I Am Us

PAGE 1

I AM US: OVERLAPPING MENTAL REPRESENTATIONS OF SELF AND COMMUNITY BY SARAH HERNANDEZ A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Michelle Barton Sarasota, Florida March, 2012

PAGE 2

ii Acknowledgements Thank you to Dr. Michelle Barton, my thesis sponsor, for being an incredible source of support and knowledge throughout the entire process. Thank you to Dr. Steven Graham and Dr. Duff Cooper, who helped me immensely with my thesis method and results. Thank you to Dr. Heidi Harley for the much appreciated work on my committee. Thanks to everyone who participated in my study. Finally, thanks to my fellow psychology students for their input, and to my family and best friends for their tremendous encouragement.

PAGE 3

iii Ta ble of Contents ACKNOWLEDGMENTS ii TABLE OF CONTENTS iii LIST OF TABLES v LIST OF FIGURES v i ABSTRACT vi i INTRODUCTION 1 Self Expansion model 2 The self 2 Theory and measurement 4 The reaction time effect 7 From other to ingroup 11 Social identification 12 Bringing models to a measure 15 Reaction time revisited 17 Overview 20 From ingroup to community 21 Psychological sense of community 21 Identification a s a distinction between ingroup and community 22 Importance of context 25 Inclusion of community in the self 28 THE CURRENT STUDY 30 METHOD 33

PAGE 4

iv RESULTS 38 DISCUSSION 42 REFERENCES 50 TABLES 54 FIGURES 57 APPENDIX 62

PAGE 5

v List of Tables TABLE 1: Mean (Standard Deviation) Positive and Negative Trait Evaluation Proportions TABLE 2: Mean (Standard Deviation) Computer Task Responses (ms) TABLE 3: Mean (Standard Deviation) Matching and Mismatching Self Error Proportions

PAGE 6

vi List of Figures FIGURE 1: The Inclusion of Other in Self (IOS) scale FIGURE 2: The Inclusion of Ingroup in Self (IIS) scale FIGURE 3: The Inclusion of Community in Self (ICS) scale FIGURE 4: Mean Reaction Time Responses Across Community FI GURE 5: Interactive activation and competition (IAC) model of self, other, and traits

PAGE 7

vii I AM US: OVERLAPPING MENTAL REPRESENTATIONS OF SELF AND COMMUNITY Sarah Hernandez New College of Florida, 2012 ABSTRACT This study investigated the inclusion of community in the self comparing a school community and a local/city community Participants were ad ministered a survey with a self descriptive task, community descriptive task, and modified Inclusion of Community in Self Scale. Participants then performed a self descriptive reaction time task R eaction time responses on traits that matched between the self and community were quicker than on trai ts that mismatch ed Participants were more connected to the school community than the local/city community, and the reaction time effect was significant only in the s chool community. These findings provide more concrete evidence for the inclusion of commun ity in self scale and support the use of connectionist models in explaining the overlap of self and c ommunity mental representations ___________________________ Dr. Michelle Barton Division of Social Sciences

PAGE 8

1 I am Us: Overlapping Mental Representations of Self and Community When someone says the word community, it may evoke a variety of images and ideas of what the concept actually means. Neighborhood communities can be pleasantly quaint houses on a block, smiling children riding tricycles down the sidewalk, middle aged dads mowing the lawn on Saturday mornings, porch conversations at dusk the street people grew up on and a place they call home. Collectively, the neighborhoods represe nt a much larger city community that can be ch aracterized by the rumbling sound of trains, bustling streets busses and cars, and the occasional street parade. S mall businesses, schools, parks, fire stations, libraries and even the best local baker on the corner street can be the groups that constitu te a local city community. Even an individual's closest group of friends can be acknowledged as a strong close knit community. People connect with a multitude of communities, based on their own personal and social identifications; community is just as indi vidualized as the self. The i nclusion of community in the self has been measured, but the degree to which these two social entities cognitively overlap with one another has yet to be investigated (Mashek, Cannaday, & Tangney, 2007). In order to understand overlap ping mental representations of s elf and community, the s elf concept is discussed first Numerous models vacillate between impressions of what con stitutes the s elf, however they collectively converge on a single idea: the s elf is complex highly expansive and inclusive ( James, 1890; Greenwald & Pratkanis, 1986; Aron, Aron, & Smollan 1992 ) This advance in theory is important because it completely opens the do or for empirical research concerning the s elf and vari ous other social entities. The s el f is not a solitary construct, but incredibly contingent on others.

