51;?
”Hex
x

a! i
.3921;
, $35....

5 ..
aaé-aazata
é’i'fi” ‘-
$813

5

i}

. 35
‘ ’35
r

 

 

14.1“ . . 1.

iii. l! at...

{iii
firinl

16']:

. mam.

s»

4:235:91
"CL

 

hapwwwr

 

 

“‘5“ lllllll'lllllllllllillllllll llllllllll

(r {.- 17 x) 3 1293 01588 0622

LIBRARY

Michigan State
University

 

 

 

 

This is to certify that the
thesis entitled
Dominance and Power Struggles in Alcoholic Marital

Interactions: Alcoholic Subtype Variations

presented by

Michelle Klotz Daugherty

has been accepted towards fulfillment
of the requirements for

M.A. Psychology

 

 

degree in

W

Major professor

Date November 19, 1996

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

PLACE II RETURN BOX to man this Maui from your record.
TO AVOID FINES return on or before date duo.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU ICMWLWMOWIM

DOMINANCE AND STRUGGLES FOR POWER IN
ALCOHOLIC MARITAL INTERACTIONS:
ALCOHOLIC SUBTYPE VARIATIONS
By

Michelle Klotz Daugherty

A THESIS
Submitted to
Michigan State University
in partial fulfillment Of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1996

ABSTRACT

DOMINANCE AND STRUGGLES FOR POWER IN
ALCOHOLIC MARITAL INTERACTIONS:
ALCOHOLIC SUBTYPE VARIATIONS
By

Michelle Klotz Daugherty

This study examined dominance and power struggles in the interactions of
alcoholic and nonalcoholic couples from a community-based sample, and further
distinguished between antisocial and nonantisocial alcoholics in order to shed light
on the conflicting findings of past studies that have conceptualized alcoholics as a
homogeneous group. Specific sub-codes from an observational measure of spousal
interaction (Communication Skills Test; Floyd & Markman, 1984) were used to
evaluate dominance and power struggles that occurred between alcoholic couples in
problem-solving sessions. Results indicate that wives of antisocial alcoholics
struggled for power more frequently than wives of both nonantisocial alcoholics and
nonalcoholic controls. In addition, findings reveal that nonalcoholic husbands
exhibited more positive and effective communication patterns than did alcoholic
husbands; the source of this effect appeared to be the lower level of positive
communication among antisocial alcoholic couples. Implications of the findings for

clinical practice are discussed.

da

far

90:

SSW

ACKNOWLEDGMENTS

I would like to thank Dr. Robert Zucker and Dr. Frank Floyd for their valuable
guidance and essential contributions throughout my work on this thesis. Without
their help, I could not have successfully completed this work. I would also like to
thank the other members of my thesis committee, Dr. Dozier Thornton and Dr.
Robert Caldwell, for their suggestions, support, and patience throughout this
process.

A special thanks also is due to Kalpana Dwarikesh for her dedication and
expertise in the coding of videotapes used for this study. I would also like to thank
Ray Bingham for the many hours he spent sharing his knowledge of statistics and
data analyses.

Finally, I would like to thank my parents, my sister, and my husband, Doug,
for their unconditional love and support. They represent to me everything that is
good and pure in life.

This work was supported in part by grants to R. A. Zucker and H. E.
Fitzgerald from the National Institute on Alcohol Abuse and Alcoholism
(ROiAA07065), from the Michigan Department of Mental Health, Prevention

Services Division, and from the Alcoholic Beverage Medical Research Foundation.

Al

RE;

TABLE OF CONTENTS

LIST OF TABLES .............................................. v
lNTRODUCTION ............................................... 1
Alcoholic Couples ........................................ 3
Alcoholism Subtypes and Behavior ............................. 8
Power and Satisfaction in Marriage ........................... 16
Dominance in Alcoholic Couples ............................. 21
Statement of the Problem .................... . .............. 26
Hypotheses ....................................... 28
METHOD .................................................. 30
Subject ............................................... 30
Measures ............................................. 32
Antisocial Behavior Checklist (ASB) ...................... 32
Measuring Wife Dominance and Power Struggles .................. 34
Communication Skills Test (CST) ........................ 35
Procedure ............................................. 39
RESULTS .................................................. 41
Hypothesis 1 ........................................... 48
Hypothesis 2 ........................................... 49
Hypothesis 3 ........................................... 49
Hypothesis 4 ........................................... 50
DISCUSSION ................................................ 54
APPENDICES
Appendix A: Antisocial Behavior Checklist ...................... 66
Appendix B: Communication Skills Test Codebook ................. 69
Appendix C: Experimenter Instructions for Marital Problem-Solving Task . 85
Appendix D: Marital Problem Areas Checklist .................... 86
REFERENCES ................................................ 87

Table 1 :

Table 2:

Table 3:

Table 4:

Table 5:

Table 6:

LIST OF TABLES

Family Demographic Characteristics of Antisocial Alcoholic,
Nonantisocial Alcoholic, and Control Couples ............... 43

Study Wave of First Marital Assessment [n's and (%l] ......... 45

lntercorrelations among Communication Skills Test Variables for
Dominance, Power Struggles, and Communication Proficiency . . . 46

lntercorrelations among Communication Skills Test Variables for
Dominance and Power Struggles ........................ 47

Analyses of Variance for relative Frequencies and Conditional
Probabilities of Dominance and Power Struggle Behaviors among
Antisocial Alcoholic, Nonantisocial Alcoholic, and Control

Couples ......................................... 52

Analyses of Variance for Communication Proficiency among Antisocial
Alcoholic, Nonantisocial Alcoholic and Control Couples ........ 53

 

 

 

as:

19:

INTRODUCTION

The quality of intimate relationships plays an important role in an individual's
happiness. In particular, marital adjustment correlates with personal well-being
(Berscheid & Peplau, 1983). In a review of literature on marriage and physical and
emotional disorders, Bloom, Asher, and White (1978) conclude that there is a strong
relationship between marital disturbance and physical and mental problems.
Separated or divorced individuals have been shown to be overrepresented among
psychiatric patients, motor vehicle accident victims, and those who commit
homicide and/or suicide. Furthermore, persons who have experienced marital
disruption have a higher incidence of illness and disability (Bloom et al., 1978).
Similarly, Burman and Margolin (1992) conclude that marital status is indirectly
related to overall mortality. They also state that support exists for the nonspecific
and indirect effects of marital status, marital quality, and marital interactions on the
etiology Of health problems. Hence, studying marital relationships through scientific
investigation can reveal information on the causes and treatment of marital
problems which in turn may provide one avenue to improving adjustment and
well-being for individuals.

It is Often claimed that individuals who abuse alcohol have turbulent
marriages and family lives (Baker, Blampied, & Haye, 1989). Alcoholism is
associated with such marital problems as spouse abuse and divorce (O’Farrell,
1992). From one perspective, abusive drinking can cause marital conflict; on the

1

2
other hand, marital problems can be viewed as a cause leading to and/or promoting
alcohol abuse. Alcoholics tend to separate and get divorced more often than
couples in the general population, and separated or divorced male alcoholics have
been found to have a shorter time until death than married alcoholics (Lewis, Smith,
Kercher, 8r Spitznagel, 1995; Paolino & McCrady, 1977). For every alcoholic, it is
estimated that there are approximately five other individuals adversely affected by
his/her alcoholism (Paolino & McCrady, 1977). It is therefore important to explore
the patterns associated with alcoholism and marriage in order to be able to help
both alcoholics and their spouses with marital problems, and possibly to have an
indirect impact on alcohol problems as well.

For these reasons, it is especially important to examine marital interactions
when studying marital relationships. Investigating actual marital interactions can
help to identify specific avenues for successful intervention. Further, according to
McClintock (1983), in order to have an adequate conceptualization of close
relationships, it is necessary to describe the regular interactions that constitute a
relationship, and to explain these regular interactions by revealing their causes and
processes by which they are maintained or Changed. Analyzing sequential dyadic
interactions can reveal behavior which increases or decreases the probability of the
occurrence of consequent behavior (Jacob & Leonard, 1992). This analysis is
valuable in that it provides information regarding the specific influence that one
spouse's actions can have on the other's reciprocal actions, as well as revealing the
interdependence of a couple's behavior. For alcoholic couples, spousal dyadic
interactions which potentially are affected by and also may influence alcohol abuse

and levels of marital satisfaction need to be explored. Thus, the present study

 

 

r8

Of

aff

inf)

little

3

examines dyadic interactions of alcoholic couples. Specifically, wife-dominance and
power struggles, which are of central relevance in alcoholic marriages, are explored.
AIcoholiflZounles

Before the 19605, research involving alcoholic couples mainly was focused
on the alcoholic's spouse (McCrady, 1987; O'Farrell & Noel, 1989). Because the
majority Of alcoholics are men, this literature dealt with wives Of alcoholics. Two
theories existed regarding these wives (Jacob, Favorini, Meisel, & Anderson, 1978).
The first stated that a particular type of personality for women lead them to be
attracted to and to marry alcoholics to fulfill their own unconscious needs to
dominate a submissive male (Jacob, 1986). In turn, because of these needs, wives
of alcoholics tended to promote their husbands' continued drinking, and sabotaged
their husbands' efforts to control their alcoholism. In contrast, the second theory
followed the first and considered the effects of a husband's alcoholism on his wife’s
well-being. It proposed that due to the stress of living with alcoholics, these wives
developed psychological disturbances (Jacob, 1986). More recently, some
researchers have begun to integrate the two opposing notions regarding the wives
of alcoholics. Instead of being viewed as either the source or victims of their
husbands' alcoholism, Moos, Finney, and Gamble (1982) proposed using a
multifactorial approach in which wives develop coping strategies because they are
affected by their husbands' alcoholism, and in which wives' behavior may partly
influence alcoholics' drinking patterns.

Jacob et al. (1978) reviewed studies using the Minnesota Multiphasic
Personality Inventory (MMPI) with wives of alcoholics, and concluded that there is

little support for the view that most of these women have psychiatric disturbances.

4
They proposed that similarities between wives of alcoholics and other groups of
women need to be examined. Generally, little evidence exists that supports the
notion that wives of alcoholics having disturbed personalities (Watts, Bush, &
Wilson, 1994). Although the idea that these women develop disturbances due to the
stress associated with living with an alcoholic has received some support, it is not
clear whether this is a cause and effect relationship (Jacob et al., 1978). Because
most studies have not controlled for confounding variables such as poverty,
unemployment, and other life stress (Jacob, 1986), these women could possibly be
experiencing stress due to a number of different confounding factors which are
independent from the effects of alcoholism. Furthermore, according to Jacob et al.
(1978), many of these studies have very small and restricted groups of subjects,
such as those who were seeking treatment and/or were recruited through social or
legal institutions. Non-help-seeking wives are extremely underrepresented. Other
problems entail the lack of reliable measures, failure to use control groups, and an
over-reliance on self-report data.

During the 1970s and 1980s, much behavioral research was conducted
examining the interpersonal interaction of alcoholics and their wives. A number of
researchers began to emphasize a family system model of drinking behavior which
proposes that drinking serves an adaptive function for the alcoholic in his familial
relationships (Davis, Berenson, Steinglass, & Davis, 1974; Jacob et al., 1978;
Steinglass, 1981 ). When an alcoholic becomes intoxicated, adaptive consequences,
such as more assertive behavior, more emotionally open contact, and increased
sexual activity, are revealed within the family (Davis et al., 1974; Steinglass, 1981).

In support of this notion, early studies found that wives looked at their husbands

 

SI

Ch

5
more during discussions of the husbands' alcohol problem than during neutral tOpic
discussions (Becker & Miller, 1976; Hersen, Miller, & Eisler, cited in McCrady,
1987). Hersen et al. hypothesized that wives' increased attention may reinforce the
husbands' alcoholism (cited in Becker 8r Miller, 1976). Additionally, Becker and
Miller (1976) found that husbands spoke more than their wives when discussing
alcohol; whereas, wives spoke more during non-alcohol related discussions which
seems to indicate that discussing alcohol is positively reinforced for the alcoholic
male. However, these findings are based upon men hospitalized for alcoholism and
their wives, contrasted with non-alcoholic psychiatric patients and their wives.The
alcoholic group did not differ significantly from the non-alcoholic psychiatric group
on these variables. Perhaps comparison of the alcoholic couples group to a normal
control group would have revealed more distinct information regarding the
communication patterns in alcoholic marriages.

Other research focuses attention on the quality, not simply the quantity, of
communication between alcoholic couples. Frankenstein, Hay, and Nathan (1985)
found that alcoholics, while intoxicated, made more problem-describing statements
and spoke more frequently than their wives. These alcoholic couples revealed more
positive verbal behaviors while the alcoholic husband was intoxicated than when he
was sober. This may represent the adaptive consequences of drinking for alcoholic
couples which Davis et al. (1974) and Steinglass (1981) feel reinforce patterns Of
abusive drinking. However, caution should be taken in generalizing these findings to
alcoholic couples in the general population as only eight couples were used for this
study and the subjects were all seeking treatment for alcoholism which is not a

characteristic representative of the majority of alcoholics.

6

In a series of studies looking at interaction with and without alcohol present,
Jacob, Ritchey, Cvitkovic, and Blane (1981), in contrast to Frankenstein et al.
(1985), found that alcoholic couples were more negative when alcohol was present
than when it was not; whereas, nonalcoholic couples expressed an equal amount of
negativity in both conditions. Although based on interactions in the presence of
alcohol, not on interactions involving alcohol-related issues, Jacob et al.'s findings
seem somewhat in opposition to Hersen et al.'s and Becker and Miller's conclusions.
Similarly, a study by Jacob and Krahn (1988) found that alcoholic couples revealed
more negative behaviors when drinking during discussion of personal problems than
did nondistressed and depressed couples. However, there was no difference
between the three groups in the non-alcohol sessions. Perhaps alcoholics can
express hostility and negative affect while intoxicated; however, they attribute their
behavior to the alcohol and avoid taking responsibility for their feelings and actions
which would not have been expressed without the alcohol. On the other hand,
negative behavior may increase in the alcoholic couples when drinking simply due to
the expectancy that negative events occur when the alcoholic is intoxicated (Jacob
& Krahn, 1988).

Using a larger number of subjects, O'Farrell and Birchler (1987) conducted a
study of alcoholic, maritally conflicted, and nonconflicted couples which included
self-report and observational data; however, unlike the studies discussed above, the
effects of the presence of alcohol was not assessed. The results revealed that
alcoholic and maritally conflicted couples had a greater amount of relationship
distress, including less positive verbal behavior and a greater desire for change in

their relationships, than nonconflicted couples, but had similar levels to one another.

7

Additionally, alcoholic husbands and maritally conflicted husbands revealed
equivalently higher levels of responsibility-avoidant communication than their wives
and than nonconflicted couples. However, alcoholic husbands reported a greater
level of marital satisfaction than their wives, and were also more satisfied than
maritally conflicted husbands. This may indicate that alcohol numbs the alcoholic's
awareness of his wife's distress or that his wife has come to view expressing her
distress with the situation as pointless. The lack of differentiation between
alcoholic and maritally conflicted couples is surprising; however, both groups were
seeking marital therapy. This may signify that these two groups were more similar
than they would be in the general population in that these subjects may have been
more apt to recognize the need for marital therapy than members of more
dysfunctional families. In general, it seems that past studies of alcoholic couples
have used subjects who were all seeking some form of treatment which creates
difficulty for generalizing results to more general populations. Furthermore, the
literature reveals that, although conflicting findings exist regarding the effects of
alcohol on these couples' interactions, alcoholic couples do tend to have distressed
relationships when compared with normal control couples.

Findings of husbands' increased negativity while intoxicated (Jacob & Krahn,
1988) and distress in the relationships of alcoholic couples (O'Farrell & Birchler,
1987) regard verbal aggression or conflict; however, similar processes may underlie
both the verbal and physical aggression that can occur in alcoholic marriages
(Leonard, 1993). In a study examining heavy drinking and marital violence, Leonard
and Blane (1992) found a strong relationship between alcohol consumption and

physical aggression, and an interaction between heavy drinking, hostility, and

8
marital dissatisfaction. Drinking patterns were shown to be associated with marital
aggression in subjects high in hostility and subjects low in hostility but high in
dissatisfaction, although only self-report data from husbands was used for this
study. Using self-report data from a sample of newlyweds, Leonard and Senchak
(1993) found that alcohol consumption and hostility were related to premarital
aggression when husbands were at the mean or higher for marital dissatisfaction.
Leonard and Senchak (1996) also found that husbands’ premarital drinking was a
strong predictor of husband marital violence, although the authors state that the
processes which link drinking alcohol to marital aggression need to be investigated
further. Thus, although husbands' alcohol consumption seems related to marital
aggression, it is important to continue to explore characteristics that may increase
the risk of marital violence when combined with specific drinking patterns (Leonard,
1993).
El I I' 5 I | I B I .

