How Do I Love Thee? Let Me Consider My Options Cognition, Verbal Strategies, and the Escalation of...

25
How Do I Love Thee? Let Me Consider My Options Co pi tion, Verbal Strategies, and the Escalation of Intimacy JAMESM. HONEYCUTT Louisiana State University JAMES G. CANTRILL Northern Michigan University PAMELA KELLY Xavier University-New Orleans DAVID LAMBKIN Louisiana State University Previous research has revealed that individuals have expectations for the deuelopment of romance based on personal expm'ence and cultural images. A series of research questions and hypotheses was generated and a sequence of hierarchical log-linear analysis models was developed to test fa) the effect of gender and knowledge of relational stage and relationship type on prediction of next-occurring actions in the development of a romance, fb) the use of compliance-gaining strategies intended to advance a relationship to the next stage, and fc) just$cation of secondary goals in which participants reported why they chose the strategies they did as opposed to other alternatives. The data revealed effects f i r relational stage progression on prediction of ensuing actions. Ingratiation, explanation, and direct requests werepopular strategies fordeveloping intimacy. Genderdi&mces in terms of secondaygoals were reported. Results are discussed in terms of relational memory structures for the develop- ment of intimacy. ndividuals vary in how mindful and strategic they are in devel- oping intimacy. Interpersonal needs for companionship and I intimacy result in a vast array of activities for facilitating and maintaining varying degrees of romantic involvement. As a result, rela- James M. Honeycutt is an associate professor in the Department of Speech Communication at Louisiana State University, lames G. Cantrill is an associate professor in the Department of Communication and Performance Studies at Northern Michigan University. Pamela filly is an instructor in the Department of Communication at Xavier University. David Lnmbkin is a doctoral candidate in the Department of Speech Communication at Louisiana State Univer- sity. Please address correspondence to James M. Honeycutt, Department of SpeechCommu- nication, Louisiana State University, Baton Rouge, LA 70803-3923; e-mail: [email protected]. Human Communication Research, Vol. 25 No. 1, September 1998 39-63 0 1998 International CommunicationAssociation 39

Transcript of How Do I Love Thee? Let Me Consider My Options Cognition, Verbal Strategies, and the Escalation of...

How Do I Love Thee? Let Me Consider My Options Co pi tion, Verbal Strategies, and the Escalation of Intimacy

JAMES M. HONEYCUTT Louisiana State University JAMES G. CANTRILL Northern Michigan University PAMELA KELLY Xavier University-New Orleans DAVID LAMBKIN Louisiana State University

Previous research has revealed that individuals have expectations for the deuelopment of romance based on personal expm'ence and cultural images. A series of research questions and hypotheses was generated and a sequence of hierarchical log-linear analysis models was developed to test fa ) the effect of gender and knowledge of relational stage and relationship type on prediction of next-occurring actions in the development of a romance, fb) the use of compliance-gaining strategies intended to advance a relationship to the next stage, and fc) just$cation of secondary goals in which participants reported why they chose the strategies they did as opposed to other alternatives. The data revealed effects f i r relational stage progression on prediction of ensuing actions. Ingratiation, explanation, and direct requests werepopular strategies fordeveloping intimacy. Genderdi&mces in terms of secondaygoals were reported. Results are discussed in terms of relational memory structures for the develop- ment of intimacy.

ndividuals vary in how mindful and strategic they are in devel- oping intimacy. Interpersonal needs for companionship and I intimacy result in a vast array of activities for facilitating and

maintaining varying degrees of romantic involvement. As a result, rela-

James M. Honeycutt is an associate professor in the Department of Speech Communication at Louisiana State University, lames G. Cantrill is an associate professor in the Department of Communication and Performance Studies at Northern Michigan University. Pamela filly is an instructor in the Department of Communication at Xavier University. David Lnmbkin is a doctoral candidate in the Department of Speech Communication at Louisiana State Univer- sity. Please address correspondence to James M. Honeycutt, Department of Speech Commu- nication, Louisiana State University, Baton Rouge, LA 70803-3923; e-mail: sphone@unixl .sncc.lsu.edu.

Human Communication Research, Vol. 25 No. 1, September 1998 39-63 0 1998 International Communication Association

39

40 HUMAN COMMUNICATION RESEARCH / September 1998

tionships develop in intimacy through a variety of means reflecting dia- lectical ebbs and flows as individuals oscillate between desiring privacy and sharing innermost needs (Baxter & Simon, 1993). Furthermore, trajec- tories for the development of intimacy in interpersonal relationships vary in complexity and specificity. There may be individual differences in terms of the ability to forecast what happens next in the development of a relationship after being taken to a certain point.

In terms of individual differences about relational trajectories, females and more experienced individuals seem to possess vivid memories of actions that characterize the development of personal relationships (Hon- eycutt, Cantrill, & Greene, 1989). These memories are rooted in cognitive structures for what occurs over the life course of different types of rela- tionships. Relationship memory structures are hierarchically ordered on the basis of recall of particular events (e.g., meeting an individual for the first time at a specific place) and scripts for behavior embedded within various scenes. Even though relationships are in constant motion, relation- ship memory structures provide a perceptual anchor in which individuals can determine where they are in a relationship (Honeycutt, 1993). One of the functions of relational memory structures is that they allow individu- als to categorize behaviors in relationships, ascribe meaning to their actions, and guide adaptive responses. Following the navigation meta- phor, relational memory structures act as a gyroscope to guide individuals through relational space by providing labels for ongoing experiences and indicating likely courses for subsequent relational moves.

The purpose of this study is to integrate three bodies of literature: expectations for the development of romantic relationships, strategies for the escalation of intimacy, and how mindful individuals are in developing intimacy. In the process of reviewing existing research, we present a series of hypotheses designed to assess the extent to which strategic choices in verbalizing the desire to escalate intimacy depend on memory structures for what typically occurs in a developing romance.

Memory Structures: A Cognitive Explanation of Relational Development

Honeycutt and his associates (Honeycutt, 1993; Honeycutt, Cantrill, & Allen, 1992; Honeycutt et al., 1989) have conducted a program of research on cognitive explanations for relational development answering Duck's (1980) challenge to relational researchers: to analyze the time that indi- viduals spend alone replaying and anticipating future relational encoun- ters to understand the growth of relationships. These studies have re- vealed that men and women have cognitive expectations for the development of intimacy, with women having slightly more differentiated

Honeycutt et al. / ESCALATION OF INTIMACY 41

or complex expectations, For example, Honeycutt et al. (1989) found that men and women agree on various activities (e.g., self-disclosure, dating, meeting parents, etc.) that characterize an escalating romance. Individuals with more dating experience had more expectations of what should happen in a developing relationship. These individuals also processed a prototypical sequence of the behaviors more quickly than did less experi- enced individuals. In addition, women generated more individual expec- tations for relational development as well as reported more expectations for the development of an escalating romance.

The memory structure approach suggests that phases exist in memory in the representation of expectations for prototypical behaviors and that the expectations represent ”internal guides for behavior that help indi- viduals recognize behaviors and promote labeling of their experiences’’ (Honeycutt, 1993, p. 80). Honeycutt (199%) found that beliefs about relationships developing in routine patterns were associated with both the necessity of self-disclosure and with not discussing future plans too early in the relationship-that is, avoiding making a commitment too quickly The participants were looking for commitment signs but noted that verbal commitment is not necessarily communicated. Verbal commitment was atypical for believing that relationships develop in predictable ways.