PAGE 9

2 The following review of literature will follow the journey of the complex self, from close other to ingroup and finally to community. The conception of the s elf as multi faceted and ex pansive is incorporated into the study o f close relationships. The Self Expa nsion m odel (Aron et al., 2005), which proposed that individuals are motivated to include their clos e others in their self concepts, is discussed first. This research is then appli ed to groups and more specifically as ingroup identification T he idea, that i ndividuals are motivated to include vari ous ingroups in their sense of selves as a means to develop an overall social identity is then reviewed. The distinction between ingroup and co mmunity will then be discussed. The concept of c ommunity will be reviewed and then exam ined within the context of the self. Overlap ping mental representations of s elf and community must be examined for research concerning t he multi faceted and inclu sive s elf to progress. Self Expansion Model The Self "#! William James' (1890) landmark book, The Principles of Psychology he offered one of the most influential discussions of the self In the chapter, The Consciousness of The Self James (1890) built a framework for one of the most important facets of social psychology the answer to the famously ubiquitous question, who am I? According to James, the s elf is not constructed internally, but through other people and possessions (James, 1890) There is a division between me' and I' the empirical self, as an ob ject (me) and the self as the pure e go, or subject (I) The empirical self may be divided into the material self, the social self, and the spiritual self. This conceptualization was the firs t i n the field to conceive of the s elf as socially complex.

PAGE 10

3 Almost a century later, Greenwald and Pratkanis (1984) sought to evaluate and integrate the past and contemporary ideas concerning the self ultimately defin ing the self as a person specific, central, complex, and attitudinal schema. The ir chapter initially focuses on evaluating the theoretical conceptions of the self as more than just a solitary construct Empirical links between the theory and evidence exist, s uch as research on self reference effects in memory, self serving attribu tio nal bias, self presentation, self schema, and self monitoring. However, an integrative conception of the self had yet to be created (Greenwald & Pratkanis, 1984). The holistic de finition that Greenwald and Pratkanis (1984) proposed posited the self as complex, an attitudinal schema, person specific, and central. The complex self consists of several facets: the diffuse self, public self, private self, and collective self. The diffu se self does not distinguish between self and others, and is most ly related to affective states. In contrast, the public self is extremely sensitive to evaluations of others and strongly seeks approval. The priv ate self is concerned with self definition an d evaluation in the absence of others T he collective self internalizes goals of the groups that the individual identifies with. T hese four facets of the complex self have an incredible amount of variance between individuals, so the self is considered to b e person specific; individual differences in self concept and cognitive biases also support the self as personalized. The self as a central construct takes into account schematic qualities of the self and ultimately views the self within a larger cognitiv e structure. Lastly, t he self is inherently connected to an individual's actions, attitudes, and memories, and could be considered an object of attitude.

PAGE 11

4 Brewer's (1991) Optimal Distinctiveness model offered a less individuated framework for the self; thi s model connects the self to groups and ultimately, the external world. The model is grounded on the assumption that social identity is derived from a tension to be both similar t o others, yet unique and distinct. I dentifying with a group allows an indivi dual to be simultaneously the same and different. An individual's self concept is defined as including both personal identity and social identity. While personal identity encompasses the qualities that differentiate the individual from others, social ident ity connects the individual to groups, thus depersonalizing the self. The self concept is a fl exible and contextual construct; depending on the level of identity, the motivations and ultimately the meaning of an individual's self changes. If individuals st rive for a moderate level of inclusion in a group, they reach a point of optimal distinctiveness: the equilibrium of both an individual's sense of assimilation and differentiation in the group. Brewer (1991) argued that for a group to survive, it must main tain distinctiveness from other groups, and that personal identity must be shifted to soci al identity. This model provided support for the self as not only diverse, but also as inherently connected to other various individuals or groups. Theory and Measur ement The idea of t he s elf as multi faceted was a monumental step in the research; the s elf is ult imately expansive and inclusive. This idea influenced close relationshi p research. The Self Expansion m odel (Aron et al., 2005) adopted the idea of the complex s elf in defining relationship closeness or the degree to which an individual includes his or her close other in the sense of self