One potential limitation of previous research on alcoholic couples is that it
has failed to make a distinction between alcoholic subtypes. In order to more
accurately assess the interactions between alcoholics and their spouses, it is
necessary to differentiate between the subtypes of alcoholism that may exist. In an
effort to explore the relationship between different subtypes of alcoholism and
marital communication, Jacob and Leonard (1988) reanalyzed the data from Jacob
and Krahn's study in an attempt to examine whether the interactions of "episodic"
alcoholics and their wives differed from those of "steady" alcoholics and their wives
in problem-solving sessions. They differentiated the two types of alcoholics based

on alcohol-specific factors (factors related to the alcohol itself, such as drinking

 

Sl

CC

CC

Kr

sh

DO

sin

9
practices) (Zucker, 1994). Episodic alcoholics were considered to have drinking
binges; whereas, steady alcoholics drank approximately the same amount of alcohol
every day. Results show that episodic alcoholics, while drinking, displayed more
negative behavior than did their wives; in contrast, steady alcoholics, while drinking,
displayed less negative behavior than their wives. The two groups did not differ
while sober. In addition, episodic couples tended to be less focused on problems
while drinking than were steady couples. It seems that drinking would be reinforced
only for the steady alcoholic couples in that they could communicate and
problem-solve more effectively. The increased negativity and decreased problem
solving focus for the episodic alcoholic couples while drinking may reflect a form of
coercive control in which the alcoholic husband is able to avoid dealing with
conflictual issues. However, because Jacob and Leonard, as well as Jacob and
Krahn, excluded alcoholics from research who had affective disorders, caution
should be taken in generalizing these results to alcoholic couples in the overall
population. Furthermore, it should be noted that Jacob and Leonard only used
simple questionnaire responses to differentiate episodic from steady alcoholics. This
may be problematic in that self-report data can be subject to socially desirable
response sets.

In 1987, Cloninger proposed a neurobiological learning model of alcoholism in
which there were two subtypes. Based on a large adoption study of alcoholism in
Sweden, he investigated alcohol-specific and alcohol-nonspecific factors (factors
which are not specifically related to alcohol, such as impulsivity (Zucker, 1994))
which could distinguish alcoholics. Cloninger concluded that "milieu-limited" (Type

I) alcoholics have a later age of onset, and are more likely to be psychologically, as

 

 

 

b

SI

Me
cor
risk
con.
dysf
alcol
nota
acoh
acohc
’eSeml
aSSOCIa
assOCIa

Severity

10

opposed to physically, dependent on alcohol, to lose control once drinking binges
begin, and feel guilty about their alcohol problem. Conversely, "male-limited" (Type
II) alcoholics develop alcohol problems at an earlier age, are more impulsive and
have developed spontaneous alcohol-seeking behavior, are more likely to exhibit
disruptive behavior when drinking, and have traits characteristic of Antisocial
Personality Disorder. These findings are similar to the findings of Jacob and
Leonard in that the "male-limited" alcoholic is the more severe subtype and is more
likely to have turbulent social relationships; however, Cloninger and Jacob and
Leonard disagree somewhat on the characteristics of the other subtype. Both
studies, though, provide support for the notion of two subtypes of alcoholism.

Similar to Cloninger's subtypes, Babor, Hofman, DeiBoca, Hesselbrock,
Meyer, Dolinsky, and Rounsaville (1992) found that two groups of alcoholics
comprised their sample. Type A is associated with a "later onset, fewer childhood
risk factors, less severe dependence, fewer alcohol-related physical and social
consequences, less previous treatment for alcohol problems, less psychopathological
dysfunction, and less distress in the areas of work and family" (p.605). Type A
alcoholism resembles Cloninger's Type I alcoholism in that it has a later onset and is
not associated with antisocial behavior (Babor et al.). However, Cloninger's Type I
alcoholics include those who lose control over drinking binges; whereas, Type A
alcoholics exhibit relatively few drinking binges. In contrast, Type B alcoholism
resembles Cloninger's Type II alcoholism in that they both begin early in life and are
associated with antisocial characteristics (Babor et al.). Type B alcoholism is
associated with "more childhood and familial risk factors, earlier onset, greater

severity of dependence, polydrug use, more serious consequences, a more chronic

11
treatment history (despite their younger age), greater psychopathological
dysfunction (current and lifetime), and more life stress" (p.605). Importantly, Babor
et al. employed subjects from a clinical population of individuals seeking alcohol
treatment, and therefore, their findings may not easily generalize to alcoholics in the
general population who do not feel or recognize the need for treatment.

In further exploration of the relationship between alcoholism and antisocial
personality, Hesselbrock, Hesselbrock, Babor, Stabenau, Meyer, and Weidenman
(1984) found that alcoholics with Antisocial Personality Disorder (ASP) are younger
than alcoholics without ASP, have a greater proportion of first and second degree
relatives who are alcoholics, have more adjustment problems as children and have
other substance abuse problems. Alcoholics without ASP have attained higher
education levels and have more professional occupations. However, unlike Babor et
al., no differences were found regarding levels of other forms of psychopathology.
Nevertheless, the similarities in the findings between the two studies are numerous,
as well as the fact that both samples were from a clinical population of alcoholics
seeking treatment. Thus, the evidence indicates that alcoholics with Antisocial
Personality Disorder or antisocial characteristics tend to have an earlier onset of
alcoholism, polydrug use, and more serious interpersonal problems, although caution
should be taken in generalizing results to alcoholics not seeking treatment.

In a review of the subtyping literature, Zucker (1987) suggested that
evidence was sufficient for the presence of at least four types of alcoholism, one of
which was antisocial alcoholism. The other types were described as
developmentally cumulative alcoholism, developmentally limited alcoholism, and

negative affect alcoholism. Antisocial alcoholics were characterized by both the

12
early presence of antisocial behavior and the early onset of alcohol problems. In
addition, antisocial alcoholism was described as occurring more frequently in males
and in lower class individuals, and evidenced by continued social and legal
difficulties in adulthood (Zucker, 1987).

The Michigan State University-University of Michigan group, in a series of
studies have continued to explore the empirical relationship among subtypes of
alcoholism. Thus, Zucker, Ellis, and Fitzgerald (1994) found that of their male
alcoholic subjects, 25% could be classified as Cloninger's Type | alcoholics and
60% as Type lls. Like Cloninger, Zucker et al. found that both alcohol-specific and
alcohol-nonspecific factors differentiated Type I and Type II alcoholics. According
to the Zucker et al. data, Type II alcoholics exhibited more antisocial behavior as
children and as adults, had a higher number of close alcoholic relatives, had higher
rates of separations and divorces, and were lower in socioeconomic status than
Type Is. In addition, support was found for the existence of two distinct
developmental path models into alcoholism. No difference, however, was found
between Type I and Type II alcoholics' current level of alcohol consumption.

Zucker et al. also noted that the Type I/Type ll classifications were unable to
categorize a significant number of alcoholics. Therefore, Zucker et al. re-Classified
their sample into antisocial alcoholics (AALs) and non-antisocial alcoholics (NAALs),
using a measure that assesses adult as well as child antisocial tendencies in order to
examine the "degree to which a developmental trajectory has been established that
is continuous across childhood and adulthood" (Zucker et al., 1994, p.13). In their
view, the pathway to alcoholism can begin early in childhood and may involve

alcohol-specific and alcohol-nonspecific factors. Examining this developmental

 

TE

bl

Fr.
(El
Ch

CO

SCc
ant
Chll
Con

al. I

1 3
pathway involved in alcoholism can also reveal the occurrence of comorbid
psychopathology (Zucker, 1993). Using the AAL and NAAL classification, AALs
have an earlier onset of alcoholic problems than NAALs, have more alcohol-related
problems and symptoms, are in a lower SES, have a greater level of depression, and
have a greater family history of alcoholism (Zucker et al., 1994). It also should be
noted that husbands who display more antisocial behavior and have had more
lifetime drinking problems exhibit higher levels of violence towards their wives
(Reider, Zucker, Noll, Maguin, & Fitzgerald, 1988). Once again, however, Zucker et
al. did not find a significant difference between the current alcohol consumption
levels of AALs and NAALs. Thus far, results indicating differences in alcohol
consumption have been found in three studies which may indicate that the more
relevant issue to interpersonal interaction is not the quantity of alcohol consumed,
but the existence of antisocial personality characteristics in the alcoholic.

In an effort to establish purer subtypes of alcoholism, Zucker, Ellis,
Fitzgerald, Bingham, and Sanford (in press) re-classified their alcoholic sample,
requiring alcoholic men to obtain at least minimal cutoff scores on both ratings of
childhood and adulthood antisociality in order to be labeled as high antisocial
continuity alcoholics (antisocial alcoholics). Low antisocial continuity alcoholics
(nonantisocial alcoholics) were considered to be those alcoholics who obtained
scores below the cutoff for both childhood and adulthood antisociality. Thus, these
antisocial alcoholics have experienced continuity of high levels of antisociality from
Childhood through adulthood; while, nonantisocial alcoholics have experienced
continuity of low levels of antisociality from childhood through adulthood. Zucker et

al. (in press) found similar differences between high antisocial continuity alcoholics

 

(A

Sf

nC

ZL
bl;
n0
rely
hug
CQU

ast

14

and low antisocial continuity alcoholics as they had previously for their more
heterogeneous sample of antisocial and nonantisocial alcoholics, including poorer
social functioning for AALs. The authors also found that high levels of antisociality
in childhood, adulthood, or both, are statistically uncommon in nonalcoholic
populations. Furthermore, subjects in both the Zucker et al. (1994) and Zucker et
al. (in press) studies are from a community-based sample who most often have not
sought treatment; therefore, the finding of the existence of two subtypes of
alcoholism generalizes the results of the previouSly discussed studies involving clinic
samples to a broader sample of alcoholics.

lchiyama, Zucker, Fitzgerald, and Bingham (in press), using this
community-based sample and perceptual data from Benjamin's Structural Analysis
of Social Behavior (SASB, 1974), were also able to differentiate between antisocial
and nonantisocial alcoholics in the way they perceived themselves and their
marriages. The authors found that antisocial alcoholic men were more
self-neglectful, more blaming, and less trusting than nonantisocial alcoholics and
nonalcoholics. The antisocial alcoholics were most likely to be rated by their wives
as being detached. Similarly, using data from Benjamin's SASB (1974), lchiyama,
Zucker, Fitzgerald, and Refior (1996) found antisocial alcoholic husbands to be more
blaming towards and less trusting of their wives than nonantisocial alcoholics and
nonalcoholics, in addition to belittling their wives more frequently. A trend was
revealed for greater disagreement among antisocial alcoholic couples regarding
husbands' hostile controlling behaviors than among nonantisocial and nonalcoholic
couples. More generally, husbands and wives across all groups rated their partners

as being more blaming and belittling than the partners rated themselves.

 

r4

vw
OF

dn

COI

Dre

DErE
0ft}
”on-
Fum

drink

15

Based on the above findings, there is a great deal of support for the
existence Of at least two alcoholism subtypes. According to Jacob (1986), families
react differently depending on the type of alcoholic with whom they live. In Jacob's
nosology, which again focuses on the distinction between alcoholics based on
alcohol-specific factors, episodic (binge) drinkers are more likely to drink out of their
home; alcohol does not seem to have any adaptive consequences with regard to
family life for these alcoholics. In contrast, steady drinkers who consume the
majority of their alcohol in their homes may gain benefits from drinking. For
example, based on observations of alcoholic couples in problem-solving sessions,
Jacob (1986) found that binge drinkers revealed more negative verbalizations and
were less likely to remain focused on problem-solving than steady drinkers. Binge
drinkers have also been found to exhibit more marital aggression than steady
drinkers (Murphy 8! O'Farrell, 1994). In a similar vein, Dunn, Jacob, Hummon, and
Seilhammer (1987) found a positive relationship between the amount of alcohol
consumed at home and wives' marital satisfaction ratings. However, as stated
previously, the failure to clearly replicate Jacob's typologies in other studies of
alcoholic samples suggests that this drinking behavior-based categorization may
actually be tapping more general characteristics of these men. That is, the
differences found in Jacob's studies may have more to do with alcoholics'
personality traits, specifically antisocial characteristics, than the quantity or location
of their drinking. Perhaps the steady in-home drinker represents Zucker's
non-antisocial alcoholic and the episodic binge drinker is the antisocial alcoholic.
Furthermore, Jacob and Krahn's findings (1988) that while drinking, episodic

drinkers displayed more negative behavior than their wives, and steady drinkers

 

re
re
Se
co

79

eXar
{in VI-
Sails:

an’UC

16
displayed less negative behavior than their wives can be thought of in terms of
adaptive consequences existing for the non-antisocial alcoholic's drinking. Thus,
this research suggests that the non-antisocial alcoholic will have a more positive and
less disturbed family life than the antisocial alcoholic.
E I S . I I' . M .

Power can be viewed as "the capacity to produce intended effects"
(Gray-Little & Burks, 1983, p.514). However, very few studies actually measure
power in marriage in this way, and instead focus on either of two issues. First, most
studies have used self-report measures to assess the balance of power that exists in
relationships. Power here is often investigated by examining which spouse is
responsible for the decision-making in specific areas, such as where to live.

Second, other studies have used observational techniques which are intended to
concentrate on the processes involved in establishing power (Gray-Little & Burks,
1983).

Early self-report studies on the outcome of power emphasized the importance
of an egalitarian relationship for marital satisfaction. Although a few studies since
the 1960s have reported that husband-dominated decision making leads to the
highest ratings of marital satisfaction, most studies indicate that egalitarian power
leads to the highest levels of marital satisfaction (Gray-Little 8r Burks, 1983). For
example, Blood and Wolfe (1960) found that wives in a syncratic egalitarian group
(in which couples make most of the decisions together) had the highest levels of
satisfaction. However, a major short-coming of this study is that husbands were not

included in the sample. Additionally, Blood and Wolfe found, similar to many

17
subsequent studies, that the women in the wife-dominant group had the lowest
levels of marital satisfaction.

More recent findings from self-report data also have demonstrated the
relationship between inequity of power and marital dissatisfaction. In a study
examining decision making power and self-evaluation, Beach and Tesser (1993)
found a marginally significant relationship between power imbalance and marital
satisfaction, regardless of which spouse held greater power. Other findings have
shown that wives have lower levels of marital satisfaction when decision making is
excessively dominated by husbands (Weisfeld, Russell, Weisfeld, & Wells, 1991)
and that the greater wives' egalitarian views are relative to their husbands' views,
the more negative the marital adjustment (Li & Caldwell, 1987). Aida and Falbo
(1991) found that equal partners were more satisfied with their marriages than
traditional partners, although the authors defined equality as both partners sharing
equal financial responsibility and felt that this represented equally shared power.

In general, studies have provided support for the relationship between
egalitarian power and high marital satisfaction, as well as to the relationship
between wife-dominance and low levels of marital satisfaction (Gray-Little & Burks).
A limitation of most of these studies is that marital power is operationalized only
through self-reports of decision making which can be subject to socially desirable
response biases. In addition, spouses often do not agree on how to rate the balance
of power that exists in their relationship. Furthermore, power-related proceSses that
occur between spouses before a final decision is made need to be examined

because important interactions occur prior to reaching a decision.

 

 

 

OC

SDI

ob:

SIU

COU

deC

con‘
later
thes
prov
level
fort)
Datte
invV.
trenc
Satls'
inter;
founr
havwr
alfllo
inter;
as bu

the n

18

Observational methods can investigate the power-related interactions that
occur between couples, and can eliminate social desirability response sets and
spouses' discrepant ratings. Despite advantages of observation, findings from
observational studies have been more equivocal than those from the self-report
studies. Levinger (1964) used an interview schedule and four tasks to observe
couples' interactions. Dominance was evaluated by way of the number of household
decisions made by each partner, the degree of acceptance or rejection of the other's
contributions, and by contrasting individual performance on the tasks with
later-Obtained couples' joint performance (Gray—Little & Burks, 1983). Although
these tasks still involved decision-making outcomes of power, observer ratings
provided information regarding the couples' interactions by evaluating each partner's
level of acceptance or rejection of the other's contributions. No support was found
for the notion that couples with high marital satisfaction would have different
patterns of dominance than couples with low levels of marital satisfaction.
However, a study by Corales (cited in Gray-Little and Burks, 1983) found slight
trends for greater marital satisfaction for egalitarian couples and the least
satisfaction for wife-dominant couples. Similarly, Kolb and Straus (1974) used an
interaction task in which spouses attempt to discover the rules of a game, and
found that couples with high-power husbands were more likely to be viewed as
having higher levels of marital satisfaction than couples with low-power husbands,
although results were not statistically significant. These findings are difficult to
interpret, however, due to unusual measures, such as the classification of a couple
as high in wife-power if the wife had a higher number of directive acts compared to

the median for all wives. Her level of power relative to her partner's level was not

 

hi
dl:
BS
Spq
su;
the
Dan
egal
CoW
598$:

19

considered. In addition, in an effort to avoid receiving socially desirable responses,
marital satisfaction levels were based on the reports of these couples' ninth grade
children which may not accurately reflect the adults' actual levels of satisfaction.

Sprenkle and Olson (1978) examined power in terms of relative control and
control efficiency. Relative control was defined as the ratio of power attempts that
modify a spouse's behavior to the total number of power attempts made, such as
the number of times a husband gave directions, instructions, etc. relative to the
total number of times these events occurred for the couple. Similar to the findings
from the self-report studies which examined decision-making outcomes, couples
with higher marital satisfaction levels tended to be egalitarian and were less likely to
be wife-dominant in terms of relative control than unhappy couples. Control
efficiency represented "the ratio of the couple's successful power attempts to the
total power attempts" (p.524). Couples who were more maritally satisfied had
higher control efficiency ratios than maritally disturbed couples, indicating that the
distressed couples' attempts to exert control wererelatively unsuccessful,
especially attempts by the husbands. This finding suggests that maritally dissatisfied
spouses, particularly wives, are less responsive to their partners' verbalizations. In
support of this finding, Gray-Little and Burks (1983) report other results showing
that wife-dominant couples, especially wives, reveal less positive regard for their
partners than other couples. More recent support for the relationship between
egalitarian power structures and marital satisfaction was revealed in study by Ball,
Cowan, and Cowan (1995), which involved couples' participating in problem-solving
sessions. These authors found that wives who perceived that they had more equal

influence on the outcome of the session were more satisfied with their marriages.