We discovered that individuals who read two stories about couples performing a sequence of behaviors ending in marriage typically divided the stories into six stages (Honeycutt et al., 1989). Readers divided the stories into different scenes by placing a mark at the end of each sentence that segmented scenes. The scenes were similar to the social penetration stages discussed by Knapp and Vangelisti (1992). The first scene, initiation, involved the story characters meeting. The second scene consisted of formal dating, showing physical affection, and sharing informal time together and was labeled intensihing. Self-disclosure emerged as a separate stage as did sexual intercourse that was accompanied by displays of physi- cal affection. The fifth stage, integrating, consisted of meeting parents, giving gifts or mementos, using other-oriented statements signifying concern for the partner’s values, and making a verbal commitment. These actions reflected a combination of Knapp and Vangelisti’s intenslfying and integrating stages. The final stage, bonding, was getting married.

Developmental expectations are derived from a variety of experiences. Some individuals recall singular experiences, whereas others recall a series of encounters (Berger, 1995). This recall is often done through retroactive imagined interactions (IIs), in which individuals replay prior conversations in their minds. The recall of dialogue helps individuals understand what occurred, keeps relationships alive in the mind, and promotes catharsis. This type of recall also serves planning and rehearsal functions (Honeycutt & Patterson, 1997).

42 HUMAN COMMUNICATION RESEARCH / September 1998

Individuals often rehearse what they will say to relational partners before giving explanations, requesting something, disclosing bad news, or giving any type of salient message that requires careful delivery. Prior studies have revealed that such proactive 11s occur most with opposite-sex intimates and prospective or former opposite-sex intimates (Honeycutt, 1995a). 11s emphasize the mental creation of relationships as individuals envision interaction scenes, messages, and what may happen next. An important point within the memory structure approach is that 11s and developmental models serve as heuristic guides for processing relation- ship behaviors.

A critical presumption behind relational memory structures is that information processing affects behavior. Baldwin (1992) notes that “rela- tional schemas should shape the individual’s expectations about and interpretations of other people’s behavior, as well as beliefs about appro- priate responses. The individual bases his or her behavior on this infor- mation, to reach valued goals” (p. 478). Relational memory structures determine what may be said in a given scene based on previous experi- ences, instruction from others, or observation of others.

The foregoing studies, especially those advanced by Honeycutt and associates, establish the existence of relational memory structures and their potential influence on the production of discourse. However, pre- vious research (e.g., Baxter & Wilmot, 1984) has been confined to a rela- tively small number of transition points in a relationship and may not account for the complexity of the cognitive maps we use to chart the entire course of a relationship. In addition, it is worthwhile to examine the specific actions that individuals think will occur next in the evolution of intimacy as well as indicating what they would say to escalate the rela- tionship from a given point in its development. This research aims to replicate previous findings as it lays the foundation for the present study.

RQl: What subsequent actions do individuals expect to occur in the develop- ment of a romance when given information on the existing stage of devel- opment in terms of the occurrence of prototypical antecedent behaviors?

Verbal Strategies Designed to Operationalize Expectations

Given that individuals are led up to a certain point in a relationship and expect subsequent behaviors, the question arises as to how these expecta- tions may be facilitated. In other words, what messages would individuals use to communicate their desires? A possible answer may be found in the body of accumulated compliance-gaining research.

The communication and persuasion literature is replete with studies investigating typologies of interpersonal influence, in which individuals

Honeycutt et al. / ESCALATION OF INTIMACY 43

report or choose from a list of strategies how they would get another individual to comply with a request (for reviews, see Dillard, 1990; Kel- lermann & Cole, 1994; Seibold, Cantrill, & Meyers, 1985,1994). As Seibold et al. (1985) observed, one of the most ambitious typologies was Schenk- Hamlin, Wiseman, and Georgacarakos’ (1982) inductively derived set of compliance-gaining strategies and tactics. Precisely because it arose out of participants’ own interpretations of the context-relevant features in influ- ence situations, this typology seems particularly well suited for our study in which respondents were asked to generate and justlfy a range of messages for the escalation of intimacy. Furthermore, a situation explicitly referencing an intimacy dimension akin to our current investigation was included in the 1982 protocol. We used the Schenk-Hamlin et al. (1982) typology to extend the interpersonal influence literature into the realm of memory structure research and to focus on the concern that individuals have for the needs of their prospective partners.

Conventional wisdom suggests that to manipulate the face wants of self and other to become a more desirable partner for the escalation of intimacy, individuals will attempt to ingratiate themselves (Stires &Jones, 1969). Schenk-Hamlin et al. (1982) found that participants valued this strategy, defined as presenting an attractive front to the target of a request. Ingratiation ranges from subtle verbal or nonverbal actions, such as praise or flattery, to more overt actions, such as the giving of gifts, supportive listening, and showing love or affection. In the context of a developing romance, pregiving is a more accurate word for these behaviors. Ingratia- tion implies a false motive to seek compliance, whereas developing rela- tionships evoke a sincere liking for the other that it is hoped the other will reciprocate. Cate, Lloyd, and Long (1988) argue that equity is more impor- tant than ingratiation because norms of fairness are more important in the early phases of a relationship when compared to later phases.

We expected that an individual’s escalating memory structure would be replete with expectations of demonstrations of love and affection over periods of time. In fact, many of the examples of ingratiation provided by Schenk-Hamlin et al. (1982) are found in the escalating metamemory structure reported by Honeycutt et al. (1989). They found expectations reflecting types of ingratiation, including the exchanging of gifts; the use of other-oriented statements in which there is an expressed interest in the partner’s goals and values; verbal expressions of love; and the display of physical affection through kissing, hugging, and touching. Therefore, we posited the following hypothesis.

H1: Individuals will report ingratiation as a popular strategy of interpersonal influence to achieve their expectancy of what will happen next in the development of a romance.

4.4 HUMAN COMMUNICATION RESEARCH / September 1998

Secondary Goals in Choosing Strategies for Intimacy Development

Dillard, Segrin, and Harden (1989) discuss how multiple goals are operating when individuals seek compliance. There is the short-term goal of achieving compliance. Yet, there may be long-term consequences that affect the relationship between the seeker and target of the attempt. In the context of intimate relationships, the short-term goal reflects a primary intent of bringing behavioral change in the partner, whereas secondary goals reflect more general motivations, such as concern for another’s feelings and concern for the stability of the relationship. Specifically, Dillard et al. identified a list of four secondary goals derived from a review of literature on impression management that reflect identity, interaction, resource, and arousal management goals.

Identity goals facilitate an individual’s self-concept. Codes of conduct, ethics, and values symbolize identity; appropriate behavior is empha- sized. Dillard et al. (1989) found that a plurality of individuals (34%) reported identity concerns as justification for not using various compli- ance-gaining strategies in the Schenk-Hamlin et al. (1982) typology.

Interaction goals represent the desire to convey a favorable impression. Individuals want to avoid threatening the other’s self-concept, ensure smooth communication, and produce coherent statements. Conversely, resource goals are the desire to increase or maintain assets in three areas: relational, material, and physical (Dillard et al., 1989). Relational assets encompass personal rewards that arise from participation in a relationship with the target. Material assets include desire for physical objects and money. Physical assets refer to health concerns that might be jeopardized in pursuing the primary goal.

