe the quantity and quality of your social life in an
interview.

We could also follow you around and write down how often you
spent time talking to other people and ask you questions about how
you felt when you interacted with these people. This approach
would be very impractical if not impossible.

However, we will use a technique which, in previous research,
has been shown to provide accurate descriptions of people’s social
interactions. There have been a few other studies like this
conducted in the U.S. Basically, this technique involves you, as
participants, completing daily records of your social interactions
which last 10 minutes of longer. This involves recording the
length of the interaction, the number of people you were with, and
a few rating scales on which you rate how you feel about the
interaction. It takes about a minute to complete each record.
These records are not too difficult to complete and we would like
you to complete them for a period of two weeks. Don’t be put off
by the idea of doing this for 2 weeks, even if you have 10
interactions per day this is only 10 minutes of recording.

Before I describe the interaction record in detail and
provide you with some examples of social interactions, I must
stress that the entire study depends on your honesty and full
cooperation. If you are not honest we cannot make valid inferences
about the data. Your honesty and cooperation are essential to the
study because you are not only subjects in this study - you are co-
investigators in the sense that you are creating, observing, and
recording the data we need. The data you give us are kept in the
strictest confidence. The only reason we need your name is for
organizational purposes. Your name is never used in a data
analysis. We enter the data anonymously and the results are

115
APPENDIX D (CONT’)

reported as aggregates. At no point of the process will you, as an
individual, ever be identified. We want the project to take place
in an atmosphere of trust. We will respect your confidentiality
and we hope that you will be honest and cooperative with us.

So let me describe the diary and a few examples of social
interaction to you [materials are administered; repeat what a
social interaction is, and describe the diary examples being
explicit about each scale].

Now at the end of the day there is another measure we would
like you to complete. It is very short. It is called the Personal
Reaction Form (hold up) and involves rating how you feel at the end
of the day, i.e., at the time you are completing the form.

In a nutshell then we want you to describe your social
experiences and how you felt on a daily basis.

We would like you to read over the instructions right now to
make sure you understand them. ANY QUESTIONS? O.K. begin your
recording tomorrow. You will receive 8 credits for your
participation. During the week, (assistant’s name) will call you
to see if all is going well with the record keeping. You will be
called again in the second week to arrange a meeting during which
you will complete a post-experimental questionnaire booklet which
includes questions about your reactions to the study. This meeting
will take about an hour and at the end you will be given your
credits. You can turn in your completed records at this meeting.

We have rooms scheduled for this meeting. We would like you to
attend the session on . If not, we have other times.

If you have any questions at any time of the study do not
hesitate to call us. There will be extra materials on my door for
anyone who needs them. If, for whatever reason, after hearing
about what your participation involves you feel that you would not
like to be a part of this study, that’s fine. You will get one
credit for coming to the meeting. If you would like to be a part
of the study please read over the consent form being administered
and sign your name. On the index card put your name and a phone
number where we can reach you and a good time to call you. These
index cards are only used for the purposes of calling you to see
how things are going and to arrange the last meeting. Look over
the instruction booklet again, it has our names and numbers, and
office location on it in case you need to call. Remember to begin
your recording tomorrow and read your instruction booklet again.

If you do not want to participate come up now and we’ll stamp your
card for one credit.

116
APPENDIX E

Information to Participants

We would like to take this opportunity to thank you very much for
participating in this study. Your responses have made an important
contribution to the study of social interaction and psychology in
general.

As we told you at the beginning of the study, social psychologists
do not know a great deal about social interaction. We will be
using the responses of all the participants in this study to
further investigate how patterns of social interaction affect other
variables such as health and academic performance and how
personality may affect these relationships.

The records you completed measured the exchange of support and the
quality and intimacy of your interactions. We are interested in
potential sex differences in these measures. Previous research has
shown that females receive more support from others than do males.
We expect to repeat this finding but we also have hypothesized that
females will report oroyioiog more support than will males. We
also expect, as previous research has shown that females will
report more pleasure and intimacy with their interactions than will
ma es.