20

A somewhat different conceptualization of power was formulated by
Gottman (1979). He stated that dominance is asymmetry in the control and
predictability of affective behavior. For example, a husband would be considered
more dominant if his affective state produces a more predictable emotional response
from his wife than her affective state produces from him. Gottman used high and
low level conflict tasks, and observers rated couples' interactions as representing
positive, negative, or neutral affect. Gottman further divided affect into expression
fluctuations, which represent immediate changes in affect, and mood fluctuations,
which involve slow, longer-term changes in affect; however, only the expressive
component on a high conflict task revealed asymmetrical patterns that were more
common for distressed than for nondistressed couples. Furthermore, unlike several
other studies, these asymmetrical patterns reflected husbands' dominance in the
distressed couples. Thus, although husbands were more dominant than their wives,
Gottman's findings reveal some support for the existence of the relationship
between unhappiness in marriage and asymmetrical power.

Findings from these self-report and observational studies generally indicate
that couples are more satisfied with symmetrical power structures, and that
wife-dominance, with the exception of Gottman's results, is related to the greatest
levels of marital dissatisfaction. Couples may be more unhappy with wife-dominated
power due to their departing from the cultural norms of power in marriage (i.e., a
strong wife taking power from her husband) which causes them to become
dissatisfied. Marital dissatisfaction can also be explained as being caused by the
husband's unwillingness or inability to share responsibilities with his wife, rather

than his relinquishing his responsibilities to a dominant wife, thus burdening her with

21

having to take care of everything on her own. These explanations of the relationship
between wife-dominance and marital dissatisfaction seem particularly relevant to
interpreting marital distress among alcoholic couples, where we expect that
frequently the wives have assumed more powerful roles than their husbands (Blood
& Wolfe, 1960).
D . . El I I. Q I

The issue of dominance in alcoholic couples has been described in varying
ways. In general, older studies, reviewed by Paolino and McCrady (1977) have
described wives of alcoholics as playing the dominant role in their marriages. Wives
were generally thought of as being inherently dominant individuals who seek out
dependent alcoholic men. In contrast, the literature on the relationship between
wife-dominance and marital dissatisfaction suggests that these wives may exhibit
dominant behavior due to their having to take control of tasks husbands cannot
complete and roles they cannot assume any longer due to their alcoholism (Paolino
& McCrady, 1977). Wives have been shown to take more responsibility for tasks
usually performed by husbands when these men become intoxicated (Gustafson,
1988). Gray -Little and Burks (1983) argue that dominance conflicts can occur in
marriage in cases where either or both the husband and the wife hold the
stereotypic view that men should be the dominant force in a marriage, but the
husband is unable or unwilling to play the dominant role. This is especially relevant
to alcoholic couples in that alcoholic husbands may be particularly unable to display
dominance due to various alcohol-related factors.

Furthermore, conflict among alcoholic couples may be heightened due to the

discrepancies which may exist between spouses' perceptions of which partner is

22
dominant or submissive. For example, in a study examining conflict and sexual
dysfunction in alcoholic and non-alcoholic couples, results from Benjamin's (1974)
Structural Analysis of Social Behavior instrument revealed that alcoholic husbands
rated themselves as being submissive, whereas, their wives perceived them as
being autonomous. This discrepancy was not found for control couples (Chiles,
Stauss, & Benjamin, 1980). This finding indicates that alcoholic husbands feel that
they are submitting to their wives' wishes, even though their wives do not perceive
their behavior in this way. As a result, conflict may increase if an alcoholic husband
becomes dissatisfied with his perceived submissive role and attempts to exert more
power in the relationship.

The issue of dominance has played an important role in theories about
alcoholic couples (Paolino 8i McCrady, 1977); however, there are a surprisingly
limited number of empirical studies examining dominance in the interactions of these
couples. From their review of older studies, Paolino and McCrady (1977) conclude
that conflicts regarding "dominance-dependence" do exist in alcoholic marriages.
However, they state that definitions of dominance and dependency in research are
unclear. Also, it is unclear whether dominance conflicts are the causes or the
effects of alcoholism.

Paolino and McCrady (1977) state that struggles for control can be assessed
on three levels: the first is the "sociological role distributions" within the marriage,
such as assessing who decides how money is spent; the second is communication
patterns, such as observing who speaks first when a question is asked; and third
level is each partner's personal view of who is dominant and who is dependent.

Three more recent studies that investigate the issue of dominance in alcoholic

23
marriages concentrate on Paolino and McCrady's second level, communication
patterns.

First, a study by Roberts, Floyd, O'Farrell, and Cutter (1985) examined the
effects of length of husbands' sobriety on struggles for control in marital
interactions. It was hypothesized that the longer the duration of sobriety, the fewer
the struggles for control would exist in the couples' interactions. During
problem-solving sessions, "dominance" was measured by the number of questions
asked by each spouse (which was considered a passive control tactic) and by the
total duration of speech for each spouse (lower levels of which were considered a
measure of passivity). Results revealed that husbands in the high sobriety group, in
contrast to those in the low sobriety group, asked fewer questions and tended to
speak for a longer total duration of time. Wives in the high sobriety group also
tended to ask fewer questions than wives in the low sobriety group. Hence, the
author's conclude that this generally provides support for the notion that there are
fewer struggles for control among couples in which the husband has remained sober
for a longer period of time. However, it is not clear if indeed neither partner in the
high sobriety group is dominant. The definition of "struggles for control" does not
adequately represent the term "dominance," because dominance implies an already
established difference in power, rather than a struggle for power. The use of
duration of speech as a measure of dominance is also questionable in terms Of
construct validity because dominance can be established through relatively brief
verbalizations as well. Additionally, the low sobriety group in this study had a lower
level of family income which may have contributed to a difference in marital

interactions.

24

A second study examined the effects of intoxication versus sobriety on
dominance in the interactions of alcoholic couples (Frankenstein, Sullivan, Hay &
Cocco, 1985). It was hypothesized that alcoholics would be more submissive and
would have less influence than their partners but would display more dominance
when intoxicated than when sober. Alcohol was given during one of two
problem-solving sessions, and multiple measures of dominance were used for each
spouse, including a self-report measure of one's typical level of assertiveness, along
with observations of duration of speech, gaze maintenance, number of questions
asked, number of interruptions, perceptions of solutions posed by the other person,
and insiders' and outsiders' global ratings of dominance of the interactions. Both
types of ratings were used in order to examine insiders' and outsiders' perceptual
differences of dominance. These ratings were then compared with the other
measures in the study. In this study, the number of questions asked was positively
associated with both insiders' and outsiders' perceptions of dominance; whereas, in
the Roberts et al. study, question-asking was considered a passive control technique
which is indicative of a struggle for control. Results of the Frankenstein et al. study
revealed that although the alcoholics viewed themselves as submissive, their wives
perceived them as autonomous. In addition, on behavioral measures overall these
men did not appear to have less influence than their spouses. While listening, the
spouses maintained their gaze for a greater length of time than did the alcoholics,
but this behavior was negatively associated with both groups of raters' perceptions
of dominance. When intoxicated, the alcoholics spoke more than their spouses and
also spoke more than they did when they were sober. Additionally, the alcoholics

asked more questions while intoxicated than while sober.

25

In a general sense, these findings support the hypothesis that alcoholics
engage in struggles for control, especially when drinking. According to the outside
raters, the spouses were dominant in the no-alcohol session, but the alcoholics
appeared to be dominant in the alcohol session. In essence, the alcoholics struggled
for control, due to a perceived imbalance of power, and seemed to have gained it in
the alcohol session. These results should be interpreted very cautiously, however.
The total sample size for this study only consisted of eight couples, which included
both male and female alcoholics. In addition, a short-coming of this study is the
failure to include a nondistressed control group. It would be interesting to see how
the alcoholic couples in this study would compare to control couples.

I A third study, by O'Farrell and Birchler (1987), also examined dominance in
the interactions of alcoholic couples during problem-solving sessions. However,
unlike the Roberts et al. study, this investigation differentiated between "struggles
for control" and "dominance." "Struggles for control" were represented by the
number of interruptions. Dominance was operationalized as the percent of a
speaker's interruptions which were successful in causing a floor to switch, and the
frequency that, when being interrupted, the speaker did not allow the floor to
switch. Alcoholic, maritally conflicted, and nonconflicted control couples were
compared. It was hypothesized that greater struggles for control and wife
dominance would exist in alcoholic couples as compared to the maritally conflicted
and control couples. No support was found for wife dominance in the alcoholic
couples. In addition, although alcoholic and maritally conflicted couples had more
interruptions than control couples, they did not differ significantly from each other.

This study provides support, as does the Frankenstein et al. study, that some

26
struggles for control do exist in alcoholic couples, but no support for the hypothesis
that these couples exhibit an imbalance of power. In addition, on most measures
taken, alcoholics do not appear to be significantly different from nonalcoholic
maritally conflicted couples. However, once again, simply using the number of
interruptions to represent struggles for control can be questioned. Interruptions
have been found to occur more frequently in normal families than in disturbed
families, and at times, can indicate that the listener is showing enthusiasm for the
speaker's statement (Jacob, 1975).

As evidenced by these studies, controversy exists over operationally defining
the concept of dominance. It is difficult to compare studies due to the varying
measures used. In addition, it is important to note that in all three of these studies,
all alcoholics were either seeking treatment or had been previously treated for
alcoholism. In the O'Farrell et al. study, alcoholic subjects were also seeking
evaluation for marital therapy. Hence, the subjects in all of these studies may have
very different characteristics than those individuals not looking for help which limits
the ability to generalize results.

W

The above review suggests several conclusions about power and dominance
for alcoholic couples. Whether due to the alcoholic couple's wife having a dominant
personality or to the alcoholic husband being unable to share responsibilities with his
wife, at least some evidence indicates that dominance-dependency issues do exist
for alcoholic couples (Paolino 8: McCrady, 1977). Alcoholic couples also experience
more struggles for control than nonalcoholic couples (Frankenstein et al., 1985;

O'Farrell & Birchler, 1987). This may be due to these wives wanting and/or needing

27
to take control which is contrary to cultural norms, and hence, causes power
struggles to occur between these couples (Gray-Little & Burks, 1983). Nevertheless,
the ambiguities and conflicting findings of previous research suggest that it is
necessary to clarify the notion that wives are more dominant and more power
struggles exist in alcoholic couples than in nonalcoholic couples.

Furthermore, more recent literature suggests that at least two subtypes of
alcoholism exist. The behavioral characteristics of these types of alcoholics suggest
that power struggles may occur more frequently among antisocial alcoholic couples
(high antisocial continuity couples) than nonantisocial alcoholic couples (low
antisocial continuity couples). On the other hand, nonantisocial alcoholics can be
viewed as having fewer alcohol-related problems, less chaotic family relations, and
possibly more stable and predictable drinking patterns, which include drinking at
home more often than do antisocial alcoholics (Zucker et al., 1994; Jacob, 1986).
These families may be able to adapt more easily to life with a nonantisocial
alcoholic. Wife-dominant behavior, whether due to her individual personality,
necessity, or both, would be a more accepted and established pattern of behavior
and should match the earlier findings in the literature of wife dominance in alcoholic
couples. For nonalcoholic couples, less relationship distress exists (O'Farrell &
Birchler, 1987) which should manifest as these couples having more positive
communication than alcoholic couples, less struggles for control, and more
egalitarian power structures.

The present study examines wife-dominance and power struggles in the
interactions of alcoholic couples from a community-based sample. Unlike previous

studies, alcoholic couples are differentiated by way of an alcoholic subtype that

28
other evidence already suggests should be effective in revealing marital dominance
and power struggle differences. The differences between antisocial and
nonantisocial alcoholics are examined, and contrasted with a non-alcoholic control
group. By distinguishing between antisocial and nonantisocial alcoholics, the study
attempts to shed light on the conflicting findings of past studies that have
conceptualized alcoholics as a homogeneous group. By utilizing subjects who were
not recruited for alcohol treatment nor marital therapy, the results of the current
work are potentially generalizable to a greater portion of the population of alcoholic
couples.

Harmless:

Based on these conclusions, the following hypotheses were tested:

1. Power struggles will occur more frequently for antisocial alcoholic coup/es than
for nonan tisocial alcoholic couples and nonalcoholic couples.

a. Specifically, antisocial alcoholic couples will exhibit a higher relative
frequency of Dyadic Power Struggles than nonantisocial alcoholic couples and
nonalcoholic couples.

b. Antisocial alcoholic couples also will display a higher relative frequency of
wives' Hostile Power Assertions followed by husbands' Hostile Power Assertions.
2. Wives of nonan tisocial alcoholics will be more dominant than wives of bo th
antisocial alcoholics and nonalcoholics. Dominance is represented by the wives'
asserting power fol/o wed by a lack of resisting these assertions by their husbands.

a. Specifically, nonantisocial alcoholic couples will exhibit a higher frequency
of wives' Positive Power Assertions followed by husbands' Avoidance, agreement,

or accepting responsibility than antisocial alcoholic couples and nonalcoholic

29
couples. The frequencies Of these three consequent behaviors will be added
together as one category.

b. Nonantisocial alcoholic couples also will have a higher frequency of
wives' Hostile Power Assertions followed by husbands' Avoidance, agreement, or
accepting responsibility than will antisocial alcoholic couples and nonalcoholic
couples. The frequencies Of these three consequent behaviors will be added
together as one category.

3. Nonalcoholic couples will have more positive and effective communication than
either the antisocial alcoholic couples or the nonan tisocial alcoholic coup/es as
represented by higher ratings of communication proficiency.

4. Nonalcoholic couples will display more positive forms of power than either the
antisocial alcoholic couples or the nonan tisocial alcoholic couples.

a. Nonalcoholic couples will have the highest relative frequency of Sharing
Control.

b. Nonalcoholic couples also will have the highest relative frequency of

Positive Power Assertions.

 

METHOD

masts.

Subjects for the current study were 126 couples from the Michigan State
University-University of Michigan Longitudinal Study (Fitzgerald et al., 1993; Zucker
et al., 1996; Zucker 8t Fitzgerald, 1991 ). The study involves court-recruited
alcoholic men and their partners, community-recruited alcoholic couples, and
nonalcoholic community control couples (their children are also part of the larger
project, but are not a focus of the work reported here). Only nonhispanic Caucasians
were included because census data in the area accessed for the study indicated that
other ethnic and racial groups would represent under 10% of the sample. Past
literature has shown a relationship between ethnic/racial status and patterns of
alcohol use which this study would be unable to effectively analyze with the sample
size. Therefore, such variation was excluded in order to decrease error (Zucker et
al., 1994; Zucker & Fitzgerald, 1991). All families in the parent study received some
compensation for their involvement. Data for the current work is from Waves 2, 3,
and 4 of the Longitudinal Study, and was collected during each couple's initial
exposure to the marital problem-solving task.

The court-recruited alcoholics were recruited via a court network in four
Mid-Michigan counties with six district courts (Zucker & Fitzgerald, 1991). When
contacted by project staff, respondents were told that the study had no connection
to the courts and that all information would be confidential. All court-recruited

30

31

alcoholics were convicted of drunk driving and all had a blood alcohol concentration
(BAC) of 0.15 percent or higher, or a BAC of 0.12 percent and at least one other
alcohol-related driving offense. In addition, because one of the major foci of the
parent study is the development of alcoholism, at the time of initial contact, the
alcoholic men all were required to have a biological son between the ages of three
and five years old; at time of first contact the men also had to be living with the
son's biological mother. The mother's alcohol involvement was not a criterion for
accepting or rejecting families. In addition to the initial BAC selection criteria,
alcoholic men were required to make at least a probable diagnosis of alcoholism
using the Feighner diagnostic criteria (Feighner, Robins, Guze, Woodruff, Winokur, 81
Munoz, 1972). This diagnosis was established with a two-stage procedure. At initial
contacts, subjects were screened for positive alcohol diagnoses using items from
the Short Michigan Alcohol Screening Test (SMAST; Selzer, 1975); later, screening
involved using items from the NIMH Diagnostic Interview Schedule (DIS-Version Ill;
Robins, Helzer, Croughan, & Ratcliffe, 1980). At Wave One, 88% of these men met
a definite diagnosis for alcoholism; the remaining men were classified with a
probable diagnosis. The court-referred alcoholics had a mean age of 30.12
(SD =4.76) years and mean years of education of 12.40 (SD =2.03). The wives had
a mean age of 29.05 (SD=4.48) years and mean years of education of 12.83
(SD =2.01). These couples had a mean annual family income of $17,200
(SD= $4,720), were married for a mean of 6.92 (SD =3.61) years, and had an
average of 2.30 (SD=0.94) children.

After obtaining each court-referred alcoholic family for the parent study, a

matched community control family from the same census tract as the alcoholic

 

32

family was located using door-to-door canvassing interviews (Zucker & Fitzgerald,
1991). Community families were screened for an approximate age appropriate male
child, and to the extent that it was possible, the control family was also selected to
match the sibling composition and birth order of the alcoholic family. The same
procedures were used for the community controls as for the court-referred
alcoholics to ensure that neither the husband nor wife in a control family met either
an alcoholism or drug dependence diagnosis. Community control families in which
the husband met a probable or definite alcoholism diagnosis were retained in the
study but were reclassified as community alcoholic families.