The final type of secondary goal is arousal management. Dillard et al. (1989) rely on discrepancy-arousal theory to explain desires for smooth interaction. Discrepancy-arousal theory assumes that individuals evalu- ate arousal as pleasant or unpleasant depending on the attractiveness and status of an interaction partner. Individuals desire to maintain a certain level of arousal, depending on their evaluation of the interaction partner (Burgoon & LePoire, 1993). In terms of compliance-gaining strategies, individuals may avoid strategies that they believe will arouse anxiety, uncertainty, or negative emotions.

Identity goals were the overwhelming choice of justifications. These goals reflect internal values and codes of conduct as well as the desire to enhance intimacy. There is the hope that the partner will be further attracted to the self. Individuals believe that they should project a persona that is seen as reflecting underlying values.

Honeycutt et al. / ESCALATION OF INTIMACY 45

Murstein (1987) discusses the importance of communicating values in relational development. He argues that in forming stable relationships, individuals initially evaluate a partner’s physical attributes and social standing, followed by the sharing of values (e.g., politics, religion) and an evaluation of how the partners function in terms of role compatibility (e.g., friends, lovers).’

Identity goals are also related to self-concept (Dillard et al., 1989). Bell and Daly (1984) found that individuals wanting others to like them used various communication strategies, such as being polite in conversations, confirming another’s self-concept, eliciting disclosure, and listening. Self- concept confirmation reflects a dimension of caring for the partner (Bell & Daly 1984). These findings led us to test the following hypothesis about identity goals in terms of escalating intimacy.

H2: Individuals will cite more identity goals than other types of secondary goals as a justification for not using alternative strategies in order to enhance intimacy at various stages of relational development.

Gender and Relational Awareness

There is evidence of gender differences in terms of superficial thought about what is occurring in personal relationships. Burnett (1990) reviews some studies on relational awareness and claims that men are not very aware about what is going on in their relationships. She surveyed 40 men and 40 women from varying social backgrounds and ages about the types of relationships they recently thought about. She found that reflection on relationships was more likely to occur when stimulated by particular events such as illnesses, periods of transition, and special occasions such as weddings.

Men were more reticent in communicating and thinking about relation- ships. Women were more concerned with assessment of both persons and relationships than were men. Men were more concerned with the business of forming and maintaining relationships in the sense of meeting and making contact. Women claimed to think more as well as know and talk more about their own closest relationship, although both genders claimed to care equally about their closest relationship. The most striking gender differences Bumett (1990) reported were males having more difficulty explaining relationships and preferring not to analyze personal relation- ships compared to the females. Males also reported having more difficulty in answering questions about their recall of a personal relationship that they had recently thought about.

We also found that women were able to sort a set of random actions reflecting the development and demise of romance faster than could men

46 HUMAN COMMUNICATION RESEARCH / September 1998

(Honeycutt et al. 1989, 1992). Relational mindfulness is reflected in the speed of sorting tasks, because ordering categories into a logical sequence requires the recognition of alternative categories. The studies that Bumett (1990) and Honeycutt (1993) review seem to demonstrate the tendency for women to monitor their relationships more than do men. Men appear to have more difficulty thinking about the daily maintenance of relationships and what makes them workcompared to women. Therefore, the following hypothesis on gender and awareness of secondary goals for enhancing intimacy in a developing relationship was tested.

H3: Women will be more aware of their secondary goals or justifications for why they did not choose alternative strategies in order to enhance intimacy at various stages of relational development when compared to men.

Awareness and Individual Differences in Typical Versus Personal Relationships

When considering the relational trajectories of potential romance, there may be differences in processing information as a consequence of the observer’s role. A pioneer study on perceived dimensions of interpersonal relationships by Wish, Deutsch, and Kaplan (1976) revealed differences in dimensions used to characterize relationships between reports of one’s own relationship (e.g., between you and casual acquaintances, you and your close friends, you and your spouse) compared to reports of others’ typical relationships (e.g., between casual acquaintances, between close friends, between husband and wife). The data revealed that individuals distinguished their own relationships from typical or other relationships. Cooperation was more important for evaluating typical relations than for one‘s own relationships. In evaluating their own relationships, the inten- sity of the relationship was more important. In this regard, individuals mentioned fewer hostile relations (e.g., your relationship with a lover compared to your relationship with a bitter enemy). On the other hand, hostile relations (e.g., business rivals, political opponents, guard/prisoner, supervisor/employee, interviewer/applicant) were perceived as charac- terizing others’ relationships.

Part of these findings can be explained by the level of information available to an observer. Attribution theory emphasizes the observer’s role in giving accounts for behavior in terms of actor-participant versus ob- server (Ross, 1977). Individuals attribute the cause of others’ negative behavior to internal dispositions, whereas situational attributions are used to explain negative aspects of their own behavior. Ross explains the attributional error in terms of the attributor focusing his or her attention

Honeycutt et al. / ESCALATION OF INTIMACY 47

on the other person and not on the environmental situation that the observed person is in. We explain our behavior in terms of the knowledge structures we have, such as previous experiences. Thus, individuals attri- bute their own behavior to external factors. This is because we know ourselves best; having access to prior experiences, we may explain our behavior in terms of the situation. In addition, our attention is concen- trated on our surroundings rather than on ourselves. The object of atten- tion tends to be perceived as the causal agent.

Therefore, individuals should be more aware of secondary goals in their justification of why they chose the intimacy enhancement strategies they did. The individual has more information available about the self, the current situation, and how this situation matches with prior experiences. Yet in evaluating typical relationships, individuals may rely on stereo- types that are derived from the escalating relational memory structures. Whatever the case, the individual would have less information available because of the inability to get into the mind or ”black box” of another person. The only information available is from stereotypes and one’s own experiences. In addition, individuals often think that their own relation- ships are uniquely different from other people’s typical romances, even though programmatic research has revealed that individuals agree on the behaviors that characterize developing romances. On the basis of these findings, the following hypothesis was tested concerning the relationship between awareness and articulating secondary goals in strategies de- signed to facilitate intimacy.

H4: Individuals will be more aware in articulating their justification of second- ary goals for developing intimacy for personal relationships as compared to typical relationships.

METHOD

Participants

A total of 538 participants were recruited from introductory communi- cation classes from two large universities located in the upper midwestem and southern part of the United States. Participants had not been exposed to any developmental model of communication at the time of the survey. After completing the survey, a brief question and answer session with participants revealed no prior class exposure to developmental models. The participants ranged in age from 18 to 52, with 21.81 being the average age. The sample consisted of 43% males and 57% females. Most of the

48 HUMAN COMMUNICATION RESEARCH / September 1998

sample consisted of single, never married adults (87%), with a smaller representation of married individuals (11%) and separated/divorced in- dividuals (3%).

Procedure

Participants were given a survey packet describing a variety of tasks, including responses to environmental concerns and predictions of what would occur in the development of two relationships. They were asked to read two stories; one involved themselves in a hypothetical relationship that was taken up to a certain point of intimate development, and the other story involved a typical relationship about two characters named Linda and Tom who met and reached various stages of intimate bonding. Wish et al. (1976) found that individuals distinguished their own relationships from other relationships in terms of intensity and cooperation.