Previous research has shown that the more support people receive
the better off they are in terms of mental and physical health. In
other words, someone who has many friends and supportive
relationships is healthier than someone who doesn’t have any
friends or only non-supportive relationships. Our measures of
health were some of the questionnaires in your booklet as well as
the daily feelings list. Research has also shown that this
relationship may depend on the level of stress an individual is
experiencing. That is, it is not always the case that supportive
relationships affect your health directly but that they act to
buffer stress such that your friends reduce the negative affects
that stress can have on your health. The measure of stress in this
study was the daily hassles scale. We hope to repeat these
findings but also investigate the affect that the provision of
support may have on your health. Previous research suggests that
providing support in excess of what you receive may adversely
affect your health. We intend to use the data you provided to test
this hypothesis.

117

APPENDIX E (CONT’)

We are also interested in how social interaction affects grades.
For instance, previous research revealed that the more time males
spent socializing the lower their GPA while the GPA of females was
unaffected by the amount of time that females spent socializing.

We hope to shed further light on this relationship by investigating
various components of social interaction, such as support, thus
giving us a more in-depth look at what may explain this sex
difference.

The post-experimental questionnaire was designed to assess your
reactions to maintaining the records and to get some background
information as to what kind of person you are. The analyses of
these data will allow us to see how various individual differences
affect social interaction. We expect to replicate the findings of
previous research showing that those who score high on the various
measures of social orientation report receiving more support. We
also expect that individuals high on the measure of emotional
intensity would be more likely to provide and need emotional
support than individuals low on emotional intensity.

You should also be aware that this method of data collection is not
that common. Many studies of social interaction and social support
use either questionnaires or interviews to assess the nature of
people’s social lives. Other researchers bring people into the lab
and observe their interaction with another person. These are all
valid ways of studying social interaction. The present method,
however, is more naturalistic than these others in that you are
recording the data as you go about your daily activities thus

providing a more dynamic approach to the study of social
interaction.

If you have any questions regarding your participation in the study
you can call Linda Sullivan at 353-3934, 436 Baker Hall.

118

APPENDIX F

POST-EXPERIMENTAL INTERVIEW
Name: Sex:

Hi my name is ---- etc. We would like to ask you a few questions
about the experience of keeping the interaction records to get a

more personal reaction to the experience. Your name was randomly
chosen from the sign up sheet along with a number of others. Do

you have a few minutes?

If no -- thank them anyway and try someone else.

(1) Did you have any difficulties using the diary to record your
social interactions?

(2) About how many minutes did you spend each day recording your
interactions?
minutes

(3) About how many minutes did you spend each day on the mood
scales?

minutes for mood
(4) How many times per day did you update the records?
times When did you do so? (What time)

(5) How did you decide what a social interaction was, that is,
what prompted you to make a recording?

(6) What situations or times made it difficult to make a recording?

(7) Did the % of interactions that lasted 10 minutes or more you
did not record refer to entire missed days or did it refer to a few
interactions missed everyday?

(8) Why didn’t you record some interactions that lasted 10 minutes
or longer? If they hesitate prompt them. E.g. "only recorded
important interactions or long ones or couldn’t remember”. Be sure

to find out why-—no one said they recorded 100% so they all missed
some.