At Wave One, husbands in the community control sample had a mean age of
31.93 (SD =4.05) years and mean years of education of 13.46 (SD = 2.18). The
wives had a mean age of 31.02 (SD=4.17) years and mean years of education of
13.83 (SD=2.55). These couples had a mean annual income of $18,240
(SD= $4,900), were married for a mean of 6.97 (SD=4.04) years, and had an

average of 2.25 (SD= 1.03) children.

Measures
EI' 'IBI . El (1' IESBI

The Antisocial Behavior Checklist (ASB; Zucker & Fitzgerald, 1992; Zucker &
Noll, 1980; see Appendix A) is a 46-item revision of an earlier antisocial behavior
inventory used in the Rutgers Community Study (Zucker 8: Barron, 1973). The
current version is a modification that allows assessment of adult antisocial activity.

The A88 is a self-report measure which examines the frequency Of an
individual's antisocial and aggressive behavior as a child and an adult. This

instrument assesses antisocial behaviors such as parental defiance, delinquent

33
behavior, and physical aggression. In a series of reliability and validity studies
ranging from male and female college students to male and female prisoners, it
achieved adequate test-retest reliability (.94 over four weeks) and internal reliability
(Cronbach‘s Standardized Alpha range -.67 to .97) (Ham, Zucker, & Fitzgerald,
1993; Zucker et al., 1994). Respondents answer each item on the A88 as occurring
"never," "rarely," "sometimes," or "often." These responses are scored 0-3,
respectively and a total score is derived by summing scores for all responses.

The degree to which a trajectory of antisocial behavior from childhood
through adulthood exists was established by way of classification of both child and
adult antisociality. Following the procedure used in Zucker et al. (in press), alcoholic
subjects who scored 10 or higher on both the Child Subscale and the Adult
Subscale of the ASB were categorized as antisocial alcoholics (AALS); these men
are also referred to as high antisocial continuity alcoholics. Men scoring below 10 on
one or both scales of the ASB were categorized as nonantisocial alcoholics (NAALS)
(Zucker et al., in press). These scores were chosen as cutoff levels based on a dual
criterion involving 3) their absolute deviation from normative levels of antisociality,
and b) their providing some parallelism to DSM-IV criteria, which requires at least
three childhood and three adulthood symptoms in order for a diagnosis of ASP to be
given (see Zucker et al., in press for more discussion on this issue). Thus, NAALs in
the study consist of alcoholics who have exhibited a pattern of low antisocial
involvement during either or both childhood and adulthood. This NAAL classification
differs from the low antisocial continuity alcoholic groups of Zucker et al. (in press)
in that it also includes those alcoholics where antisocial involvement was high in

childhood but low in adulthood. This was done in order to have less restrictive

 

34

classifications in an effort to enhance the generalizability of findings. Thus, the AAL
group used here have sustained life course antisocial involvement, and the NAAL
group is reflective of the more general variation among nonantisocial alcoholic men.
M . M” D . I E El |

Interactions during problem-solving sessions were rated according to five
categories of behavior: sharing control, positive power assertions, hostile power
assertions, avoidance, and dyadic power struggles. These categories have been
derived from a microanalytic coding scheme of couples' communication behavior
(Communication Skills Test (Floyd & Markman, 1984)) and each consists of several
more specific behaviors. For example, the category of "sharing control" consists of
behaviors such as a spouse checking whether he/she has understood his/her
partner's viewpoint and asking the partner to clarify an earlier statement (Floyd &
Markman, 1984). "Positive power assertions" include suggestions of how to solve a
problem and attempts to return a conversation to its original topic. The category of
"hostile power assertions" consists of behaviors such as insulting one another and
blaming the other for causing a problem. "Avoidance" refers to such behaviors as
making statements which are unrelated to the topic of discussion and describing
problems in an ambiguous way. "Agreement" and "accepting responsibility" are
included in this category when these behaviors are preceded by either Positive or
Hostile Power Assertions. "Dyadic power struggles" include disagreeing without
offering an explanation and complaining about a partner in response to the partner's
complaint (Floyd & Markman, 1984). The use of an observational measure of these
types of behavior in the current study allows a more comprehensive and objective

assessment of couples' interaction than would self-report measures. Furthermore,

 

35
the various behaviors included in the above categories more accurately represent the
complexity of the concepts of dominance and power struggles than several previous
studies have done by using single types of behavior as operational definitions. In
addition, ratings of statements for communication proficiency were used as an
estimate of the overall effectiveness of couples' communication (Floyd & Markman,
l

g . I' Sl'll I IIQSII

The CST was used to assess the quality of the subjects' marital interaction in

 

problem-solving sessions. The CST is an observational measure designed to 4;
evaluate verbal and nonverbal communication in videotaped problem-solving
sessions of approximately ten minutes in duration (Floyd & Markman, 1984; see
Appendix B). This measure uses as a coding unit each spouse's verbal and
nonverbal behaviors when he/she has the floor and the other spouse is listening.
Observers rate each speaker's statement as very negative (1), negative (2), neutral
(3), positive (4), or very positive (5). These ratings represent the communication
proficiency of the speaker, and are based on tracking the occurrence of a number
of more specific actions considered to be instances of either facilitative,
problem-solving behaviors (e.g., proposing solutions to problems, requesting the
statement of a feeling or opinion, giving a compliment) or disruptive, potentially
destructive behaviors that impede problem-resolution (e.g., stating opinions without
offering explanations, making global complaints about personality traits, giving a
command). Statements containing facilitative behaviors receive a rating of positive
or very positive depending upon whether they represent general facilitative

interactions or more specific problem-solving skills, respectively. In order to assign

36
the ratings of positive or very positive, statements must reflect neutral to positive
affect. Statements containing disruptive behaviors receive a rating of negative or
very negative depending upon whether they are generally nonfacilitative to resolving
problems or represent more specific actions which are destructive to
problem-solving, respectively. Negative ratings are assigned to statements made
with neutral to negative affect. Very negative ratings can also be given to !
statements made with neutral effect, but are more often associated with negative to

very negative affect (Floyd & Markman, 1984).

 

The CST ratings have been found to have an interrater reliability ranging from a
.71 to .95 which was calculated based on correlations between two coders' scores
of couples' interactions on 20 tapes, as well as an interrater correlation of .81 for
the mean of the 5-point CST scores for all statements of both husbands and wives
(Floyd & Markman; Floyd & Zmich, 1991). In addition, the CST was shown to be
sensitive to Change in couples' communication skills due to therapy (Floyd &
Markman), was able to distinguish between distressed and nondistressed couples
(Floyd, O'Farrell, & Goldberg, 1987), and revealed significantly more negative
interactions between parents of mentally retarded children than between parents of
normally developing children (Floyd & Zmich, 1991).

In order to use the CST to code videotapes of the marital problem-solving
sessions, coders were trained in six sessions during which they became familiar
with the rating categories and subcategories of the CST and their definitions (Floyd
& Markman, 1984). The coders then used the CST scoring to code five videotaped
practice problem-solving sessions. When the interrater reliability between every

coder and the investigator reached kappa = .70, training was ended. These

37
individuals then utilized the CST scoring criteria to code the actual marital
problem-solving task videotapes for this study. Interrater reliability for coding of
these tapes using the CST ranged from .60 to .97 with an average interrater
reliability of .83.

In the current study, coders not only recorded their 5-point ratings of each
statement made, but also recorded their scores for each specific behavioral event
that occurred. These specific event codes were used to create new categories of
behavior which are more relevant to the present investigation. In addition, ratings of
communication proficiency were calculated for each couple based on the summation
of all positive, negative, and neutral ratings they received relative to their total
number of coded behaviors.

For the present work, the CST codes were collapsed into five new categories
relating to dominance and power struggles. The new categories were developed
collaboratively by the author and Frank J. Floyd; however, they do not include all of
the behavior codes used by the CST. They are as follows:

W This category is based on behavior codes from the
CST categories Of Very Positive and Positive. The code is given to behaviors where
power is shared by couples. It consists of the behavior codes "checking out,"
"opinion/feeling probe," "agree," and "clarification request." These behaviors
represent attempts by partners to seek each other's opinions and advice and to
acknowledge the validity of their opinions during the problem-solving sessions.

WM Also based on the CST behavior codes
from the categories of Very Positive and Positive, this category reflects statements

of power made which have a positive effect on the couple's interactions. It

38
consists of "solution proposal," "back on beam," and "plan suggestion -
nonspecific" which represent efforts to think of ways to solve problems and to keep
the focus of conversation on relevant issues.

W. This is predominantly based on the CST
codes from the category of Very Negative, with the exception of two codes being
from the category of Negative. These statements reflect the speaker's established - !
power; however, they are made with hostile/negative intentions and have a

detrimental effect on the problem-solving abilities of the couple. This category

 

consists of the behavior codes "mindreading," "putdown," "blaming, character I:
assassination," "command/patronizing," "very negative nonverbal - short, sarcastic,
condescending," "leading question," and "opinion without rationale." These
behaviors represent attempts to make the listener feel inferior and responsible for
the couples' problems.

mm This is based on the CST codes from the categories of Very
Negative and Negative. Couple members receiving this code are not overtly hostile,
but attempt to avoid dealing with problems. It consists Of the behavior codes "off

beam," "deny responsibility, confused problem-talk," "disruptive extraneous
comments," and "avoidant question" which represent the speaker's efforts to
change the topic of conversation or hinder discussion altogether and to avoid taking
responsibility for a problem.

WWW Based on CST codes from the categories of
VeryNegative and Negative, this code indicates that the couple is involved in a

battle for power. It consists of the behavior codes " kitchen sinking - cross

complaining" and "disagree without rationale" which represent a negative response

39
by the speaker to the previous statement of the listener. The speaker responds to

the listener's complaint or comment with a complaint about the partner or by
negating it without any explanation.

Emcedure

The protocol for the parent investigation required subject families to
participate in nine sessions which lasted for a total of approximately 15 hours of
data collection. Follow-up assessments occur at three year intervals. The research
staff remained blind to the diagnostic classification of the subjects, and the
investigators received informed consents from all subjects before beginning the
study and at each new study wave (Zucker & Fitzgerald, 1991).

Relevant to the current work, as part of the nine session protocol of Wave 2,
all subjects completed the ASB, a background Demographics Questionnaire, and
also took part in a marital problem-solving task. The marital problem-solving task
was videotaped primarily from behind a one-way mirror in a university research
laboratory. The administrator of this task always was an advanced doctoral level
graduate student in clinical psychology with a substantial background in marital
therapy work. Before the task began (see Appendix C for experimenter instructions),
each subject was asked to complete a Marital Problem Areas checklist (see
Appendix D) which permitted subjects to rate 10 problem areas (e.g., money,
religion, relatives) on a scale of 0 to 5 (0 representing no problem; 5 signifying a
major problem). Two problem areas could be added and rated if a subject felt it was
necessary.

After the inventory was completed by each partner, the administrator

determined the most important problem area for the couple to discuss by selecting

 

40
the topic that had the highest shared score between partners. Subjects were then
instructed to spend the next ten minutes discussing the problem and attempting to
reach a mutually satisfying solution to the specified problem. The administrator left
the couple at this point, after making it clear that their discussion was private and
that he/she would be available in a nearby room if needed. At the end of the ten
minute time period, the administrator completed the videotaping and returned to the
room. He/she then discussed with the couple how it felt to participate in the task in

an effort to alleviate any distress that the session had caused them. For

 

circumstances in which the couple remained involved in conflictual interactions at fir
the end of the task, the administrator worked with them as long as was necessary

to resolve the conflict. On very rare occasions, when there was uncertainty that the

conflict would be raised again after the couple returned home, a follow-up phone

call was made to insure that no further intervention was required. In the running of

this procedure over the last five years, this level of conflict has been reached in less

than 3% of the protocols.

RESULTS

The basic design for the present study is a three-group comparison on the A
behavioral variables. As indicated in the hypotheses, specific variables for the study

are based on the relative frequencies of the new codes as well as on the conditional

 

probabilities of these codes occurring as sequential responses calculated with lag E
sequential analyses. Hypotheses regarding relative frequency scores were examined

for the couple and then for each member, given the possibility that the couple score

could obscure what was occurring individually. The total number of behaviors

exhibited by each couple was based on the number of codes they received from the

CST, as the five newly created categories did not include all of the CST codes.

Since the hypotheses proposed specific group differences, one-way ANOVAs with

planned comparisons were used in the data analysis.

The sample consisted of antisocial alcoholic couples (n: 22), nonantisocial
alcoholic couples (n=53), and nonalcoholic control couples ([1: 51). Demographic
data for the groups are presented in Table 1. Results indicate that the three groups
of husbands differed in age (EI2,124) =3.41, p< .05), in years of education
(El2,124) = 5.14, p< .05), and in number of separations and divorces,

E(2,124) =3.19, p< .05. No significant correlations were found for husbands' age,
education, or number of separations and divorces on any dependent measure;

therefore, none of these variables was covaried in analyzing data.

41

42

A greater number of nonalcoholic control couples were assessed at Wave 2
than AALs and NAALs (See Table 2). So as not to confound group membership and
family Iifestage differences, SES and income were calculated for the three groups at
Wave 2. There were no significant differences between groups in SE8 or income,
unlike those that have been found for the overall longitudinal study sample
(lchiyama et al., 1996; Zucker et al., in press). This may be due to the lower
statistical power of the present study, since the means for the demographic
variables are in the direction similar to those observed in these previous analyses.

The five collapsed categories that were used in testing the hypotheses were
Sharing Control (SC), Positive Power Assertions (PPA), Hostile Power Assertions
(HPA), Avoidance (A), and Dyadic Power Struggles (DPS). lntercorrelations among
those categories which were analyzed using relative frequency scores (SC, PPA, &
DPS) and conditional probabilities, as well as communication proficiency are shown
in Tables 3 and 4. Means and standard deviations for the relative frequency scores
and z-scores obtained from sequential analysis (based on conditional probabilities)

which were used in ANOVAs are presented in Table 5.

Table 1

43

Family Demographic CharacteriStics of Antisocial Alcoholic, Nonantisocial Alcoholic,

and Control Couples

 

 

AALs NAALs Controls _E
ln=22l ln=53l ln=51l
Age ..
Mean 37.55 37.14 34.87 3.41 *
(SD) (7.50) (4.94) (3.85)
Education I"
Mean 12.32 13.70 14.33 5.14* * '
(SD) (3.90) (2.15) (1.94)
Prior Separations/Divorces
Mean .38 .26 .19 3.19*
(SD) (.59) (.45) (.33)
MILES;
Age
Mean 33.07 34.65 33.55 1.47
(SD) (5.10) (4.00) (3.93)
Educaflon
Mean 13.00 13.60 13.55 .72
(SD) (1.95) (2.38) (1.72)
Prior Separations/Divorces
Mean .19 .19 .08 1.07
(SD) (.68) (.40) (.27)

 

(Table 1 continued)

 

Table 1 (continued)

Family Demographic Characteristics of Antisocial Alcoholic, Nonantisocial Alcoholic,
and Control Couples

 

AALS NAALS Controls E
(n=22l ln=53l ln=51)

 

 

Family SES (Duncan) at Wave 2:
Mean 331.41 360.53 379.26 .90
(SD) (157.75) (144.18) (128.81)
Family Income at Wave 2:
Mean $39,047 $44,169 $45,843 1.15 S
(SD) ($18,026) ($18,459) ($15,767)

Years Married:
Mean 8.20 8.91 8.91 .30
(SD) (3.75) (4.00) (3.78)

 

* p< .05; **Q<.01

45
Table 2

Study Wave of First Marital Assessment [n's and (%)l

 

 

AALs NAALs Controls

(n=22) (n=53) (n=51)
Wave 2 14 (64) 37 (70) 47 (92)
Wave 3 7 (32) 14 (26) 4 (8)
Wave 4 1 (4) 2 (4) 0 (0)

 

 

Note. Percentages are within group he column percents)

“H.11 ‘L

46

Table 3

lntercorrelations among Communication Skills Test Variables for Dominance, Power
Struggles, and Communication Proficiency

 

 

 

1. sc - I
2. Hsc .75" - '
3. wsc .78“ .18 -

4. DPS -.10 -.12 -.03 -

5. HDPS -.04 -.11 .05 .95" -

6. WDPS -.14 -.12 -.09 .95" .80” -

7. PPA .04 .11 -.04 -.10 -.05 -.13 - la
8. HPPA .14 .16 .06 -.09 -.02 -.14 .79" - F
9. WPPA -.04 .04 -.11 -.08 -.06 -.09 .88” .41" -

10. CP .44" .34“ .35**-.29* -.20* -.34** .40" .33“ .31"

11. HCP .39" .41" .19* -.27*-.23* -.29* .37" .40" .25*

12. WCP .44’” .20" .46'” -.26* -.14 -.34’* .37‘” .30" .33”

 

Note. All individual variable n's = 126;

SC = Sharing Control; HSC = Husband Sharing Control; WSC = Wife Sharing
Control; DPS = Dyadic Power Struggles; HDPS = Husband Dyadic Power Struggle;
WDPS = Wife Dyadic Power Struggle; PPA = Positive Power Assertions; HPPA =
Husband Positive Power Assertion; WPPA = Wife Positive Power Assertion; CP =
Communication Proficiency; HCP = Husband Communication Proficiency; WCP =
Wife Communication Proficiency;

*n< .001

47
Table 4

lntercorrelations among Communication Skills Test Variables for Dominance and
Power Struggles

 

 

 

WHPA WPPA WHPA
followed followed followed
by HHPA by HA by HA
WHPA followed -
by HHPA (27)
WPPA followed -.06 -
by HA (45)
WHPA followed -.36 -.27 -
by HA (20)
SC (126) -.35+ .17 -.13
HSC (126) -.25 -.03 .02
WSC (126) -.32 .26 -.16
DPS (126) .35 + -.03 -.12
HDPS (126) .24 .01 -.11
WDPS (126) .40* -.05 -.08
PPA (126) -.05 -.02 -.17
HPPA (126) ' -.03 .03 -.19
WPPA (126) -.05 -.05 -.09

 

Note. Numbers in parentheses are individual variable n's; therefore, correlation n's
are the lower of the two variable n's;

WHPA = Wife Hostile Power Assertion; HHPA = Husband Hostile Power Assertion;
HA = Husband Avoidance, Agreement, Accept Responsibility; SC = Sharing
Control; HSC = Husband Sharing Control; WSC = Wife Sharing Control; DPS =
Dyadic Power Struggles; HDPS = Husband Dyadic Power Struggle; WDPS = Wife
Dyadic Power Struggle; PPA = Positive Power Assertions; HPPA = Husband
Positive Power Assertion; WPPA = Wife Positive Power Assertion;

*p<.05; +n<.10

48

Hypothesis 1

This hypothesis stated that power struggles would occur more frequently for
antisocial alcoholic couples than for nonantisocial alcoholic couples and nonalcoholic
couples. The hypothesis was tested by using one-way ANOVAs using planned
comparisons by: a) comparing the relative frequency of Dyadic Power Struggles
across groups and b) comparing the conditional probability that wives' Hostile Power k
Assertions would be followed by husbands' Hostile Power Assertions across groups.