The order of typical and personal stories was counterbalanced across the surveys. Each story (see the appendix) ended at a particular point in time, and the respondents were queried as to the next major thing they thought would happen in the relationship. Participants were asked what they or the story characters would say if such development was desired. A third open-ended question asked the participants why they thought they or the individual would say this instead of something else. This question was posed to reflect secondary goals.

The points in time that ended each story were derived from Study 4 of Honeycutt et al. (1989), based on their story that contained escalating, prototypical relational memory structure actions. The completed story consisted of actions that had been previously identified as being common in the development of romance.

Each of the stories in the current study had an average of 89 readers (range 82-96).’ These cell sizes provided good statistical power. For exam- ple, assuming a medium effect size, the power to detect stage differences was .88 using a conservative, two-tailed test with alpha specified at the .01 level (Cohen, 1969). There were no differential effects based on the order of the stories. After reading the first story, participants filled out a social and environmental issues attitudes questionnaire that took about 5 minutes. This questionnaire contained a series of statements using Likert scales with anchors of strongly disagree and strongly agree. A sample state- ment was ”Regulation of business by government usually does more harm than good.” This task acted as a distractor between Stories 1 and 2. After reading the second story, participants were thanked and briefly instructed on the idea of communication dimensions that characterize individuals at varying stages of romance.

Honeycutt et al. / ESCALATION OF INTIMACY 49

Coding of Memory Structure Actions

The first open-ended question was designed to have participants report their relational memory structure expectations about what actions would be forthcoming in the development of the relationship after being led to the given point in time. These responses were coded by two research assistants using the escalating memory structure actions described by Honeycutt et al. (1989). They coded 17 behaviors as well as a miscellaneous category.

Instead of using a miscellaneous category, we were able to code three new categories. One of the new categories was breaking off or ending the relationship despite its intimacy trajectory. There is cynicism about the future development of relationships. Following is a sample excerpt of this category after the reader was taken up to the final bonding stage in which the characters were exclusively committed to each other.

I think that now they have slept together, problems will arise. They will start finding out things about each other (family, personality traits, drugs, past relationships, different goals) that will make them fight. Honesty from the beginning and communication throughout the relationship is important.

Another category was an emotional inference of the relationship growing closer without any reference to a behavioral action. An example of this follows in which the participant was taken up to the bonding stage: "It will continue to grow because they are very honest and open. The rela- tionship started slowly and gradually built up." A final category that was coded seemed simple yet was theoretically important. We coded blanks or null responses. We were interested in the nonmentioning of actions as a type of inability to articulate actions and demonstrate awareness about relational trajectories. For example, one male respondent left the open- ended questions blank but at the end of the survey packet wrote "I don't think about this. It's too hard."

A subset of 20 randomly selected protocols was recoded as a reliability check. Category kappas were quite high, ranging from .67 to 1.00, with most categories reflecting perfect agreement. The category-by-category kappas were as follows: meeting (1.00); calling (1.00); small talk (.86); displaying physical affection (1.00); dating (1.00); engaging in joint activi- ties and sharing time together (1.00); self-disclosure (37); overcoming a crisis (.93); meeting parents or family (1.00); talking about future plans as a couple (1.00); verbal expression of love (1.00); exchanging gifts and/or tokens (.98); verbal commitment (1.00); living together (1.00); sexual inter- course (1.00); statements indicating an interest in the other's goals, orien- tations, and/or values (33); getting married (1.00); inference of the rela-

50 HUMAN COMMUNICATION RESEARCH / September 1998

tionship getting closer (.98); nothing or the inability to say what will happen next (.67); and breakup (1.00).

Coding of Compliance-Gaining Escalating Strategies

The second open-ended question asked participants what the story characters would say or do to make the next action happen. Coders used the Schenk-Hamlin et al. (1982) scheme (Wiseman & Schenk-Hamlin, 1981). Unlike other schemes that have been criticized, this typology con- tains direct request and explanation as categories as well as decision rules for differentiating strategies from another (Kellermann & Cole, 1994). Overall, kappa for the coding of these strategies was .90. Following are brief category labels, definitions, kappa reliabilities, and examples.

Ingratiation is the offering of gifts, love, affection, and doing favors or services that precede the request (1.00). An example statement is “I like the way things are going. I have a rose for you. I love you.” Promise is similar to ingratiation. However, an exchange is involved in terms of a quid-pro-quo arrangement. Both partners give to receive (1.00). An exam- ple is “I’d like to be friends first, then move on if we see it‘s fit to do so.’’ Direct request is asking the partner to accede to one’s request (33). An example is ”I think we should talk about getting married.’’ Explanation (.92) provides several reasons for responding to the request. The reason for growing closer is because the prior history of the relational develop- ment makes this a logical step. In addition, the appeal is based on trust- worthiness and integrity. An example is ”This is a nice setup so far. I really like you, but let‘s finish school.” Hinting as an indirect statement that is designed to have the partner respond in the desired way (e.g., “Maybe we could spend more time together”) was less reliably coded (.57). Hinting request was a new category in which an indirect question was used (.70).

Coding of Secondary Goals

As previously indicated, responses to the third open-ended questions were coded using Dillard et al.‘s (1982) secondary goals coding examples. Overall, kappa was high (.84), as were most category kappas: identity goals (.74), interaction goals (.86), personal goals (.74), and arousal goals (1.00). The lowest kappa was for relational goals (.67). However, instead of the expected (yet not mentioned) resource goals, a new goal emerged that reflected personal fulfillment (.83). Statements such as ”What’s in it for me? I would do this for myself. I have a need for closeness” are examples. Additional examples of the other secondary goals follow: ”I have my own standards. I know what I’d do, and this is the way I’d do it”

Honeycutt et al. / ESCALATION OF INTIMACY 51

(identity); “I’d be nice and not try to embarrass me or her” (interaction); “I’d give him a chance to make a decision about which way he wants to go with the relationship” (relational); “I would feel anxious about doing something else” (arousal).

Coding of Awareness of Secondary Goals

The degree to which participants were aware of secondary goal re- sponses was coded by two research assistants who read excerpts from Langer’s (1989) discussion of mindfulness and awareness of creating new categories. A lack of awareness was judged on the basis of a lack of explanation about goals and acknowledging the use of categories that already exist? Sample, prototypical statements reflecting lack of aware- ness are as follows: “I have always done it this way. It just works”; “This is the only way I’d do it”; ”I don’t think much about it.”

A 5-point Likert-type scale was used with high scores reflecting more awareness (1 = not very aware, 5 = extremely aware). There was high inter- coder reliability as indexed by an intraclass correlation coefficient of .78. The assistants were unaware of the sex of the participants when evaluating the amount of awareness.