119

APPENDIX G

NOMENCLATURE OF RIRAP VARIABLES

SEX,OVSSN,OVSSR,OVINT,OVQUAL,OVSSP,OVDIFF,OVSON,OVPD,
OVTPD,OVLEN,OVPERSI,OVPERSZ,OVPERS3,SLIST,OLIST,
GSSN,GSSR,GINT,GQUAL,GSSP,GDIFF,GSON,GPD,GTPD,GLEN,GPCT,
SSSN,SSSR,SINT,SQUAL,SSSP,SDIFF,SSON,SPD,STPD,SLEN,SPCT,SPERI,
SPER2,SPER3,
OSSN,OSSR,OINT,OQUAL,OSSP,ODIFF,OSON,OPD,OTPD,OLEN,OPCT,OPERl,
OPER2,0PER3,
MSSN,MSSR,MINT,MQUAL,MSSP,MDIFF,MSON,MPD,MTPD,MLEN,MPCT,NPER1,
MPERZ,MPER3,
SFSSN,SFSSR,SFINT,SFQUAL,SFSSP,SFDIFF,SFSON,SFPD,SFTPD,SFLEN,SFPCT,
SFBPCT,SFPER1,SFPER2,SFPER3,
OFSSN,OFSSR,OFINT,OFQUAL,OFSSP,OFDIFF,OFSON,OFPD,OFTPD,OFLEN,OFPCT,
OFBPCT,OFPER1,0FPER2,0FPER3,
SlSSN,SlSSR,SlINT,SlQUAL,SISSP,SlDIFF,SlSON,SlPD,SlTPD,SlLEN,SIPCT,
SlBPCT,SlPER1,SlPER2,SlPER3,
SZSSN,SZSSR,SZINT,SZQUAL,SZSSP,SZDIFF,SZSON,SZPD,SZTPD,SZLEN,SZPCT,
SZBCPT,SZPER1,SZPER2,SPER3,
SBSSN,S3SSR,S3INT,S3QUAL,S3SSP,S3DIFF,S3SON,S3PD,53TPD,S3LEN,S3PCT,
S3BPCT,S3PER1,S3PER2,S3PER3,
OlSSN,OlSSR,OlINT,OIQUAL,OlSSP,OlDIFF,OlSON,OlPD,OlTPD,OlLEN,OlPCT,
OlBPCT,OlPERl,OlPER2,0lPER3,
OZSSN,OZSSR,OZINT,OZQUAL,OZSSP,OZDIFF,OZSON,OZPD,OZTPD,OZLEN,OZPCT,
OBBPCT,OZPER1,0ZPER2,02PER3,
O3SSN,O3SSR,O3INT,O3QUAL,O3SSP,O3DIFF,O3SON,O3PD,03TPD,O3LEN,O3PCT,
O3BPCT,03PERI,03PER2,03PER3,
STSSN,STSSR,STINT,STQUAL,STSSP,STDIFF,STSON,STPD,STTPD,STLEN,STPCT,
STBPCT,STPER1,STPER2,STPER3,
OTSSN,OTSSR,OTINT,OTQUAL,OTSSP,OTDIFF,OTSON,OTPD,OTTPD,OTLEN,OTPCT,
OTBPCT,OTPER1,0TPER2,0TPER3,
OVWRK,OVSOC,OVREL,OVLIF,OVOTH,GWRK,GSOC,GREL,GLIF,GOTH,
GPER05,GPERIO,GPERIS,GPERZO,GPER25,GPER30,GPER35,
SWRK,SSOC,SREL,SLIF,SOTH,OWRK,OSOC,OREL,OLIF,OOTH,
MWRK,MSOC,MREL,MLIF,MOTH,
SFWRK,SFSOC,SFREL,SFLIF,SFOTH,OFWRK,OFSOC,OFREL,OFLIF,OFOTH,
STWRK,STSOC,STREL,STLIF,STOTH,OTWRK,OTSOC,OTREL,OTLIF,OTOTH,
SlWRK,SlSOC,SlREL,SlLIF,SlOTH,SZWRK,SZSOC,SZREL,SZLIF,SZOTH,
S3WRK,S3SOC,S3REL,S3LIF,SBOTH,OlWRK,OlSOC,OlREL,OlLIF,OIOTH,
OZWRK,OZSOC,OZREL,OZLIF,OZOTH,O3WRK,O3SOC,03REL,OBLIF,O3OTH,