No group effect was present for Dyadic Power Struggles (E(2,125) = 2.04, as) and

 

the planned comparison did not support Hypothesis 1a that antisocial alcoholic F
couples would exhibit a higher frequency of Dyadic Power Struggles, 1(125) = 1.73,
as (See Table 5). However, analyses of the relative contribution of husbands versus
wives to this variable showed group differences in wives' behavior. Since the
variance for wives' DPS was not homogeneous, this analysis was run using a
nonparametric test (Kruskal-Wallis one-way ANOVA). This analysis showed that
wives of antisocial alcoholics had a higher frequency of Dyadic Power Struggles
than wives of nonantisocial alcoholics and nonalcoholics, X2(2, 126) =6.63, p< .05.
A group effect was not found for husbands, E(2,125) = 1.09, 115. Results of a
planned comparison for husbands also were not significant, 1(125)= 1.29, as.
Hypothesis 1b, that antisocial alcoholic couples would have a greater
frequency of reciprocal Hostile Power Assertions, was not supported. NO main
effect was found, E(2,26) = .29, us (See Table 5). Moreover, planned comparisons
did not support the prediction of a greater frequency of reciprocal Hostile Power
Assertions (based on conditional probabilities) for antisocial alcoholic couples than

nonantisocial alcoholic and nonalcoholic couples, 1(26) =-.57, gs.

49

W

This hypothesis stated that wives of nonantisocial alcoholics would be more
dominant than wives of both antisocial alcoholics and nonalcoholics. The hypothesis
was tested by conducting one-way ANOVAs using planned comparison by: a)
comparing the conditional probability that nonantisocial alcoholic wives' Positive
Power Assertions would be followed by husbands' Avoidance, agreement, or
accepting responsibility to that of the antisocial and nonalcoholic couples and b)
comparing the conditional probability that nonantisocial alcoholic wives' Hostile
Power Assertions would be followed by husbands' Avoidance, agreement, or
accepting responsibility to that of the antisocial and nonalcoholic couples. As
shown in Table 5, no main effect was found for Hypothesis 2a (E(2,44) = .88, as) or
Hypothesis 2b, E(2,1 9) = .44, 115. Results of planned comparisons also did not
support Hypothesis 2a (1(44)= 1.05, as) or Hypothesis 2b, 1(19) =-0.59, gs.
11W

This hypothesis stated that nonalcoholic couples would have more positive
and effective communication than either the antisocial or nonantisocial alcoholic
couples. The hypothesis was tested using one-way ANOVAs with planned
comparisons which compared ratings of communication proficiency across groups.
A group effect was present for communication proficiency (E(2,125) = 3.03,
n= .05), and a planned comparison revealed a trend in the predicted direction,
1(125) = 1.87, p= .06 (See Table 6). Additional analyses were performed in order to
assess the relative contribution of husbands and wives. These analyses revealed a
significant group effect for husbands on communication proficiency

(El2,125) =3.41, n< .05), and a planned comparison here supported the prediction

 

50

that nonalcoholic husbands would have more positive and effective communication
than the antisocial and nonantisocial alcoholic husbands, 1(125) =2.34, Q< .05.
However, no overall group effect was found for wives, E(2,125) = 2.29, [15. Thus,
the prediction that nonalcoholic couples would have more positive and effective
communication than alcoholic couples was confirmed, and the source of this effect
is attributable to differences among the husbands. Inspection of the means,
however, suggests that the major contribution to this effect is the lower
communication proficiency of the antisocial alcoholic group. In fact, a significant
main effect was found between AALs versus NAALs, E(1, 125) =3.83, p: .05.
titanium

This hypothesis stated that nonalcoholic couples would display a higher
relative frequency of positive power than would either the antisocial alcoholic or the
nonantisocial alcoholic couples. The hypothesis was tested using one-way ANOVAs
with planned comparisons by: a) comparing the relative frequencies of Sharing
Control across groups, and b) comparing the relative frequencies of Positive Power
Assertions across groups. No significant group effect was found for Hypothesis 3a
(El2,125) = .08, us) or for Hypothesis 3b, E(2,125) = .63, 115. Results of planned
comparisons did not support the prediction that nonalcoholic couples would display
a higher relative frequency of Sharing Control (1(125) =0.27, as) or Positive Power
Assertions (1(125)=-1.07, us) than antisocial alcoholic and nonantisocial alcoholic
couples (See Table 5). These nonsignificant findings also held true for the relative
contributions of husbands and wives.

After inspection of the data from Tables 5 and 6, it is clear that effect sizes

were small (DPS: r2= .03; SC: r2 = .00; PPA: r2 = .01; WHPA followed by HHPA:

51
r2 = .02; WPPA followed by HA: r2= .04; WHPA followed by HA: r2 = .05;
Communication Proficiency: r2 = .05), and the power for these tests was low
(DPS= .41; SC=.06; PPA=.16; WHPA followed by HHPA=.09; WPPA followed by
HA=.19; WHPA followed by HA: .1 1; Communication Proficiency= .58) (Cohen,

1977).

 

Table 5

52

Analyses of Variance for Relative Frequencies and Conditional Probabilities of

Dominance and Power Struggle Behaviors among Antisocial Alcoholic, Nonantisocial

Alcoholic, and Control Couples

 

 

8.2mm AAL: NAALS EQUILQIS E t'

E I . E .

DPS .030 (.054) .017 (.043) .009 (.032) 2.04 1.73
HDPS1 .025 (.057) .015 (.040) .009 (.037) 1.09 1.29
WDPS2 .035 (.057) .020 (.051) .009 (.029) 2.73 + 1.97*

SC .070 (.069) .064 (.054) .068 (.065) .08 .27
HSC .062 (.095) .062 (.070) .062 (.075) .01 .02
WSC .076 (.085) .066 (.074) .074 (.089) .17 .32

PPA .029 (.033) .041 (.046) .038 (.037) .63 -1.07
HPPA .027 (.039) .033 (.044) .039 (.042) .72 -.91
WPPA .031 (.037) .049 (.066) .037 (.046) 1.07 -.92

WHPA followed
by HHPA .662 (1.387l1.166 (1.847) .870 (1.096) 29 -.57

WPPA followed
by HA 2.015 (2.439)1.324 (2.042) .741 (2.157) .88 1.05

WHPA followed
by HA -.443 (.183) .061 (.881) .990 (3.755) .44 -.59

 

 

* Q=.05; +p<.10
HI = husbands; W2 = wives;

a=planned comparison of AALs vs. NAALs and Controls

Table 6

53

Analyses of Variance for Communication Proficiency among Antisocial Alcoholic,
Nonantisocial Alcoholic, and Control Couples

 

 

AALs NAAL: Camels E t'

Communication

Proficiency 1.899 (.316) 2.046 (.288) 2.072 (.258) 3.031 1.87+
Husbands'

Communication

Proficiency 1.882 (.335) 2.020 (.320) 2.087 (.281) 3.41* 2.34"
Wives'

Communication

Proficiency 1.916 (.352) 2.071 (.295) 2.057 (.273) 2.29 1.14

 

*p<.05;
‘p=.05;
+Q< .10;

"=planned comparison of Controls vs. AALs and NAALs

 

DISCUSSION

The results of the present study provide some support for the prediction that El
power struggles would occur more frequently for antisocial alcoholic couples than

for either nonantisocial alcoholic couples or nonalcoholic couples, and overall,

 

highlight the importance of distinguishing between antisocial and nonantisocial ,
alcoholics. Specifically, this prediction was supported by way of wives' behavior, as J
wives of antisocial alcoholics were shown to exhibit a higher relative frequency of
Dyadic Power Struggles than wives of nonantisocial alcoholics and nonalcoholics.
However, antisocial alcoholic couples did not have a higher frequency of reciprocal
Hostile Power Assertions than nonantisocial and nonalcoholic couples. The
prediction that wives of nonantisocial alcoholics would be more dominant than
wives of antisocial alcoholics and nonalcoholics, as represented by their exhibiting
either a higher frequency of Positive Power Assertions or Hostile Power Assertions
followed by husbands' Avoidance, agreement, or accepting responsibility, was not
supported.

In the realm of positive communication, the prediction that nonalcoholic
couples would have more positive and effective communication than either
antisocial alcoholic couples or nonantisocial alcoholic couples was supported.
Examination of husbands' versus wives' relative contribution to this variable
revealed that this effect was attributable to more positive and effective
communication among the nonalcoholic husbands. Further analyses exploring the

54

55

source of this effect suggested that the major contribution to this variable was the
lower level of positive communication among AALs. However, given the post-hoc
nature of this finding, it needs to be regarded with some tentativeness and it also
requires replication. At the same time, this result is consistent with self-report
differences observed by lchiyama et al. (1996) who found that AAL husbands
blamed and belittled their wives more frequently than NAAL and nonalcoholic .1
husbands. Finally, the prediction that nonalcoholic couples would display more A

positive forms of power than either antisocial or nonantisocial alcoholics, as

 

represented by their having the highest relative frequency of Sharing Control and bl
Positive Power Assertions, was not supported. These results are discussed in turn
below.

The expectation that power struggles would occur more frequently for
antisocial alcoholic couples than for nonantisocial alcoholic and nonalcoholic couples
was supported only by way of differences among wives' behavior; the wives of
AALs had a higher relative frequency of Dyadic Power Struggles than wives of
NAALs and nonalcoholics. Contrary to prediction, however, AAL husbands did not
have a higher frequency of Dyadic Power Struggles than NAAL and nonalcoholic
husbands. These results may indicate that antisocial alcoholic men have more
established power than their wives due to the more antisocial nature of their
behavior. It also may indicate that these men are less invested in their relationships,
and thus do not feel the need to struggle. The wives of antisocial alcoholics, to a
greater degree, struggle to increase their level of power when interacting with their
husbands, perhaps in an attempt to control social and/or financial problems created

by the more turbulent nature of their husbands' alcoholism as compared to that of

56
nonantisocial alcoholics. However, as power struggles were found to be greater only
for AAL wives, but contrary to prediction, not for AAL husbands, these findings
need to be replicated.

The above findings have some commonalities and some differences with
earlier studies in this area. In both the Frankenstein, et al. (1985) and O'Farrell and
Birchler (1987) studies, struggles for control were found to occur in the interactions
of alcoholic couples. However, the former study found that these struggles for
control were exhibited by the alcoholic only when he/she was intoxicated and did
not compare alcoholic couples to a control group. The latter study reported a greater
number of struggles for control for alcoholic couples than for control couples, but
did not differentiate between the frequency of husbands' versus wives' struggles for
control. The previous two studies operationalized "struggles for control" as either
the frequency of interruptions or questions asked. The validity of defining struggles
for control in this manner is questionable, as discussed earlier. The present study
used the CST's codes from the categories of Very Negative and Negative to create
the new category of Dyadic Power Struggles. This new category consisted of the
behaviors of "kitchen sinking-cross complaining," in which the speaker stated a
series of complaints about the partner or responded with a complaint about the
partner to the partner's complaint about him/her, and "disagree without rationale,"
in which the speaker negated the previous statement by the partner without an
explanation. These codes more adequately represent negative struggles for control
and their use allows a greater examination of antecedent and consequent behaviors.
However, these behaviors do not reflect established power of the wives of the

antisocial alcoholics (as other behaviors in this coding scheme were proposed to do),

 

57
but instead indicate that these wives are involved in a struggle to obtain power. As
these results have shown, wives of AALS struggle to gain power more frequently
with their husbands than wives of NAALs and nonalcoholics. According to Zucker
et al. (1994), antisocial alcoholics have more serious alcohol-related problems, an
earlier history of alcohol abuse, and more interpersonal problems than nonantisocial
alcoholics; thus, perhaps due to the more serious nature of the problems associated
with their husbands' alcoholism, these wives need to struggle more.

In contrast to the above results, the prediction that power struggles would
occur more frequently for antisocial alcoholic couples than nonantisocial and
nonalcoholic couples was not confirmed by the data on Hostile Power Assertions.
Given the nature of the coding system used here, where the collapsed CST category
consists of very negative and negative behaviors which are more reflective of the
speaker's established power, these results may indicate that neither member of the
antisocial alcoholic couple has established power and thus, cannot struggle to
increase his/her power in this manner. Instead both spouses attempting to gain
more power by berating one another, both may struggle together by exhibiting
negative "powerless" behaviors.

The second hypothesis, that wives of nonantisocial alcoholics would be more
dominant than wives of antisocial alcoholics and nonalcoholics, was not supported.
Based on the conflicting findings of earlier reports of wives' dominance of alcoholic
husbands (Paolino & McCrady, 1977) versus later studies which did not find
evidence of wives' dominance (O'Farrell 81 Birchler, 1987), it was proposed that
distinguishing between subtypes of alcoholics (AALs vs. NAALs) would help to

clarify the controversy in the literature. As the nonantisocial alcoholics have been

 

58

reported to have a later onset of alcohol problems, fewer alcohol-related problems,
and higher levels of social functioning, as well as being more trusting and less
blaming of their wives (Zucker et al., 1994; Zucker et al., in press; lchiyama et al.,
in press), it was believed that the NAAL couples would exhibit more stable patterns
of wife dominance and husband submissiveness than AAL couples and nonalcoholic
couples. However, results indicate that no difference between the levels of I]
dominant-submissive interactions exists between the three groups.

The finding that wives in the NAAL group were not more dominant than

wives in the AAL and nonalcoholic groups is parallel to earlier work reported by LJ

 

O'Farrell and Birchler (1987). As similar results were found in both the O'Farrell and
Birchler and present studies using different operational definitions of "dominance,"
more general support has been provided to dispute the earlier notion that wives of
alcoholics are dominant individuals. Wives of alcoholics may have more household
and family responsibilities as a result of their husbands' alcoholism; hence, they may
seem more "in control" of their homes. However, this was not examined by the
present study nor does this represent the same type of dominance that relates to
marital problem-solving. In addition, the current study did not explore subjective
perceptions of dominance by either partner. As such, a comparison could not be
made to Frankenstein et al.'s findings (1985) that although wives view their
alcoholic husbands as autonomous, the husbands perceived themselves as
submissive. Other studies also have shown that husbands and wives rate their
partners as being more blaming than the partners rate themselves and that a greater
discrepancy exists among antisocial alcoholic couples regarding their perception of

husbands' hostile controlling behavior than among nonantisocial and nonalcoholic

59

couples (lchiyama et al., 1996) Subjectively, husbands may view themselves as
being submissive, less hostile, or less blaming than their wives do, although
exploration of this is beyond the present study. Analysis of behavioral interactions
are not indicative of patterns of wife dominance and husband submissiveness.

The prediction that nonalcoholic couples would have more positive and
effective communication than the antisocial alcoholic and nonantisocial alcoholic a
couples, represented by higher ratings of communication proficiency, was supported
by way of husbands' communication. Examination of the relative contributions of

husbands versus wives to this proficiency score revealed that nonalcoholic i

 

husbands had more effective communication patterns than antisocial alcoholic and
nonantisocial alcoholic husbands. Furthermore, NAAL husbands were found to have
more effective communication than AAL husbands. These results support the
findings in the literature that alcoholic husbands have difficulties with effective
communication, but also point to the nonhomogeneity of the alcoholic population.
Communication difficulties have been represented by the alcoholic communicating
more effectively when intoxicated as some studies have shown (Davis et al., 1974;
Steinglass, 1981; Frankenstein et al., 1985) and by the alcoholic's drinking having
negative effects on communication as other studies have found (Ritchey et al.,
1981; Jacob & Krahn, 1988). These results also again highlight the importance of
differentiating antisocial alcoholics from nonantisocial alcoholics.