Hierarchical Log-Linear Analyses

The open-ended questions were nominally coded. Therefore, the re- search question and the first and second hypotheses were tested using a series of three hierarchical log-linear models. These models were used to explore the relationship between gender, type of relationship story (typical/ personal), stage of relationship, each of the dependent variables, next-oc- curring escalating memory structure actions, compliance-gaining strate- gies, and secondary goals. Log-linear models are similar to multiple regression models, in that all variables that are used for classification are independent variables, and the dependent variable is the proportion of cases in crosstabulation cells (Bishop, Fienberg, & Holland, 1975). In hierarchical log-linear analysis, a saturated model is initially tested that contains all multiple-factor interactions and subsumes underlying models with factors deleted. As in a step-down procedure, the models are ar- ranged in descending complexity, with the multiple-factor (saturated) model at the top. A saturated model accounts for all conditional propor- tions. The question arises of whether less complex models without higher order interactions or factors account for the frequencies in the dependent variables.

A likelihood ratio chi square is reported for each model, and a likeli- hood ratio difference test is used to determine the best fitting model. By

52 HUMAN COMMUNICATION RESEARCH / September 1998

specifying the open-ended question response categories as the dependent variables, a two-way interaction between them and gender is analogous to a main effect for gender in ANOVA terminology. A three-way log-linear interaction reduces to a two-way interaction in ANOVA terms. Finally, H3 and H4 were tested using a two-factor ANOVA, with gender and relation- ship type as the independent variables and awareness ratings as the dependent variable.

RESULTS

Most participants mentioned at least one or two memory structure actions (M = 1.35, SD = .53) and more than one compliance-gaining strategy (M = 1.76, SD = .76). On the other hand, the number of secondary goals averaged roughly one (M = 1.11, SD = .49). An example of four multiple actions is in the following case of a 21-year-old female who read about her own relationship in which the story ended with sexual inter- course. The parenthetical statements represent the label for the coded escalating memory structure action.

I think our relationship would grow stronger [inference of relationship growing closer]. We would continue to see each other; we would go to church together, spend more free time together [sharing informal activities], and disclose more about ourselves [self-disclosure of intimate information]. We may even consider moving in together [cohabitation].

Individual Differences in Use of Compliance-Gaining Strategies

Separate results are presented for story type in Stories 1 and 2 due to the counterbalancing of typical and personal stories. In Story 1, a typical story was followed by a personal story, whereas in Story 2, the reverse order was presented (see the appendix for typical and personal stories). Thus, any differences in the results between Story 1 and Story 2 can be attributed to the order in which typical and personal stories were read and responded to. An ANOVA on the total number of compliance-gaining strategies revealed a significant gender effect in both types of stories, Story 1: F(1,492) = 5.52, p < .019, eta' = .01; Story 2 F(1,449) = 7.83, p < .007, eta' = .02. Regardless of whether they were reading about their own hypothetical relationship or the development of a typical romance, women reported more compliance-gaining strategies (Story 1: M = 1.97, Story 2: M = 1.88) than did men (Story 1: M = 1.77; Story 2 M = 1.64). There was also an effect for relational stage [Story 1: F(5,442) = 2.41, p < .036, eta' = .02; Story 2: F(5,449) = 2.89, p < .014, eta' = .04], in which individuals reading stories in early stages reported fewer strategies (Story 1: M = 1.64, Story 2 M =

Honeycutt et al. / ESCALATION OF INTIMACY 53

1.50) than did individuals reading stories in later stages (Story 1: M = 2.06, Story 2: M = 1.95).

RQ1: Prediction of Next-Occurring Memory Structure Action

In addressing the RQl and H1 and H2, we tested saturated log-linear models in which multiple factor interactions existed among type of rela- tionship, gender, and prediction of the next-occurring actions, choice of compliance strategies, and justification of secondary goal. More simple models accounted for the conditional proportions.

The log-linear analysis for testing the first research question revealed that the conditional proportions among the escalating memory structure actions were accounted for by relational stage in both types of stories, Story 1: likelihood ratio (LR) F(100) = 541.29, p < .001; Story 2 LR F(lO0) = 494.92, p c ,001.4 Table 1 presents the first-order conditional probabilities in both stories. Probabilities that have a plus sign are statistically sigrufi- cant based on the z tests for the effect parameters or lambdas (Goodman, 1984).5

As revealed in Table 1, individuals who read a story that ended after initial meeting (initiation stage) expected telephone calling and dating to occur in typical and personal romantic development. Individuals who read a story ending in the intensdying stage with the story characters dating expected more sharing of activities in Story 1 as well as expecting sexual intercourse as the next action in both stories. Furthermore, indi- viduals who read a story that ended with disclosure about personal values and failures (disclosure stage) also expected intercourse to ensue. Story 1 readers inferred that the relationship would get closer as the next action after reading that the story characters slept together.

Individuals taken up to the integrating stage in which the last actions by the story characters involved becoming more interested in each other's lives expected that verbal commitment messages would be forthcoming in both stories. Story 1 readers expected the story characters to move in or live together (cohabitation action), whereas Story 2 readers expected that the characters would get married. Finally, individuals taken to the bond- ing stage in which the story ended with a verbal commitment expected additional statements signdying commitment and getting married.

An additional log-linear analysis was conducted on choice of a second escalating action. There was an effect in both stories for relational stage, Story 1: LR F(lO0) = 195.50, p < .001; Story 2: LR p(lO0) = 156.43, p < .001. Similar to the results for the first-order actions, dating was mentioned as a second choice after meeting and calling. Additional self-disclosure was a second choice after reading about the story characters disclosing per- sonal information. The inference of the relationship getting closer was a

54 HUMAN COMMUNICATION RESEARCH / September 1998

TABLE 1 Significant First-Order Conditional Proportions of Next-Occurring Escalating Relationship Behavior as a Consequence of Existing Stage of Intimacy

Relational Stage

Initiation Intensifying Disclosure Intercourse Integrating Bonding Relational Memory STl/ST2a STlIST2 STl/ST2 STl/ST2 STl/ST2 STl/ST2

Structural expectancy Ask for other’s phone number and call later +.27/+.31 .W/.OO .00/.01 Dating +.56/+.59 .15/.07 .05/.04 Joint activities (informal time spent together) .02/.00 +.17/.10 .10/.08 Selfdisclosure of intimate information .03/.04 .04/+.14 .10/.09 Verbal commitment (talk about a long-term rela tionship) .OO/.OO .20/.17 .22/.13 Cohabitation .oo/.oo .oo/.oo .01/.01 Sexual intercourse .01/.00 +.15/+.25 +.21/+.28 Marriage .oo/.oo .oo/.oo .01/.02 Inference of relationship getting closer .Ol/.OO .08/.07 +.15/.13 Breakup .01/.01 .03/.01 .01/.01

.00/.01

.02/.05

.05/.09

.07/.10

.21/.25

.02/.02

.08/.07

.07/.05

+.17/.14 +.16/.06

.oo/.oo .oo/.oo

.05/.02 .05/.05

.06/.06 .05/.06

.01/.02 .01/.02

+.33/+.32 .37/+.34 +.15/.07 .12/.10 .01/.06 .OO/.OO .12/+.21 +.21/+.20

.04/.07 .07/.09

.05/.01 .01/.04

NOTE: ST1 = Story 1; ST2 = Story 2. a. Distinctions between Stories 1 and 2. +Significant at the .01 alpha level.

second choice among other participants after reading about disclosure among the story characters. Finally, verbal commitment was expected as a second choice in the integrating stage.