120
APPENDIX G CONT’

DESCRIPTION OF RIRAP VARIABLES

The first part of an index name indicates what group of behaviors was
used to calculate the index and the second part indicates what aspect of
those behaviors the index represents. Indices beginning with OV
represent overall (or all) interactions, including group interactions.
Those beginning with 6 represent group interactions, that is, those
interactions where more than three people were present. Interactions
that included only same sex others are represented by indices beginning
with S, opposite sex only with an O, and those of mixed sex composition
with M. All of these three include appropriately composed group
interactions. Interactions that included any of the three most frequent
same sex initials are represented by indices beginning with SF, with the
corresponding opposite sex summaries beginning with OF. Interactions in
which the most frequent same sex interactant was present are represented
by indices beginning with 51, those in which the second most frequent
same sex interactant was present are represented by variables beginning
with $2, and the third with S3; with corresponding indices for the three
most frequent opposite sex interactants. The last prefixes are ST and OT
and these indices represent interactions with same and opposite sex
(respectively) individuals who were not among the three most frequently
mentioned interactants. These may be considered to be acquaintances or
strangers. Because no initials are recorded for group interactions,
group interactions are not included in calculation of the SF, OF, SI-3,
01-3, ST and OT indices.

The second part of each index name indicates what quantity was used
to calculate the index. The various scales used on the RIR are
represented rather clearly: SSN, social support needed; SSR, social
support received; DUAL, quality; INT, intimacy; SSP, social support
provided; DIFF, difficulty of providing social support; SON, perception
of others’ need for social support. All of these indices are simple
averages of all the valid values for a variable within a class of
interactions. So, for example, OVINT, is the average intimacy across
all of the interactions reported by a diary keeper, whereas SlINT is the
average intimacy across only those interactions in which the most
frequently appearing same sex initial was recorded. The index suffix PD
represents per-day, and stands for the average number of interactions of
a certain type that were recorded for each day the RIR was maintained.
Therefore, OPD is the average number of opposite sex interactions
recorded each day by a particular diary keeper. Similarly, the suffix
TPD represents time per-day and stands for the average amount of time
spent in interactions of a certain type. For example, SFTPD, represents
the average amount of time a diary keeper spent per day in interactions
that included at least one of his or her three closest friends.

121
APPENDIX G CONT’

Most of the other index suffixes represent percentage breakdowns of
different kinds. The most common of these are the PERI, PERZ, PER3
suffixes, and they stand for the percent of interactions of a certain
type in which one, two, or three people were represented as being
present. Therefore, SPERl represents the percent of all same sex
interactions in which only one other person was recorded as present. The
second major percentage breakdown is represented by the PCT indices, and
they stand for the percent of all interactions that fell into a certain
category. The SPCT, OPCT, MPCT, and GPCT use the total number of all
interactions as a denominator. Therefore, MPCT is the percent of a
participants’s interactions that were mixed sex in composition. However,
the friendship percent indices, SF, OF, Sl-3, 01-3, ST, and OT, use only
the total number of non-group interactions in the denominator. Group
interactions were not included in these denominators because individual
interactants are not recorded for group interactions and it would be
misleading to count an event in a denominator that could not be counted
in the numerator. A similar set of indices are the BPCT indices. These
percents use the same numerator as the PCT indices, but use a slightly
different denominator. For the same sex BPCT indices the denominator is
all non-group interactions minus opposite sex non-group interactions,
while for the opposite sex indices it is all minus same sex. The
rationale for these additional indices was that since a same sex person
could not be present in an opposite sex interaction, such events should
not be included in the denominator and vice versa for opposite sex. In
essence, the BPCT indices are the PCT indices adjusted for relative
differences in the amount of same and opposite sex contact that has been
reported. As an example, OZPCT is the percent of e11 non-group
interactions in which the number two opposite sex set of initials
appeared, whereas OZBPCT is the percent of opposite and mixed sex non-
group interactions in which this person appeared. The third major
percent breakdown represents the nature categorical scale used on the
RIR: WRK, work; SOC, socializing; REL, relaxation; LIF, life necessity,
and OTH, other. Denominators used to compute these percents are the
totals of different behaviors of a certain type of interaction. For
example, SREL is the percent of all same sex interactions that were
recorded as being relaxation.

The only two groups of indices not covered by the previous
description were LIST and GPER sets. The SLIST index represents the
number of different same sex individuals recorded per day in all of a
diary keepers’ interactions and OLIST represents the corresponding
opposite sex figure. The GPEROS index represents the percent of group
interactions recorded as having up to 5 others present, GPERIO is the
percent of group interactions with 6-10 others, etc.