The results regarding communication proficiency are closest to O'Farrell and
Birchler's findings that, under laboratory conditions, alcoholic couples (who were not
given alcohol) exhibited less positive verbal behavior than normal control couples. In

addition, the alcoholic husbands had more responsibility-avoidant communication

60
than did control husbands. The wives of alcoholics in the O'Farrell and Birchler
study did not differ from the control wives in their pattern of responsibility-avoidant
communication. Thus, the findings of the present study provide further support for
the notion that alcoholic husbands have more disturbed communication patterns
than nonalcoholic husbands, but that wives, regardless of the alcoholic status of
their husbands, do not seem to differ in their overall communication proficiency. a
However, the O'Farrell and Birchler study did not distinguish between antisocial and "

nonantisocial alcoholics. As the current study found that NAAL husbands exhibited

 

more effective communication than AAL husbands, examining the relationship J
between alcoholic subtype and communication differences is an important element L
in developing a greater understanding of the communication patterns of alcoholics.
Whether the alcoholic husband's communication patterns contribute to or are a
result of his alcoholism (or both) cannot be differentiated in the present work,
although analyses of changes in communication patterns over time, as the
longitudinal study and these families' alcoholism progresses, will allow exploration
Of this issue.
The final prediction that nonalcoholic couples would display more positive
forms of power, as represented by having the highest relative frequency of Sharing
Control and Positive Power Assertions, than either antisocial alcoholic couples or
nonantisocial alcoholic couples was not supported. These categories reflect
behaviors where power is shared by the couple and statements of power which
have a positive effect on the couples' interactions. It may be that the codes chosen
to be collapsed into these two new categories of behavior do not adequately

distinguish the interactions of nonalcoholic couples from alcoholic couples. For

61
example, behaviors, such as agreeing with one's spouse or requesting clarification
of a statement, may be common in the interactions of all couples. It is important to
note, however, that significant positive correlations exist among positive power
(Positive Power Assertions and Sharing Control) and Communication Proficiency.
Differences found between groups for Communication Proficiency and the lack
thereof for positive power may be due to the number of codes included in these Pl
variables. Calculation of Communication Proficiency scores involved the summation

of all behaviors; whereas, Positive Power Assertions and Sharing Control consisted

 

Wit

of only three and four subcategories of behavior, respectively. On the other hand,
differences in this area may simply not be present.

In general, these results indicate that wives' power struggles exist to a
greater degree in the interactions of antisocial alcoholic couples when compared to
the interactions of nonantisocial alcoholic and nonalcoholic couples. Also,
nonalcoholic husbands exhibit more positive and effective overall communication
than alcoholic husbands, particularly antisocial alcoholic husbands. The struggle for
power by the wives of antisocial alcoholics may have contributed to their overall
poorer rating of communication proficiency as compared to wives of nonantisocial
alcoholics and nonalcoholics However, power struggles (DPS) were represented by
the occurrence of only two behavior codes; whereas, the rating of overall
communication proficiency involved the summation of all of the codes used in the
CST.

The sample used for the present study was recruited from a community
population of alcoholics who were at an early stage of family development, where

they have largely not yet have sought out help for alcohol and/or marital problems-

62
(Zucker & Fitzgerald, 1991). As the subjects were not drawn from a treatment
program, they are more representative of alcoholics in the general population. A
control group consisting of community nonalcoholics was also used. Thus, results
from this study are more generalizable than those from most previous research.
However, it should be noted that the findings of the present study may not
generalize beyond non-Hispanic, Caucasian couples who have young children. In 1
addition, women were not accepted nor rejected from the sample of Wave One of

the larger MSU-UM Study on the basis of their own levels of alcohol use. Zucker

 

and Fitzgerald (1991) reported that approximately 44 percent of these women met J
criteria for alcohol dependence or abuse. As these couples were recruited from the V
community, the fact that the alcoholic men frequently were married to women who
had alcohol problems as well seems an important consideration in generalizing
findings of studies involving alcoholic couples. The interactions of these couples
may more adequately represent the behavior exhibited by more typical alcoholic
couples, as opposed to the behavior of couples from "purer" samples used by
previous studies which have excluded couples in which both partners met criteria
for alcoholism.
One limitation of the present study may be that the newly-derived codes do
not adequately represent the intended concepts of power and dominance. These
codes operationalize power and dominance differently than has been done in other
studies. However, the coding scheme seems to be more representative of the
complexity of interactions that occur in marital relationships than measures used in
previous studies, and also allows a more objective assessment of couples' behavior

than self-report measures. It also should be noted that the frequency of specific

63

power-related behaviors exhibited by each couple relative to their total number of
behaviors exhibited was very low and may have been a factor in obtaining
nonsignificant results. Furthermore, as described earlier, effect sizes were small and
power was low. On these grounds, these findings are best considered as lower
bound indicators of the true differences that exist between groups, that can be more
fully articulated when the available Ns are larger. n

One issue relating to the present study concerns the extent to which the

interactional task challenged these couples sufficiently to provide adequate sampling

 

of marital variations under stress. Given that these families were not selected based

1.":-

on their current active problems nor on their need for treatment, the level of
expressed conflict may be somewhat lower for these couples than couples in
greater distress. Thus, stronger differences between alcoholic groups may only be
present under times of greater stress or with a task that has a greater capacity for
more immediate affective conflict arousal. In addition, as the task ran for only ten
minutes, there is the possibility that these couples were able to behave in a more
controlled manner than they would normally; it is not uncommon for conflicted
couples to be able to suppress their level of conflict while being observed. These
issues, as well as the relatively low number in the antisocial alcoholic group, may
also have contributed to the relative lack of strength of the current findings. Even
so, those that existed were as predicted.

The finding that nonalcoholic husbands are able to communicate more
effectively than alcoholic husbands, particularly antisocial alcoholic husbands, is
relevant for clinicians, especially marital therapists. The lower level of positive

communication for antisocial alcoholic husbands, as compared to control husbands

64
and to NAAL husbands, suggests that building positive communication skills should
be an important part of marital therapy, particularly for antisocial alcoholic couples.
As wives of antisocial alcoholics were shown to exhibit a higher frequency of power
struggles in their interactions than wives Of nonantisocial alcoholics and
nonalcoholics, additional evidence has been provided for the importance of
differentiating between alcoholic subtypes. Developmental trajectories from
childhood through adulthood involving antisocial behavior differ between the two
groups of alcoholics (Zucker et al., 1994), and the more serious nature Of antisocial
alcoholics' problems may influence their wives interactions with them. It is clinically
relevant to note the distinction between antisocial alcoholic and nonantisocial
alcoholic couples in the frequency of wives' struggles for power, and to recognize
that power struggles may be more of a presenting or underlying problem in therapy
for antisocial alcoholic couples. The clinical significance of this finding indicates that
approximately 3-4% of the statements made in a ten minute period by wives of
antisocial alcoholics signify struggles for power. This may be seen in therapy, for
example, as a wife making a series of complaints about her husband, complaining
about him in response to a complaint about her, or disagreeing without any

rationale, such as "No, you're wrong. Although these verbalizations reflect a low
percentage of the total number of statements made by these women, they do occur
more frequently for wives of antisocial alcoholics than for wives of nonantisocial

alcoholics and nonalcoholics. The fact that these wives feel the need to struggle for
power may be reflective of inequity in their marital power structures. This inequity in

levels of power, in turn, may cause these women to experience lower levels Of

marital satisfaction. Overall, these findings provide further support for the notion

65
that antisocial alcoholic couples have more turbulent marital interactions (Zucker et
al., in press).

Future research should employ additional coding schemes of interactional
behavior to evaluate differences in the frequency of struggles for power and
dominance between antisocial and nonantisocial alcoholics. It is also important to
compare results from interactional data with that from self-report data (e.g., a
regarding marital satisfaction) in order to evaluate differences between more

Objective ratings of couple's behavior and the subjective experiences of each

 

partner. As discussed earlier, these objective and subjective views can be quite J
iv

I

discrepant (Frankenstein et al., 1985). Additionally, the use of larger sample sizes
will be helpful in establishing greater statistical power. Moreover, the more frequent
use of community-recruited samples is needed to adequately compare future results
to the current findings. A greater level of focus should be directed toward the
relationships of alcoholic couples who are not seeking treatment; thus, future
findings would more likely generalize to the larger segment of alcoholic couples in

the general population.

APPENDICES

 

Appendix A: Antisocial Behavior Checklist
Zucker & Noll (1980)

Many of us have had adventures during our lives...times that were exciting and

carefree, even though they may have been a bit impulsive or happy-to-lucky. Please
read each of the following items. Indicate (with a check) if you have ever done any

of the following activities and how often.

NEVER

RARELY
SOMETIMES
OFTEN

NEV

I

RARELY

OMETIMES

OFTEN

You have never done this

Once or twice in your life

Three (3) to nine (9) times in your life
More than ten (10) times in your life

 

l

1. Skipped school with a legitimate excuse for more than 5
days in one school year.

 

 

2. Been suspended or expelled from school for fighting.

 

3. Been suspending or expelled from school for reasons
other than fighting.

 

Lied to a teacher or principal.

 

Cursed at a teacher or principal (to their face).

 

Hit a teacher or a principal.

 

Repeated a grade in school.

 

9°.“9’9‘P

Taken part in a gang fight.

 

9. “Beaten up” another person.

 

10. Broken street lights, car windows, or car antennas just
for the fun of it.

 

1 1. Gone for a ride in a car someone else stole.

 

12. Teased or killed an animal (like a dog or cat) just for the
fun of it.

 

13. Defied your parent's authority (to their face).

 

14. Hit your parents.

 

 

 

 

 

 

 

15. Cursed at your parents (to their face).

 

66

 

  

OMETIMES l

l

 

 

 

16. Stayed overnight without your parent's permission.

 

17. Run away from home for more than 24 hours.

 

18. Lied to your parents.

 

19. Snatched a woman’s purse.

 

20. Rolled drunks just for the fun of it.

 

21. Shoplifted merchandise valued over $25.

 

22. Shoplifted merchandise valued under $25.

 

23. Received a speeding ticket.

 

24. Been questioned by the police.

 

25. Taken part in a robbery.

 

26. Taken part in a robbery involving physical force or a
weapon.

 

27. Been arrested for a felony.

 

28. Resisted arrest.

 

29. Been arrested for any other non-traffic police offenses
(except fighting or a felony).

 

30. Been convicted of any non-traffic police Offense.

 

31. Defaulted on a debt.

 

32. Passed bad checks for the fun of it.

 

33. Ever used an alias.

 

34. Gone AWOL from the military.

 

35. Received a bad conduct or undesirable discharge from
the military.

 

36. Performed sexual acts for money.

 

37. Engaged in homosexual acts.

 

 

 

 

 

 

38. Had intercourse with more than one person in a single
day.

 

 

 

68

I

 

 

 

c: >- (I) Z
LU r)“ LIJ DJ
5 c: 2 I:
2 g E o
2
O
L==JT (D 1 :I
39. “Fooled around” with other women/men after your were
married.
40. Hit your husband/wife during an argument.
41. Lied to your spouse. 1

 

42. Spent six months without any job or permanent home.

 

43. Been fired for excessive absenteeism.

 

 

44. Been fired for poor job performance (except a
absenteeism). if

I .

 

45. Changed jobs more than 3 times in one year.

 

 

46. Lied to your boss.

 

 

 

 

 

 

Thank you very much for your cooperation.

Appendix B: CST Codebook
CST Code Book

Communication Skills Test (CST)
Scoring Criteria

Frank Floyd & Howard Markman, 1984
Prepared by Glenn Phillips, September, 1990

The Communication Skills Test (CST) was designed to be a relatively
economical observational measure of spouses' communication behaviors that
reflect their communication proficiency when solving problems together. The
system is set up to evaluate the content and affect of partner's statements
and to assess whether each statement represents an instance of the use of
either positive, facilitative communication behavior or negative, disruptive
communication behavior. This system was developed to be used by researchers
and clinicians as an alternative to more time consuming microanalytical
measures.

This system was intended to evaluate the same constructs as the Marital
Interaction Coding System (MICS; Weiss, Hops & Patterson, 1973) and the
Couples Interaction Scoring System (CISS: Gottman, 1979; Notarius & Markman,
1981). The CST borrows specific codes from these measures because they are
the most frequently used and the most carefully researched systems available.

The CST differs from many other systems in that it uses an entire
statement as the coding unit. These statements are rated on an interval scale
Of measurement that reflects how facilitative that statement is. The MICS in
contrast assigns behavioral units, several Of which can occur in one
statement, to nominal categories of positive, negative, or problem-solving
behaviors.

The system is used by assigning a rating to each statement, defined as
all behaviors that occur between floor switches. The statement is rated
very positive, positive, neutral, negative, or very negative depending on how
facilitative the statement is determined to be in meeting the problem-solving
goals of the conversation. The choice of a rating is based on a set of
guidelines about levels of problem solving proficiency outlined in this
manual.

Regarding validity, the relative frequency scores for categories of
behaviors and the overall mean scores were shown to be sensitive measures of
pre- to posttreatment improvements in communication skills for premarital
couples who completed a cognitive/behavioral relationship enhancement
intervention (Floyd & Markman, 1984; Markman, Floyd, Stanley & Storaasli,
1988). These scores showed only modest correlations with couples' own
evaluations of the positiveness of their interactions (Floyd & Markman, 1984;
Floyd, 1988), and with scores obtained from a different observational coding
system that examines more discrete, molecular events (Floyd, O'Farrell &
Goldberg, 1987). However, CST scores were moderately correlated with measures
of marital satisfaction and divorce potential for a heterogeneous sample Of
married couples, and they successfully discriminated distressed couples from
happily married couples (Floyd et al., 1987). These relationships were
particularly strong for indices of negative, disruptive behaviors. Most
recently, mean CST scores have been shown to correlate with marital
satisfaction for Older married couples (Floyd & Haynes, 1989), and the CST
coded marital discussions were shown to be less positive and more negative for
couples raising a mentally retarded child as opposed to those with_typically
developing children (Floyd & Zmich, 1990).

69

 

 

7O

CST Code Book

References

Floyd, F.J., & Markman, H.J. (1984). An economical Observational measure
of couples' communication skill. Jgurnal of Congulting and
W. 9.2.. 97-103-

Floyd, F.J., O’Farrell, T.J., & Goldberg, M. (1987). Comparison of
marital observational measures: The Marital Interaction Coding
System and the Communication Skills Test. Jou na 0 '

d c 'n 1 p , §§, 423-429.

Floyd, F.J. (1988). Couples' cognitive/affective reactions to

communication behaviors. gggrnal gf Marriage agd ghg Family, §g,
523-532.

Markman, H.J., Floyd, F.J., Stanley, S.M., & Storaasli, R.D. (1988).
Prevention of marital distress: A longitudinal investigation.

gggrngl gf gongglting agd gliniggl Esxghglggx. ié. 210-217.

Haynes, S.N., Floyd, F.J., et al. (1992). The Marital Satisfaction

Questionnaire for Older Persons. ngghglggigal Agggsgmgng: A
Journal of Consulting and Clinical Psycholggy, g, 473-482.

Floyd, F.J., & Zmich, 0.8. (1991). Marriage and the parenting
partnership: Perceptions and interactions of parents with mentally
retarded and typically developing children. Child Develgpmeng,
gg, 1434-1448.

Egg! POSITIVE (5)

1) Summarize Other/both
2) Checking out

3) Opinion/feeling probe
4) Solution Proposal

5) x,r,z - feedback

6) Back on Beam

7) Metacommunication

8) Validation

NEUTRAL 3

1) Problem-talk
2) Question
3) Information

VERY NEGATIVE (l)

1) Off Beam

2) Kitchen sinking - Cross
complaining

3) Mindreading

4) Putdown

S) Summarizing self— no

clarification needed/requested

6) Blaming
7) Character Assassination
8) Deny Responsibility -

"Yes, but..."
9) Command/patronizing

10) Very negative nonverbal
sarcastic,

voice tone - short,
condescending
11) Avoidance/Withdrawal

71

CST Code Book

CST Scoring Criteria

POSITI 4

Feeling statement

Agree

Disagree with rationale

Plan suggestion - nonspecific

Compliment

Clarification Request

Accept Responsibility

Empathy

Summarize self - concise or to
clarify

10) Humor
11) Positive nonverbal

a) face - smile, laugh

b) voice tone - caring, warm,
empathic, concerned,
affectionate, cheerful

c) positive physical contact
- touch, pat

NEGATIVE 2

1)
2)
3)
4)
S)
6)

7)

Confused problem-talk
Opinion without rationale
Disagree without rationale
Disruptive extraneous comments
Avoidant question
Negative nonverbal
a) face - frown, sneer,
glare, angry, cry,
no eye contact
b) voice tone - disinterested,
cold, impatient, clipped,
whining, angry, gruff
c) body position - rude
gestures, turn away
Leading question

72

CST Code Book

VERY EQSITIVE (5)

Very Positive behaviors are demonstrations of the specific positive
communication/problem-solving skills outlined by Gottman, Notarius, Gonso, &
Markman, 1976). For a statement to receive a very positive it must be given
with neutral to positive affect.

1) §gmmgrigigg_9§hgg£§ggh: A statement in which the speaker either

summarizes his/her partner's previously expressed statement or opinion, or t,
summarizes the couple's conversation is rated "very positive". Note that I]
"summarizing other/both" statements are nonjudgmental, such that a statement .
of agreement or disagreement, followed by a summary of the point being judged, 5

is not rated "very positive" (see "Agree/disagree With/Without Rational").
Nevertheless, following a summary statement, the speaker may offer an opinion.

Examples (Bx):
"So we agree that Mary is the cause of the problem."
"Sounds like you're saying that it really upsets you "
when I ignore you at parties." i
"So you think that Mary is the cause of the problem.
Well, I disagree. I think that ..."