This analysis provides robust yet methodologically different support for previous findings regarding the shape of relational memory structures for escalating intimacy. It also suggests that when our respondents re- ported what they would say to pursue the instrumental goal of relation- ship development and why they chose a given strategy instead of alter- native strategies, the strategic moves were grounded in a tacit consensus as to what developmental stage was in the offing.

H1: Test of Popular Strategies Designed to Escalate Intimacy

The first hypothesis posited that individuals would report pregiving as the most popular strategy of interpersonal influence to achieve their expectancy of what would happen next in the development of a romance. This hypothesis was partially confirmed. The hierarchical log-linear

Honeycutt et al. / ESCALATION OF INTIMACY 55

analysis of the use of compliance-gaining strategies to escalate intimacy revealed no effects for gender, relational stage, or relationship type. How- ever, a simple effect emerged among the patterning of the strategies themselves. This was the case for three log-linear models based on choice of a first, second, and third strategy across both stories [first strategy, Story 1: LR p(16) = 1277.50, p < .001; first strategy, Story 2: LR p(16) = 1140.16, p < .001; second strategy, Story 1: LR p(16) = 711.92, p < .001; second strategy, Story 2 LR p(16) = 608.67, p c .001; third strategy, Story 1 LR p(16) = 335.88, p < .001; third strategy, Story 2 LR p(16) = 273.65, p < .001].

Table 2 presents the conditional proportions of the compliance-gaining strategies. Pregiving emerged beyond chance probability as a popular strategy. Further examination of the lambda parameter estimates revealed that there were three popular strategies: pregiving, direct request, and explanation.

Further examination of Table 2 reveals that hinting request emerged as a second choice strategy. In addition, some Story 2 readers could not think of a strategy to escalate intimacy (nothing category). There was an appar- ent inability to articulate an initial strategy.

H2: Test of Secondary Goal Justifications

The second hypothesis posited that individuals would report more identity goals than other types of secondary goals as a justification for not using alternative strategies to enhance intimacy at various stages of rela- tional development. This hypothesis was partially confirmed. Although these goals werr cited, additional justifications were also reported. Table 3 presents the conditional proportions reflecting the secondary goals.

The hierarchical log-linear analysis revealed an effect for type of rela- tionship on secondary goal justification in both stories, Story 1: LR P(4) = 57.46, p < .001; Story 2: LR p ( 4 ) = 45.80, p c .001. As revealed in Table 3, individuals cited identity goals when reading about their own romantic development. However, interaction goals were reported when reading about a typical relationship and less used in own, personal relationships. Personal goals were used when reading about own romantic involvement and less used when reading typical stories. Finally, there were no sigmfi- cant models for reporting a second, secondary goal.

H3 and H4: Awareness in Secondary Goal Justification

There was support for H3 that women would be more aware about their secondary goal justifications than men. A slight gender effect emerged [Story 1: F(1,492) = 6.92, p < .009, eta’ = .01; Story 2 F(1,449) = 13.38, p < .001, eta’ = .03], in which men were rated as being less aware in justifying their strategies (Story 1: M = 2.72, Story 2 M = 2.58) than were women

56 HUMAN COMMUNICATION RESEARCH / September 1998

TABLE 2 Significant Conditional Proportions of Strategies Used to Escalate Intimacy

First Strategy

Strategy STl/SRa

Ingratiation +.18/+.14 Direct request +.27/+.28 Explanation +.43/+.38 Nothing .03/+.10 Hinting request .06/.06

Second Strategy Third Strategy

S T l / S R STl/Sr;!

+.15/+.18 .ll/.ll +.38/+.33 +.41/+.41 +.25/+.23 +.30/+.26

.oo/.cil .oo/.oo +.14/+.16 .10/.12

a. Distinctions between Stories 1 and 2. +Significant at the .01 alpha level.

TABLE 3 Significant First-Order Conditional Proportions of Secondary Goals Given Type of Relationship

Type of Relationship

'Ispical Personal Seconda y Goal S T l / S R STI/ST2

Identity Interaction Relational Personal

-.17/-.17 +.38/+.3a +.60/+.51 -.30/-.32

.12/+.19 .08/-.07 -.09/-.10 +.21/+.19

NOTE: These are orthogonal contrasts between typical and personal relationships so that a + sign reflects a significant goal beyond the .01 alpha level, whereas a -sign signifies a goal occurring significantly less often.

(Story 1: M = 2.95, Story 2: M = 2.93). The fourth hypothesis posited that individuals would be more aware in articulating their justification of secondary goals for developing intimacy for personal relationships com- pared to typical relationships. This hypothesis was partially supported. A significant effect emerged for relational type in Story 1, F(1,492) = 7.33, p < .007, eta' = .01, in which typical story readers were rated as less aware about justdying secondary goals (Story 1: M = 2.74) than when reading about personal involvement in a developing romance (Story 1: M = 2.96). There were no significant differences between typical and personal in- volvement for Story 2 readers.

DISCUSSION

The log-hear models of the next-occurring escalating memory struc- ture action revealed that knowledge of relational stage predicted several

Honeycutt et al. / ESCALATION OF INTIMACY 57

actions. By counterbalancing the comparison of personal with typical involvement in two stories, we were able to replicate many of the findings in log-linear analyses. These findings add support to relational memory structure theory, which proposes that individuals have knowledge struc- tures that allow them to interpret behaviors in personal relationships and to interpret the meaning of the behaviors as a consequence of their experiences. Memory structures for the development of intimacy act as perceptual anchors to guide individuals through relational space. In this regard, developmental stages are useful, heuristic devices that help us navigate our way through the development of intimacy.

The relational memory structure approach places importance on the mental creations of relationships based on memory, imagined interactions, talk, and expectations that sustain individuals in everyday living. In addition, current relational experiences affect reconstruction of prior events in relationships. Duck (1986) also argues that relationships should be regarded as changing mental and behavioral creations of participants that play a crucial role in creating and sustaining relationships. The time spent alone analyzing future encounters reflects individuals’ use of rela- tional memory structures to understand and differentiate among different types of relationships, such as distinguishing a casual dating relationship from an exclusive romance.

An observation that should not be dismissed as trivial is that although actions such as verbal commitment are expected in more developed stages such as integrating or bonding, many relationships consist of partners with different views of commitment. Critics could argue that findings such as these are expected and intuitive. Yet, we live in a pluralistic society that is fragmented with diverse information, lifestyle choices, political inter- ests, and religious values. Furthermore, the differences between individu- als’ schemata for relationships are likely to be greater. As a result of such pluralism, each of us constructs our own reality from a diverse set of resources (Stephen, 1994). Because the information is quite diverse and open to contradictory interpretation, individuals’ sense of uniqueness comes from a wide range of selection and prioritization of informational sources.

Similarly expectations for the development of romance are both idi- osyncratic and consensual. Berger (1993) makes a similar argument when discussing goals, plans, and mutual understanding in relationships. He reviews studies that indicate that most relationship partners do not spend much time negotiating meanings or shared understanding. Individuals tend to use easily retrieved general knowledge to interpret behaviors of relational partners before they build knowledge structures exclusively focused on the particular individual in the relationship. General knowl- edge about relational development is more readily used than is creating

58 HUMAN COMMUNICATION RESEARCH / September 1998

new knowledge based on observation of the other individual (Berger, 1993).