122

FOOTNOTES

1 For example, some subjects reported recording only important
interactions, others recorded only a small percentage of their
interactions, and others made consistent errors of omission such as not
recording any interactions with their roommate.

2 Ten minutes was chosen as a minimum length so as to capture as
much as possible of an individual’s social life without rendering the
recording overburdensome or trivial. Further, a pretesting session
conducted with 36 females and 18 males revealed that the average length
of an interaction that was considered to be supportive was considerably
longer than ten minutes (females, M - 38.94, SD - 36.31; Males, M -
58.56, SD - 79.82; t(1,52) - -1.25, o = n.s.). Thus, the ten-minute
minimum was considered to be appropriate.

3 Aoooreey of goo Beeotjone towards Recording

On the post-experimental questionnaire and interview, subjects were
asked about the experience of maintaining the records. T-tests revealed
some differences between males and females on these measures. All items
were responded to on a 7-point scale (1 = not at all; 7 . extremely)
unless otherwise indicated.

1. Was it difficult to maintain the records? 1(1,122) - -2.26, o < .05
(females M - 3.01, males M - 3.59).

2. Were your records accurate? 1(1,121) - 3.01, o < .01 (females M -
5.72, males M - 5.28).

3. Was maintaining the records time consuming? Females M - 3.36,
males M . 3.63; o = n.s.

4. Did maintaining the records interfere with your daily events?
Females M - 2.35, males M - 2.61; o = n.s.

5. How interesting did you find this research project? Females M -
4.72, males M - 4.39; o - n.s.

6. Percentage of interactions, lasting ten minutes or longer, that were
not recorded? Females M - 16.03, males M = 17.91; p - n.s.

7. Minutes per day spent recording interactions? Females M - 13.87,
males M - 12.06; o a n.s.

8.3 ginutes per day spent recording feelings? Females M - 3.73, males M
= . ; o = n.s.

9. Times per day that interaction records were updated? Females M -
1.81, males M = 1.95; o = n.s. ‘

Subjects were asked if maintaining the records drew attention to any
aspect of their daily behavior and/or changed that behavior. Twenty-
eight females and eighteen males responded to this question. The
responses constituted "a mixed bag" of reactions, ranging from some
negative ”... created anxiety to keep up to date, did not look forward to
filling out sheets, made end of day annoying" through the generally more
positive "it drew my attention to the people I was spending most of my

123
FOOTNOTES (CONT’)

time with and how happy I was with that time" and the enlightening "it
made me realize that approximately 75% of my social interactions were
with females“ to the encouraging "the questionnaire that asks for a
response at the end of the day about how much we felt a certain emotion,
such as happy, needy or depressed, seemed very useful. Even after this
project was finished I still asked myself how I felt about several
emotional inside characteristics before I went to sleep".

Subjects who were randomly selected for interviews were asked if
they had any specific difficulty with their recordings. Very few did,
and of those who did, the amount of time spent recording was most often
mentioned. These subjects were also asked what times created difficulty
with recordings. Most subjects mentioned the weekends as the most
problematic of times due to increased nighttime activity.

4 Due to a coding error at the level of program input it was
decided to exclude four subjects from analyses that involved the various
levels of friendship. The exclusion of these subjects did not affect the
pattern of results. They were included in analyses presented in
subsequent sections.

5 The 24 ratings were factor analyzed by the principal components
method followed by varimax rotation. Reliability analyses were then
conducted on the emergent factors. The results of these analyses
revealed that the ’dependent’ and ’selfish’ items did not load
substantially on any factor and greatly reduced the alpha coefficient of
any scale they were entered into. Upon reviewing the remaining items it
was decided that the ’social,’ ’intimate,’ and ’lonely’ items would
create overlap of measurement when entered into any analyses with the
social variables. It was therefore decided upon to drop these five
variables from analyses. The factor analyses were then rerun on the
remaining 19 items.