 

IE.-

Not "Summarizing Other/Both":
"I agree with that. Your dad is kind of close-
minded." (Agree)

2) Checking Cup: .A statement requesting the partner to verify that the
speaker correctly understands the partner's viewpoint, is rated "very
positive". In most cases, checking out statements follow summaries, therefore
the statement would be rated "very positive" anyway. However, checking out
also includes questions about the speakers' meaning that incorporate an
implied summary.

Ex:
"We agree that Bill is wrong here. Is that right?"
(Summary and checking out)
"So you think we should put off buying a new car for
a year?” (Checking out with implied summary)

Not "Checking Out":
"So, what is your opinion again?" (Clarification
Request)

73

CST Code Book

O inion Feelin Probe: A request in which the speaker asks the listener
to state a feeling, opinion, or any non-factual information. Also any request
for acknowledgment of or expansion on the speaker's viewpoint is considered an
opinion probe and is rated "very positive". However, request for restatement
or clarification of an opinion are not "opinion/feeling probes" (see
Clarification Request), nor are statements to the effect of "I'm right aren't
I".

Ex: '

"I think John is the cause of the problem, do you
agree?"

"What time do you want to leave tomorrow?"

"I said that I think Jack is the cause of the
problem, do you see what I mean?" (Request for
acknowledgment/expansion)

Not "Opinion/Feeling Probe."
"You just wanted me to leave, didn't you“
(Mindreading or empathy)
"John is the cause of the problem, right?” (Problem
talk)

4) Solution Propgsal: A proposal that offers a realistic way to solve or
diminish an existing problem, including specific suggestions about behavioral
change, is rated "very positive". These proposals commonly occur late in the
interaction when the problem has been discussed in depth. They can be
unilateral attempts to give advice.

Bx:
"So why don't we both agree to save $20 out of our
checks each week?"
"How about if we allow the kids to only watch T.V.
after they have finished their homework?"
"It seems you would be better off to put 10% of your monthly
check in an investment account."

Note: See "Plan Suggestion" and "Problem Talk" for
examples of solutions which are presented too
early or which are actually part of an opinion
statement, rather than a solution proposal
about a problem which has already been addressed.
Note in particular, statements of complaints or problems are
sometimes stated in the form of suggestions for change, but
are not really sincere or helpful proposals about how to solve
a problem that has been discussed. Often, these are "problem
talk" or "blaming" statements:

Not "Solution Proposal":
"I want you to help out with the housework every night
after work." (Problem Talk)
"The problem is that you need to cut down your spending
by 10% on a weekly basis." (Blaming)

74

CST Code Book

5) X,Y,z - Feedback: A statement in which the speaker identifies a specific
behavior of the listener, and expresses his/her feelings in response to that
behavior is rated "very positive". The statement can refer to either an
annoying behavior to which the speaker has a negative reaction, a pleasing
behavior to which the speaker has a positive reaction, or a suggested
behavioral change to which the speaker would have a positive reaction. ”X,Y,z
- feedback" statements must be distinguished from "Compliments", which either
include no reference to the speaker's emotional response, or are not in
reference to a specific behavior.

Bx:
"When you make jokes about me at parties, I feel
really angry."
"It makes me feel good to see you help me keep the
house clean."

6) §§95_Qg_§§§m: A statement that acknowledges that the conversation has
gotten off the subject and attempts to return the discussion to the topic, is
rated "very positive". If the speaker is referring only to his/her own
current statement as being off the subject, the statement does not receive a
"very positive" rating for "back on beam".

Ex:
"We're getting off the subject. We need to talk
about..."
"But that is not the point, we need to decide. ."

7) Mgtacommunigation: A statement that is a comment about the flow, style,
or course of the interaction, stated with neutral to positive affect, is rated
"very positive". These statements may be either past or future oriented.

Ex:
"It seems like we're spending too much time on the
details and missing the main point."
"I feel like we've really accomplished a lot in this
conversation."
"I think we're getting no where with this.“

8) Validatign: A statement indicating that the speaker accepts and
understands a previously stated feeling or opinion of the listener, is rated
"very positive". There is no need for a summary statement or a statement of
agreement or disagreement in a "validation" statement.

EX:
"I can see how you would feel nervous about that."
"I understand how you could see my drinking as a
potential problem."

75

CST Code Book

PQSITIVE (4)

Positive behaviors are those that are generally facilitative interactions, but
are not any of the specific skills included in the very positive category.

For a statement to receive a positive code it must be given with neutral to
positive affect.

1) Feeling Statement: A statement that reports on a past, present, or
future emotional reaction of the speaker which is directly related to the
topic of the discussion is rated "positive". Reports of feelings that are
unrelated to the topic, or are related to an ”off beam" topic are considered
"neutral" (see "Information"). Also, statements that begin with "I feel
that...", followed by a statement of an opinion rather than a report of an
emotional reaction are considered "neutral" (see "Problem Talk"). Sometimes
the speaker fails to label the feeling, but makes the statement along with
nonverbal cues that clearly communicate the feeling. This is particularly
true for instances of nervousness, hurt, or worry. These statements are also
coded as feeling statements and rated "positive”.

Ex:
"I got really angry when Sherry said that."
"This conversation is making me upset."
"It really worries me when Billy comes home late."
(said with obvious signs of anxiety)

Not "Feeling Statement":
"I feel that Burt is the problem here." (Problem
Talk)
"When you do that, I feel really happy." (x,Y,Z -
Feedback)

2) Agree: A statement of agreement, which may or may not include rationale
for the agreement. No matter how emphatic the agreement, it is rated
positive. .

Ex:
"Yes, you are absolutely right."
"Yeah, Jim does cause part of the problem because he
just won't listen."

Note: These can include short agreements or acknowledgements
such as "Yea", "Uh huh", or head nods, if significant to the
conversation.

76

CST Code Book

3) Disagree With Rational: A statement of disagreement which includes the
reason for the disagreement is rated "positive". It is important to note that
the rationale must include reference to why the speaker disagrees. These are
statements that are facilitative to the conversation even though they are
disagreements.

EX:
"I disagree, I don't think we see my mother too
often. Once a week isn't too often..."
"I would rather see our budget organized quarterly because..."

Not "Disagree With Rational":
"No, I don't think it was all Brian’s fault."
(Disagreement Without Rationale)

4) Plan suggestion - nonspecific er early: A proposal about a realistic way

to solve or diminish a problem situation, presented in a vague, non-specific
way is rated "positive". Also included here are proposals given early in the
conversation before the topic has been fully explored. This does not include
opinionated proposals. (see "Problem talk") The plan can be for both, or it
can be offering advice to the other spouse.

EX:
"Maybe we need to talk about this."
"Let's try to spend less money."
"Why don't you try coming home earlier?"

5) Compliment: A statement complimenting the listener is rated "positive".
Compliments can be general, or can refer to specific behaviors. This does not
include compliments given in a sarcastic way. (see "Very negative nonverbal)

Ex:
"I think you handle your mother well."
"I like the way you said that."

6) Clarifieation Regpest: A request for the listener to restate, explain,
or clarify a previous statement is rated "positive". Note that a "
clarification request" must be distinguished from a "summarizing other"
followed by a "checking out" statement, although both statements may serve the
same function for the speaker. A request for additional inflormation would
not fall under "Clarification Request."

Ex:
"What do you mean by trying too hard?"
"So, what is your opinion again?"
"I don't understand what you mean, could you make it
more clear?"

Not "Clarification Request":
"So you think Burt is the cause of the problem.
right?" ("summarizing other" and "checking out")
"In what ways am I neglecting my responsibilities?" (opinion probe)

77

CST Code Book

7) Accept Responsibility: A statement that acknowledges the speaker's role
in the current conflict, as previously pointed-out by the listener is rated
positive. A " yes, but" type statement (see "avoid responsibility"), in which
the.Speaker admits to a behavior, but denies responsibility is not rated
positive. .

Ex: »
."Ok, you're right, I don't keep the kitchen clean."

"Yes, I can see how that wasn't good on my part."

"I guess I could try to be more flexible about spending money."

Not "Accept Responsibility":
"Yes, I don't always wash my dishes, but I don't
think they need to be washed every time."
(Yes, But)

8) Empathy - mipdreading with pesitive affect: A statement that attributes

a previously unstated emotional response to the listener, stated with a
positive voice tone (e.g., caring, warm, empathic, etc.), is rated "positive".
"Empathy" also includes statements that imply emotion if stated in an empathic
way. "Empathy" statements followed by denial by the listener are nonetheless
considered "positive".

EX:
"It seems that you feel really hurt when I do that."
"Sounds like that was really hard for you to say."
"I think that you work too much and its a lot for you to handle
at this time.“

9) Summarize Self - concise or to clarify: A precise restatement of the

speaker's opinion, which clarifies a previous statement, or summarizes the
speaker's point of view, is rated "positive".

Ex:
"So, I guess I'm saying that I think Matt is the
cause of the problem."
"I guess what I'm saying is, that I could go either
way on this one."

10) Humor: A statement that appropriately injects humor into the
conversation, without being disruptive or off-beam, is rated "positive".
Inappropriate or disruptive humor statements however receive a "negative"
rating (see "disruptive extraneous comments"). To evaluate whether a humorous
statement is positive or negative the context in which the statement occurs
must be evaluated. Positive humorous statements usually are attempts to
affiliate, and they bring the partners together in a positive way. Disruptive
humorous statement often are attempts to avoid discussing the problem at hand.

78

CST Code Book

11).Positive Nonverbal: Facial expressions. tones of voice, and body
positions or gestures that convey positive emotional expressions are rated

"positive"- If paired with negative eentent. seoge the statemept in
accordance with the content.

"Positive Nonverbal" expressions include:
a) face - smile, laugh: note this does not include
' nervous laughter
, b) voice tone - caring, warm, empathic, concerned,
affectionate, cheerful, etc.
c) positive physical contact - touching, patting,
hugging, etc.

Note: Mixtures of positive and negative nonverbal
expressions are evaluated hierarchically with
facial expressions considered first, voice tone
next, and body position last. For example,
impatient voice tone, along with a patting
movement is rated "negative" (see "negative
nonverbal").

79

CST Code Book

NEgsz'AL ( 3 )

Neutral behaviors are those that meet the minimum demands of the
problem-discussion task. These statements are made with neutral, or only very
mildly positive or negative affect. '

1) Prgblem Ielk: All statements defining a problem or offering an opinion
with rationale about a problem, stated with neutral or mildly positive affect
(e.g., pleasant voice tone, eye contact, forward lean), are rated "neutral”.
A concise problem statement is rated neutral unless the statement clearly
contains one or more elements needed for "positive" or "very positive"
ratings. However, a very confused statement may be rated "negative" (see
"Confused problem talk"). Also included in "problem talk" are plans that are
offered before a topic has been thoroughly discussed, or plans that are
opinionated. These plans are included here because they are one-sided in
there viewpoint rather than being reasonable solutions that address both
persons' concerns.

Ex:
"The way I see it, Nancy is the cause of the problem
because..."
"I think we spend too much money on unnecessary
things, like..."
"The problem is that Jimmy won't behave, and I think
we ought to spank his butt when he doesn't."

2) Quespion: A question that is neither a clarification request nor an
opinion probe is rated "neutral". This category generally includes questions
where the answers are factual information.

Ex:
"What time do we have to leave?"
"What did your mother say about my job?"

3) Information: Any statement of fact that is either directly or indirectly
related to the current conversation is rated "neutral". Nondisruptive
extraneous comments are also rated neutral.

Ex:
"I only spent $10 at the grocery yesterday."
"I didn’t pick up the mail today."

80

CST Code Book

NEGATIVE (2)

Negative behaviors are either nonfacilitative or disruptive to reaching a
successful resolution for the problem under discussion. These statements are
made with neutral to negative affect.

1) ggnfuseg preplem talk. A statement about a problem, or an opinion about
a problem, stated in a confusing, unclear, or ambiguous manner is rated
"negative".»

Ex:
"Well...I don’t know...it's sort of like that,
but not really."

2) Qpinion wippout Rationale: A statement of opinion or belief, neither
preceded nor followed by a verbal explanation, is rated "negative". These are
blunt, dogmatic, and close-minded statements that the speaker seems to feel
need no justification. These statements frequently are paired with mildly
negative nonverbal cues such as clipped speech or gruff demeanor.

3) Disagree without Rationale: A statement of disagreement, neither preceded
nor followed by an explanation, is rated "negative". In these statements the
speaker is being dogmatic and not explaining why s/he disagrees.

Ex:
"No, you're wrong."
"I disagree, I don't think Tommy is the problem."

4) Dierpptive expraneous comments: A comment that is unrelated to the topic
of conversation, and that serves to disrupt the exchange is rated "negative".
Inappropriate humorous statements that tend to take the conversation off beam
are also rated "negative".

Ex:
"Oh, look, the ceiling is cracked."
"What are we having for dinner?"

5) Avoidant Question: A question, that attempts to take the focus off the
speaker and place it on the listener is rated "negative". These questions are
often given in response to probes, and thus are ways of avoiding a direct
reply.

Ex:
F: "What do you plan to do the next time Burt acts
up? II
M: "What do you plan to do about it?"

81

CST Code Book

6) Negative Nonverbal: Facial expressions, VOice tones, and 965tures or bOdY
positions that convey negative emotions are rated "negative".

Negative nonverbal expressions include:
a) face - frown, sneer, glare, angry expression,
cry, no eye contact.
b) voice tone - disinterested, cold, impatient,
_ clipped, whining, angry, gruff
c) body position - rude gestures, pointing

See note under "Positive nonverbal."

ea in , ue ion: These are statements of opinion in question form;
questions in which the speaker is leading the other to an implied answer.
These are not genuine attempts from the speaker to find a answer or
information from the listener. They attempt to put words in the listener's
mouth, and are frequently paired with mildly negative nonverbals.

Ex:
"Don't you think it's reasonable to help me out with this?"
"So, you should have to do the dishes every evening, right?"

82

CST Code Book

VEBY NEGATIVE (1)

Very negative behaviors are demonstrations of specific negative actions that
are destructive to problem-solving. These statements can be conveyed with
neutral affect, but they frequently involve negative to very negative affect.

1) fo beam: A statement that is either tangentially related or unrelated to
the topic of conversation, serving to move the conversation away from the
issue at hand is rated "very negative".

2) Kitchen einking - Cross cemplaining: A series of complaints about the
partner ("kitchen sinking), or a complaint about the partner given in response

to the partner's complaint ("cross complaining").

Bx:

F: "I don't think you should spend that much money
on fixing the car."

M: "Well, you spend a lot of money on things you
want to do. Besides, you do things I don't
agree with, like buying lots of clothes."

F: "Well, you're always buying new tools."

Note: The last two statements would receive a "very
negative" code.

3) Mindreading: An attribution of an opinion, emotion, or motive (usually
undesirable) to the listener, which has not been previously expressed by the
listener, is rated fivery negative". Mindreading statements are often hostile
and accusatory. Note that, in many cases, it is unclear whether the speaker
is referring to a previous incident or statement by the partner that occurred
prior to the laboratory interaction tasks, or whether the speaker is making an
assumption about the partner's opinion, emotion, or motive. Only statements
that clearly reflect assumptions (often indicated by the listener's reaction)
are rated very negative.

Ex:

"Well, you just think that you don't have to worry
about how I feel!"
"I think you do that because you hate my parents."

4) Putdown: A statement that is directed toward the listener and is intended
to make the listener appear inferior to the speaker is rated "very negative".

Ex:
"You're just no good at handling money."
"I'd starve if I had to eat your cooking every day."

83

CST Code Book

5) Summarizing-self. No clarification needed/requested: A repetition of the
speaker's point of view which adds no additional information, is not requested
by his/her partner, and displays a failure to respond to the partner's
concerns is rated "very negative". This type of statement suggests a dogmatic
approach on the part of the speaker, an unwillingness to compromise or
actively work on an area of disagreement.

6) Blaming: Attributing responsibility to the listener for a problem or
disagreement is rated "very negative". Note: These statements place all fault

on the listener.

EX:
"This wouldn't have happened if you would have been
nicer."
"It's your fault for spending too much money."

7) Charecter aesaseination: A global statement of complaint which attributes
a negative personality trait to the listener is rated "very negative".
Statements that include the words "always" or "never" in reference to the
listener's behaviors or beliefs are included in this category.

Ex:
"You never do anything I ask you to."
"Why do you always ignore me when I’m talking to
you?"
8) Deny Responsibility - "YesI but...": A statement, usually made in

response to a complaint by the partner, in which the speaker attempts to
remove all responsibility for a problem from him/herself. Also included in
this category are statements that superficially acknowledge the validity of a
problem, then offer excuses that dismiss the criticism ("Yes, but").

Ex:
M: "I think you spend too much money on unneeded
things."
F: "Yes, but I try to save what I can."
"It's the kids fault that the house always messy."

9) gommand - Patronizing: A statement that directly instructs the listener
to do something, without discussion is rated "very negative". Note: The
statement should be made with at least "negative nonverbal" voice tone.

Ex:
"You just leave Howard alone!"
"Don't ever touch my tools again."
"Well, you’d better just get over this problem."

84

CST Code Book

10) Very Negative Nonverbal: An emotional expression that disrupts the
problem-solving interaction or serves as a nonverbal put down is rated "very
negative".

Very negative nonverbal behaviors include:
Short, sarcastic, or condescending voice tone.