Individuals vary in their expectations of how relationships should develop due to the variety of informational sources that form the founda- tion for the expectations. Differences exist among what people feel a person should do, what they feel he or she will do, and what the person actually does. When these differences are minimized, persons will feel more satisfied in their romantic relationships. Stephen (1994) notes that relational partners who share a worldview come to appreciate their unique bond of shared knowledge, create an interpretive framework, and at the same time reinterpret themselves within the framework.

An intriguing finding that may not immediately stand out but is important from a social cognition and information processing perspective is that individuals reported inferences of the relationship becoming closer despite instructions to report only on the next occurring behavior. On one hand, it could be argued that this is additional evidence of superficial thought given to relationship processing. Some individuals may not have a readily accessible category available that accurately describes the pro- gression of behaviors. If anything, the instructions to report behaviors and not inferences combined with the given examples of each would enhance awareness of relational behaviors. Honeycutt (1993) reports on the inabil- ity of individuals to overcome the intermixing of inferences with reports of objective behaviors. In addition, women tend to report more inferences and objective behaviors than do men (Honeycutt et al., 1992). This finding is consistent with the findings in support of H3, in which men were rated as less aware in citing secondary goal justifications.

The first hypothesis was partially supported, in which ingratiation was the most popular strategy to escalate intimacy. In addition, direct request and explanation were also used. Aprototypical example of these strategies is in the following example from a 26-year-old male who read a story about personal romantic involvement that ended in the bonding stage. He is delivering his marriage proposal that contains explanation, ingratiation, and direct request strategies. He wrote, "We've been together so long, I can't imagine life without you. I love you, [name].Will you marry me?"

One of the Limitations of this study was the coding of compliance-gaining strategies. Kellermann and Cole (1994) criticize compliance-gaining tax- onomies for being atheoretical collections of strategies. Although the Schenk-Hamlin et al. (1982) scheme has a set of properties that urufy and distinguish strategies, the taxonomy has problems in strategy overlap (e.g., promise and ingratiation). However, a prime focus of our study was the examination of secondary goals and awareness of goal articulation. Kellermann and Cole (1994) have argued for more research on the differ- ences in compliance gaining in different situations for different goals. By identifying the secondary goal justifications of our participants, we were

Honeycutt et al. / ESCALATION OF INTIhUCY 59

able to identlfy concerns for self-identity and personal goals in personal relationships. And in keeping with the tradition of a goal limitation model (Seibold et al., 1994), this study further supports the idea that the choice of compliance strategies is a function of how one’s current objectives mediate choices. If one wants to use a verbal strategy to escalate a rela- tionship, decisions are based on assumptions of stage-based effectiveness. Individuals clearly think of relationships developing in stages (Honeycutt, 1993). Our participants were able to identify a variety of subsequent actions, such as calling, sharing time together, and pessimism about breaking up, as possible ensuing actions.

In addition, we found partial support for the hypothesis that individu- als were somewhat unaware in processing secondary goals for typical relationships compared to cognition about personal relationships. This finding is partially consistent with conclusions by Kellermann and Cole (1994), who speculate that individuals are cognitive misers who often use rules of thumb and attend to limited information when making decisions across a variety of contexts. However, this process may be attenuated in reading about personal romantic development. Furthermore, our coding of the awareness of secondary goals does not preclude the possibility that some of our participants were inarticulate or insufficiently motivated to elaborate on the hypothetical scenarios.

Another limitation of this study concerns the limits of the claims that can be made based on a story-based and hypothetical self-report. Even if we have relational development schemata, it does not mean that we use them in our real lives or that relational schema predicts actual behavior. There is a small positive correlation between the number of intimate relationships an individual has been in and the time it takes to sort a set of random escalating behaviors into a temporal sequence that results in a romantic bond such as marriage (Honeycutt et al., 1989).

Through focusing on differences in messages for escalating intimacy, this research has resulted in a better understanding of relational memory structure expectations, compliance gaining, secondary goals, awareness, and how they affect each other. Future research needs to examine when relational memory structures are modified or accommodated to new observations that do not clearly reflect an existing category. Although individuals often use existing knowledge of relationship characteristics (e.g., intimates often hold hands in public to signal they are a couple) to interpret behaviors in particular instances rather than building a bank of new knowledge for each relationship in which one is involved, the ques- tion arises concerning the creation of new categories for relationships. Are categories created in times of relational crises, in dialectical moments when individuals have opposing needs in the relationship (Baxter & Simon, 1993), or in times of quiet reflection? Yet, the fact that our partici-

60 HUMAN COMMUNICATION RESEARCH / September 1998

pants were able to articulate ensuing relational memory structure actions and the use of the influence strategies they would use to carry out the actions reveal how interpersonal relationships consist of mental creations and are not simple outcomes of interconnected, communication patterns. The idea that communication is the relationship needs to be considered not only in the interpersonal and behavioral realm but also understood at an intrapersonal level within the minds and imagined interactions of relational partners.

APPENDIX

Story About Linda and Tom

Linda met Tom in a biology lab during the fall semester. They talked about the course, their majors, and what they wanted to do after graduation. Tom asked Linda if he could call her and see about going on a date to one of the home football games (1). They enjoyed the game and continued going to sporting events as well as having other dates (2). They held hands and kissed each other. They did informal activities together such as studying, biking, and jogging. They disclosed personal information about such things as family crises, failures, and fantasies (3). They slept together (4). Tom invited Linda to visit his hometown and meet his family. They periodically exchanged little gifts and favors such as jewelry and candy. They talked about each other‘s needs and how to help each other whenever they could (5). They told each other that they were exclusively committed to each other (6).

Story About Personal Involvement in a Developing Romance

You and another person met for the first time at a party. The two of you engaged in small talk such as commenting on how nice the party was and who you knew at the party, and learned a little about where each other was from. You asked your acquaintance for his/her phone number and if you could call him/her later to see about going on a date (1). The first date went well and you continued dating. You held hands, kissed, and hugged each other. The two of you also did informal activities such as watching TV, studying together, playing tennis, and going to lunch (2). You began to self-disclose to each other and reveal personal information such as likes and dislikes, successes and failures, values, and the like (3). You had sexual intercourse (4). Your partner took you on a weekend trip to his/her hometown to meet his/her family. You gave each other candy, “love cards,” and jewelry. You became more interested in each other‘s goals and lives because you felt like you needed each other and had an exclusive relationship (5). You made a verbal commitment so you could make the relationship long term and stable (6).

NOTE: The numbers in parentheses represent various time points that the readers were led up to before reporting what would happen next.

Honeycutt et al. / ESCALATION OF INTIMACY 61

NOTES

1. Murstein's (1987) model of dyadic pairing has been criticized in a number of reports (e.g., Leigh, Holman, & Burr, 1987; Stephen, 1987). The major point of criticism has centered on Murstein's (1987) contention that the sharing of values follows the evaluation of a partner's stimulus qualities and precedes evaluating role compatibility. Murstein has amended his model to allow value information to accrue continuously, with the rate of acceleration determining each stage.

2. There were a few isolated instances of missing data due to participants providing no response to the open-ended question on the next-occurring action in the development of the romance.

3. Langer (1989) discusses how a concern with outcomes rather than process induces mindlessness and lack of awareness. Responding to familiar situations, we notice only the minimal cues necessary to carry out the scenario. Langer discusses how mindfulness involves the creation of new categories, openness to alternative information or ideas, and awareness of more than one perspective. Mindlessness is the rigid reliance on old categories and reflects a lack of awareness of alternative categories; mindfulness involves the continual creation of new categories.

4. The hierarchical log-linear analysis tests that main effects and higher order effects are zero. Small significance levels indicate rejection of the hypothesis that the terms of particular orders is zero. None of these tests were significant beyond a second-order effect. Hence, only a first-order or main effect was significant.

5. The log-linear analysis yields estimates for parameters (lambdas) and standard emrs for each estimate. The ratio of the parameter estimate to the corresponding standard error results in a z core. Lambdas with z values greater than 1.96 in absolute value are significant at the .05 alpha level. When tests for many lambdas are conducted, the probability of a Type I statistical error inaeases. Special procedures for testing these lambdas are discussed by Goodman (1984) and were used in the models.

REFERENCES

Baldwin, M. W. (1992). Relational schemas and the processingof social information. Psycho- logical Bulletin, 112,461-484.

Baxter, L. A., & Simon, E. P. (1993). Relationship maintenance strategies and dialectical contradictions in personal relationships. journal of Social and Personal Relationships, 10, 225242.

Baxter, L. A., & Wilmot, W. (1984). "Secret tests": Strategies for acquiring information about the state of the relationship. Human Communication Research, 11,171-201.

Bell, R. A., & Daly, J. A. (1984). The affinity-seeking function of communication. Cornmunica- tion Monographs, 51,91-115.

Berger, C. R. (1993). Goals, plans, and mutual understanding in relationships. In S. Duck (Ed.), Individuals in rehztionships (pp. 30-59). Newbury Park, CA Sage.

Berger, C. R (1995). Aplan-based approach to strategic communication. In D. E. Hewes (Ed.), Cognitive buses of interpersonal Communication (pp. 141-179). Hillsdale, NJ: Lawrence Erlbaum.

Bishop, Y.M.M., Fienberg, S. E., & Holland, P. W. (1975). Discrete multivariate analysis: Theory and practice. Cambridge, MA: MIT.

Burgoon, J. K., & LePoire, B. A. (1993). Effects of communication expectancies, actual communication, and expectancy disconfirmation on evaluations of communicators and their communication behavior. Human Communication Research, 20,57-96.

62 HUMAN COMMUNICATION RESEARCH / September 1998

Burnett, R. (1990). Reflection in personal relationships. In R. Burnett, P. McGhee, & D. Clarke

Cate R M., Lloyd, S. A., & Long, E. (1988). The role of rewards and fairness in developing

Cohen, J. (1969). Statisticalpower analysisfor the behavioral sciences. New York Academic Press. Dillard, J. I? (1990). Seeking compliance: The production of interpersonal inf2uence messages.

Scottsdale, Az: Gorsuch Scarisbrick. Dillard, J. I?, Segrin, C., & Harden, J. M. (1989). Primary and secondary goals in the

production of interpersonal influence messages. Communication Monographs, S6,19-38. Duck, S. (1980). Personal relationships in the 1980s: Towards an understanding of complex

human socially. Western Journal of Speech Communication, 44,114-119. Duck, S. (1986). Human relationships: A n introduction to social psychology. London: Sage. Goodman, L. A. (1984). The analysis of mss-classifieddata havingordered categories. Cambridge,

MA: Harvard University Press. Honeycutt, J. M. (1993). Memory structures for the rise and fall of personal relationships. In

D. Duck (Ed.), lndividuals in relationships (pp. 60-86). Newbury Park, C A Sage. Honeycutt, J. M. (1995a). Imagined interactions, recurrent conflict and thought about per-

sonal relationships: Amemory structure approach. In J. E. Aitken & L. J. Shedletsky (Eds.), Intrapersonal communication processes (pp. 138-150). Plymouth, MI: Midnight Oil & Speech Communication Association.

Honeycutt, J. M. (199%). Predicting relational trajectory beliefs as a consequence of typicality and necessity rating of relationship behaviors. Communication Research Reports, 12,3-14.

Honeycutt, J. M., Cantrill, J. G., & Allen, T. (1992). Memory structures for relational decay: A cognitive test of sequencing of de-escalating actions and stages. Human Communication Research, 18,528562.

Honeycutt, J. M., Cantrill, J. G., & Greene, R. W. (1989). Memory structures for relational escalation: A cognitive test of the sequencing of relational actions and stages. Human Communication Research, 1662-90.

Honeycutt, J. M., &Patterson, J. (1997). Affinity strategies in relationships: The role of gender and imagined interactions in maintaining liking among college roommates. Personal Relationships, 4,3546.

Kellermann, K., & Cole, T. (1994). Classifying compliance gaining messages: Taxonomic disorder and strategic confusion. Communication Theory, 4,340.

Knapp, M. L., &Vangelisti, A. (1992). Znterpersonal communication and human relationships (2nd ed.). Boston: Allyn & Bacon.

Langer, E. J. (1989). Mindfulness. Reading, MA: Addison-Wesley. Leigh, G. K., Holman, T. B., & Burr, W. R. (1987). Some confusions and exclusions of the SVR

theory of dyadic pairing: A response to Murstein. Journal of Mnrriage and the Family, 49, 933-937.

Murstein, B. I. (1987). A clarification and extension of the SVR theory of dyadic pairing. ]ournu1 of Mam'age and the Family, 49,929-933.

Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 10, pp. 173- 220). New York Academic Press.

Schenk-Hamlin, W. J., Wiseman, R. L., & Georgacarakos, G. N. (1982). Amodelof properties of compliancegaining strategies. Communication Quarterly, 30,92-100.

Seibold, D. S., Cantrill, J. G., & Meyers, R A. (1985). Communication and interpersonal influence. In M. L. Knapp & G. R. Miller (Eds.), Handbwk of intetperwnal communication (pp. 551-611). Beverly Hills, CA. Sage.

Seibold, D. R., Canhill, J. G., & Meyers, R. A. (1994). Communication and interpersonal influence. In M. L. Knapp & G. R Miller (Eds.), Handbwk of interpersonal communication (2nd ed., pp. 542-588). Beverly Hills, CA: Sage.

(Eds.), Accounting for relationships (pp. 73-94). London: Methuen.

premarital relationships. Journal of Mam'age and the Family, SO, 443-452.

Honeycutt et al. / ESCALATION OF INTIMACY 63

Stephen, T. (1987). Taking communication seriously: A reply to Murstein. lournal ofMurriage and the Family, 49,937-938.

Stephen, T. (1994). Communication in the shifting context of intimacy: Marriage, meaning, and modernity. Communication Theory, 4,191-218.

Stires, L. K., &Jones, E. E. (1969). Modesty versus self-enhancement as alternative forms of ingratiation. ]ournal ofExper'menta1 Social Psychology, 5,79-82.

Wiseman, R. L., & Schenk-Hamlii, W. (1981). A multidimensional scaling validation of an inductively-derived set of compliance-gainiig strategies. Communication Monographs, 48,

Wish, M., Deutsch, M., & Kaplan, S. J. (1976). Perceived dimensions of interpersonal relations. 251-270.

Journal of Personality and Social Psychology, 33,404-420.