 

 

Appendix C: Experimenter Instructions for Marital Problem-Solving Task

After couple has completed Marital Problem Areas checklist, select an issue
for the couple to discuss, and orient them to the protocol as follows:

"What we want to do in these next tasks is to continue to get more
information on your family in two other ways. We're going to ask you to give us an
idea of how you solve problems, in a few minutes as a whole family, and right now
as a couple. These exercises will help us understand how childhood is the same or
different across different types of families, or different at different times even in the
same family, for example, during a period when a family member has a number of
health problems as compared to when it has none."

(Additional/optional introduction of task for couples asking for further explanation):

"What you just did with (target child) was a way of showing us one very
important part of your child's life within the family. Although it is certainly
important for us to get this sense about play and learning between your son and the
two of you, there is more to family life than that, because the family is larger than
just the child and one of his parents. Even though our interest is ultimately about
your child, we ask you to show us how you solve problems, first as a couple,
because it is clear from the work of other family researchers that the parents are in
charge of the family."

Ask the couple to discuss and resolve the issue you have picked from the checklist:

"I would like you to spend the next ten minutes discussing (the topic). I will
be in the next room and will return at the end of this time period. While I'm gone,
please raise issues on both sides of the problem. Also try to think of different ways
to solve the problem. Try to reach a solution you can both agree should make the
situation better. O.K., any questions? (If there are questions, simply repeat and
rephrase the instructions; if questions persist, simply say 'Just go about this the
way you would typically solve any problem") See you in ten minutes."

85

Appendix D: Marital Problem Areas Checklist

Listed below are some areas that often cause problems for couples. Please
rate how important each of these problems is in your relationship 31.111154103- Make
your rating on the 0 to 5 scale beside each item.

A 0 indicates that the area is not a problem at all in your relationship right
now; a 5 indicates that it is a major problem right now; and a 3 indicates that it is a
middle level problem right now.

1. Money 0 1 2 3 4 5
2. Communication 0 1 2 3 4 5
3. Relatives O 1 2 3 4 5
4. Sex 0 1 2 3 4 5
5. Religion 0 1 2 3 4 5
6. Recreation 0 1 2 3 4 5
7. Friends 0 1 2 3 4 5
8. Alcohol and drugs 0 1 2 3 4 5
9. Children 0 1 2 3 4 5
10. Jealousy O 1 2 3 4 5

Please write in any other problem areas that you may feel are relevant to your
relationship.

11. O 1 2 3 4 5

 

12. O 1 2 3 4 5

 

(Adapted from F. Floyd, 1986)

86

 

REFERENCES

REFERENCES

Aida, Y & Falbo, T. (1991). Relationships between marital satisfaction,
resources, and power strategies. W. 24(1-2), 43-56.

Babor, T. F., Hofmann, M., DeiBoca, F. K., Hesselbrock, V., Meyer, R. E.,
Dolinsky, Z. S. Rounsaville, B. (1992). Types of alcoholics, l: Evidence for an
empirically derived typology based on indicators of vulnerability and severity.

Wm 42, 599- 608.

Baker, M. P., Blampied, N. M., 8: Haye, L. (1989). Behavioural marital
therapy for alcoholics: Effects on communication skills and marital satisfaction.

W. 613-4), 178-186.

Ball, F. L. J., Cowan, P., & Cowan, C. P. (1995). Who's got the power?
Gender differences in partners' perceptions of influence during marital

problem-solving discussions. EaijLELchss, BA, 303-321.

Beach, 8. R. H. &Tesser, A. (1993). Decision making power and marital
satisfaction: A self— evaluation maintenance perspective. WM
91111911351:thch 12(4) 471-494.

Becker, J. V. & Miller, P. M. (1976). Verbal and nonverbal marital interaction

patterns of alcoholics and nonalcoholics. WWW, 31(11),
1616- 1624.

Berscheid, E. & Peplau, L. A. (1983). The emerging science of relationships.
In H. H. Kelley, E. Berscheid, A. Christensen, J. H. Harvey,T. L. Huston, G.
Levinger, E. McClintock, L. A. Peplau, &D. R. Peterson, Wing”.
1- 19). New York: W. H. Freeman and Company.

Blood. R. 0. & Wolfe, D. M. (1960). WWW
marfledex’mg. Illinois: The Free Press.

Bloom, B. L, Asher, S. J., &White, S. W. (1978). Maritaldisruption as a
stressor: A review and analysis. WW, 35(4), 867- 894.

Burman, B. 81 Margolin, G. (1992). Analysis of the association between
marital relationships and health problems: An interactional perspective.

WW. 112(1), 39-63.

87

88

Chiles, J. A., Stauss, R. S., 8: Benjamin, L. S. (1980). Marital conflict and
sexual dysfunction in alcoholic and non-alcoholic couples. W
mm. 1.31: 266-273-

Cloninger, C. R. (1987). Neurogenetic adaptive mechanisms in alcoholism.

5.918.093: 2.3.5: 410-416-
Cohen. J. (1977). W

New York: Academic Press.

Davis, D. l., Berenson, D., Steinglass, P., & Davis, S. ( 1974). The adaptive
consequences of drinking. Psychiatry, 31, 209-215.

Dunn, N. J., Jacob, T., Hummon, N., & Seilhammer, R. A. (1987). Marital
stability in alcoholic-spouse relationships as a function of drinking pattern and

location. JoumaloLAonormaLEsxoholoox. 9.612). 99-107.

Feighner, J. P., Robins, E., Guze, S. B., Woodruff, R. A., Winokur, G., &
Munoz, R. (1972). ArommoLGonoLaLEsxobiatu 25. 57-63.

F'OYdr F- J- & Markman, H. J. (1984).Communioanon_$|sill§_1os_t_l§§1_'l
W. Brockton, MA.

Fitzgerald, H. E., Sullivan, L. A., Ham, H. P., Zucker, R. A., Bruckel, 8.,
Schneider, A. M., & Noll, R. B. (1993). Predictors of behavior problems in
three-year-old sons of alcoholics: Early evidence for the onset of risk. thjd

Qexeloomont. 6.4., 110-123.

Floyd, F. J. & Markman, H. J. (1984). An economical observation measure of

coupleS' communication skills .LoumaLoLConsultinojndflimcaLEsoohologx 5.211)
97- 103.

Floyd, F. J, O'Farrell, T. J. 8: Goldberg, M. (1987). Comparison of marital
observation measures: The Marital Interaction Coding System and the

Communication Skills Test JoumaLoJSoosultmundflImoaLEsxohologx 55(3)
423-429.

Floyd, F. J. & Zmich, D. E. (1991). Marriage and the parenting partnership:
Perceptions and interactions of parents with mentally retarded and typically

developing children. W, 52, 1434-1448.

Frankenstein, W, Hay, W. M., Nathan, P. E. (1985). Effects of intoxication
on alcoholics' marital communication and problem solving. MW
Aloohol 46(1). 1 -6

 

89

Frankenstein, W., Nathan, P. E., Hay, W. M., Sullivan, R. F., 8: Cocco, K.
(1985). Asymmetry of influence in alcoholics' marital communication: Alcohol's

effects on interaction dominance. WW, 11(4),
399-410.

Gottman. J. M. (1979). W New

York: Academic Press.

Gray-Little, B. & Burks, N. (1983). Power and satisfaction in marriage. A

review and critique. WM, 93(3), 51 3- 538.

Gustafson, R. (1988). The relationship between alcohol intoxication and

power in real-life non-alcoholic couples. W, 22, 55-62.

Ham, H. P, Zucker, R. A., 8: Fitzgerald, H. E. (1993, June). Assessmg

.I0 . 0‘I. '0 I I I'.I 0 . "I- 0 I‘ ' a‘ .0. ._I0 ._0
slum. Paper presented at the annual meeting of the American Psychological
Society, Chicago, IL.

Hesselbrock, M. N., Hesselbrock, V. M., Babor, T. F., Stabenau, J. R.,
Meyer, R. E., 8: Weidenman, M. (1984). Antisocial behavior, psychopathology and
problem drinking in the natural history of alcoholism. In D. W. Goodwin, K. T.

VanDusen. & S. A. Mednick (Eds.),_LonojIIIoinal_BssoaLoh_ln_Aloobolism (pp-
197-214). Boston: Kluwer-Nijhoff.

lchiyama, M. A., Zucker, R. A, &Fitzgerald, H. E., &Bingham, C. R. (in
press). Articulating subtype differences in self and relational experience among

alcoholic men via Structural Analysis of Social Behavior. MW
Cflnioalfisxcholoox.

chhiyama, M. A. Zucker, R. A, Fitzgerald, H. E., & Refior, S. K. (1996,
June) :0 0I o I;0; 0I: 0I 0 I. 0I0 o 0; ‘: 0I I0 Paper
presented at the Joint Scientific Meeting of RSA and ISBRA, Washington, D. C.

Jacob, T. (1975). Family interaction in disturbed and normal families: A
methodological and substantive review. Win, 82(1), 33-65.

Jacob, T. (1986). Alcoholism. Afamily interaction perspective. Nebraska
W. 3.4.. 159 206.

Jacob, T, Favorini, A., Meisel, S. 5., &Anderson, C. M. (1978). The

alcoholic' s spouse, children, and family interactions. JoumaLotfimdjmnAlgpbgj,
19(7), 1231- 1251.

Jacob, T. 8: Krahn, G. L. (1988). Marital interactions of alcoholic couples:

Comparison with depressed and nondistressed couples. WW
CliniooLEsxoholoox. 56(1) 73- 79.

 

90

Jacob, T. 8: Leonard, K. E. (1988). Alcoholic-spouse interaction as a function
of alcoholism subtype and alcohol consumption interaction. WM

Emholoox. 9112). 231-237.

Jacob, T. 8: Leonard, K. E. (1992). Sequential analysis of marital interactions
involving alcoholic, depressed, and nondistressed men. WM

Emholoox. 1521(4). 647-656.

Jacob, T, Ritchey, D., Cvitkovic, J. F, 8: Blane, H. T. (1981).
Communication styles of alcoholic and nonalcoholic families when drinking and not

drinking JoumaLoLStuoiosonfloohol. 42(5). 466- 482.

Jacob, T. 8: Seilhammer, R. A. (1987). Alcoholism and family interaction. In

T. Jacob (Ed.). EamIbleoLaotIonjnoosxohootholooxLIheorioLmomodojno
flndjngilpp. 535- 627). New York: Plenum.

Kolb, T. M. & Straus, M. A. ( 1974) Marital power and marital happiness in
relation to problem- solving ability. W, .35, 756- 765.

Leonard, K. E. (1993). Drinking patterns and intoxication in marital violence:
Review, critique, and future directions for research. In U. S. Department of Health

and Human Services BosomhMonooraohJichoboljndJmoLoersooaliiolonoot

WWW Publication No. 93- 3496. pp
253- 280). Rockville, MD. National Institutes of Health.

Leonard, K. E. 8: Blane, H. T. (1992). Alcohol and marital aggression in a
national sample of young men. JoumaLoleorooLsonaLMioloncoJll I. 19-30.

Leonard, K. E. &Senchak, M. (1993). Alcohol and premarital aggression
among newlywed couples .LoumalgLSjudjmnAloohol. suooltll. 96- 108

Leonard, K. E. 8: Senchak, M. (1996). Prospective prediction of husband

marital aggression within newlywed couples. WW.
195(3). 001-0012.

Levinger, G. (1964). Task and social behavior in marriage. W, 21,
433-449.

Lewis, C. E., Smith, E., Kercher, C, 8: Spitznagel, E. (1995). Assessing
gender interactions in the prediction of mortality in alcoholic men and women: A

20-vearfollow-up studv WW 1915)
1162- 1172.

Li, J. T. 8: Caldwell, R. A. (1987). Magnitude and directional effects of

marital sex-role incongruence on marital adjustment. Wes, 8(1),
97-1 10.

 

 

91

Moos, R. H., Finney, J. W. & Gamble, W. (1982). The process of recovery
from alcoholism: ll. Comparing spouses of alcoholic patients and matched

community controls JoumaLoLStudjomAmohol. .4319). 888- 909

Murphy, C. M. 8: 0' Farrell, T. J. (1994). Factors associated with marital
aggression in male alcoholics. W, 3(3), 321 -335.

McClintock, E. (1983). Interaction. In H. H. Kelley, E. Berscheid, A.
Christensen, J. H. Harvey, T. L. Huston, G. Levinger, E. McClintock, L. A. Peplau, &

D. R. Peterson, MIMI-3.11125 (pp. 68-109). New York: W. H. Freeman and
Company.

McCrady, B. S. (1987). Interpersonal contexts, alcohol consumption and

alcoholism. Implications for treatment. WM”,
&(2-3), 127- 143.

0' Farrell, T. J. (1992). Families and alcohol problems: An overview of

treatment research. 199W, 9(3&4), 339- 359.

0' Farrell, T. J. & Birchler, G. R. (1987). Marital relationships of alcoholic,

conflicted, and nonconflicted couples. Wm, 13(3),
259- 274.

O' Farrell, T. J. 8: Noel, N. E. (1989). Families and alcoholism: Developments
in practice, research, and theory. W, 3(1), 69- 81.

Paolino, T. J. Jr. & McCrady, B. S. (1977).In9_9199n9hem_am_agez
W. New York: Grune 8: Stratton.

Reider, E. E., Zucker, R. A, Noll, R. B., Maguin, E. T., 8: Fitzgerald, H. E.

(1988, August). Wm

WM. Paper presented at the 96th Annual Meeting of the American
Psychological Association, Atlanta, GA.

Roberts, M. 0., Floyd, F. J., O'Farrell, T. J., 8: Cutter, H. S. (1985). Marital
interactions and the duration of alcoholic husbands' sobriety. Amengamnajpj

QwundAlooholeuse. _1_1.(38:4). 303-313.

Robins, L. N., Helzer, J. E., Croughan, J., & Ratcliff, K. S. (1981). National
Institute of Mental Health Diagnostic Interview Schedule. Wm

Emhialm. 3.8. 381-389.

Selzer, M. L., Vinokur, A., 8: van Rooijen, L. (1975) A self- administered

Short Michigan Alcoholism Screening Test (SMAST)..LQum_aLQf_SnIdLe§_Qn_A199|391.
35(1),1 17-126.

 

92

Sprenkle, D. H. 8: Olson, D. H. (1978). Circumplex model of marital systems:
An empirical study of clinic and nonclinic couples.,19_u;n91_9_t_M9m91_an_d_Eam11y

IheLapx._4121. 59- 74.

Steinglass, P. (1981). The alcohol family at home: Patterns of interaction in

dry, wet, and transitional stages of alcoholism. MW, 33,
578-584.

Watts, 8., Busch, R., 8: Wilson, P. (1994). Partners of problem drinkers:
moving into the 19903. W, 13, 401-407.

Weisfeld, G. E., Russell, R. J. H., Weisfeld, C. C., & Wells, P. A. (1992).

Correlates of satisfaction in British marriages. _E1n919gy_&_899i9919|9gy, 13(2),
125-145.

Zucker, R. A. (1987). The four alcoholisms: A developmental account of the

etiologic process. In P C Rivers (Ednl WWW
Wlpp. 27- 83). Lincoln: University of Nebraska Press.

Zucker, R. A. (1994). Pathways to alcohol problems and alcoholism: A
developmental account of the evidence for multiple alcoholisms and for contextual
contributions to risk. In R. A. Zucker, J. Howard, 8: G. Boyd (Eds), 1W1

otalsoholmolomsijolodnojhobioosxobosooiolmflxotflsk (pp 255- 289)

Rockville, MD: National Institute on Alcohol Abuse and Alcoholism.

Zucker, R. A. 8: Barron, F. H. (1973). Parental behaviors associated with
problem drinking and antisocial behavior among adolescent males. In M. E. Chafetz
(Ed.,) “0|! 0 I- -II 2.; 0I0“ 0I‘ I -0 I‘L- 0I-

InsmutoonfloohoLAoumndAlcobolIsmlpp. 276- 296). Rockville. MD: National

Institute on Alcohol Abuse and Alcoholism.

Zucker, R. A., Ellis, D. A., & Fitzgerald, H. E. (1994). Developmental
evidence for at least two alcoholisms: I. Biopsychosocial variation among pathways

into symptomatic difficulty. WM. 108.
1 34-146.

Zucker, R. A., Ellis, D. A., &Fitzgerald, H. E. (1993). W

' IO-OIOII I -=0 'lI-I 0- 0- 0
dIfleLennanQn. Unpublished manuscript, Michigan State University, Department of
Psychologv. E. Lansing.

Zucker, R. A, Ellis, D. A., Fitzgerald, H. E., Bingham, C. R., 8: Sanford, K. (in
press). Other evidence for at least two alcoholisms ll: Life course variation in
antisociality and heterogeneity of alcoholic outcome. D919199n19n1_an9

13519092611191.9911

 

93
Zucker, R. A. 8: Fitzgerald, H. E. (1991). W

0I0 0 0 .0 I3“. I0.I . 0I0 JI- 0 (Grant
No. 2R01 AA07065). Washington, D. 0.: National Institute on Alcohol Abuse and
Alcoholism.

Zucker, R. A. 8: Fitzgerald, H. E. (1992). The Antisocial Behavior Checklist.
Michigan State University Family Study, Department of Psychologv. E. Lansing.

Zucker, R. A. 8: Noll, R. B. (1980). The Antisocial Behavior Checklist.
Unpublished instrument, Michigan State University, Department of PsychologY. E.
Lansing.

Zucker, 8- A.. Noll, R. B., & Fitzgerald, H. E. (1986). W
childLeanaloobolics (Grant No. R01 AA07065). Washington, D.C.: National
Institute on Alcohol Abuse and Alcoholism.

 

"‘IIIIIIIIIIIIIIIII