PAGE 12

5 Aron et al. (2005) summari zed the main points of the Self Expansion model. They proposed a motivational theory, in which the s elf is motivated to expand in order to explore, enhance, and increase potential efficacy, through the inclusion of others. Another important facet of the model proposes that the cogni tion related to the s elf is inseparable from that of others In other words, when a person thinks of his or her self, this thought process is related to thinking about his or her close other. Ultimately, the three main aspects of the other that the s elf can strive to achieve are resources, perspectives, and identiti es. Resources imply the perception of another's resources as one's own, and a basic inheritance of the other's rewards and costs. Perspectives are defined as experiencing the world from the other's point of view, while identity focuses more on the other pe rson's features central to his or her identity A person's resources are central to the motivational aspect of the model, and it's arguably the main benefit of includ ing the other in the self. The Self E xpansion model not only provided a new method of thin king about relationships and overall closeness between the self and other, but also offered a new perspective about cognitive aspects of the self as inclusive, expansive, and ultimately complex. In support of the Self Expansion model, Aron et al., (1992) created a single item pictorial measure of closeness. One of the main goals in developing this measure was to supplement the shortcomings of the Relationship Closeness Inventory (RCI), which measures closeness as a multidimensional construct and includes factors of frequency (time spent together), diversity (varying interactions), and strength (influence of plans). Because the RCI is only appropriate for college students and takes 10 15 minutes to complete, a more efficient single item pictorial measure of relationship closeness was

PAGE 13

6 created the Inclusion of Other in the Self (IOS) scale. The measure presents participants with a set of seven Venn like diagrams, with s elf labeled on one circle and o ther for the other circle From the first diagram to the se venth, the degree of overlap between the self and o ther circles increases (s ee Figure 1 ). The idea of overlapping selves has been popular in the field, and Aron et al. (1992) argued that the IOS appropriately represents closeness between the self and other The main purpose of Aron et al.'s (1992) research was to evaluate the utility of the IOS scale and ultimately to expand upon the concept of closeness, through a variety of separate studies. In the first study, the IOS' reliability and validity (convergent, discriminant, and predictive) were exam ined, as well as the possibility of the measure eliciting social desirability effects. Participants were given a questionnaire containing the IOS, RCI, and Subjective Closeness Index in the first section, and emotional tone items, Sternberg's Intimacy Scal e, an alternate IOS scale using diamonds, the Expected Distress Scale, and several social desirability measures for the second section. Participants were then retested with the same questionnaire two weeks later. The IOS showed test retest reliability, int ernal consistency reliability, concurrent validity, convergent validity, discriminant validity, and predictive validity. Social desirability was shown to have very insignificant effects on the IOS responses. Additionally, exploratory factor analyse s of the IOS, RCI, Subjective Closeness Index, and the Sternberg Intimacy Scale, allowed for the development of two overall factors that the IOS had a high to moderate effect on: feeling close and behaving close. Through a second replication study, Aron et al. (19 92) provided even more support for their measure of closeness and the overall empirical potential of the IOS To support the generalizeability of the IOS, another series of studies

PAGE 14

7 were performed concerning the use of the IOS in a variety of populations an d experimentally created relationships, while also assessing participants' subjective meaning of the IOS; the results from the study indeed supported the use of IOS in a variety of populations. Overall, Aron et al. (1992) presented strong empirical suppo rt for the psychometric properties of the IOS. The interconnectednes s of the self and other offers an elaboration on the variable of closeness, while al so clearly supporting the Self E xpansion model, through its pictorial representation of an inclusive rel ationship. The Reaction Time Effect One of the first studies that provided a promising way of testing the concept of IOS utilized reaction time methods to demonstrate evidence for the self as a prototypical representation (Kuiper, 1981). The primary focus of this study was to provide insight on the self as a prototype based on a proposed pattern of data. In a computer based cognitive task it was predicted that participants' r eaction time response s to traits most descriptive of themselves should be faster than reaction time responses to traits that only moderately described themselves Participants first completed a computer task that recorded reaction time responses. Participants responded either yes or no for each of the 24 trait words that were presented on the computer screen to indicate whether the words de scribe d themselves After a 5 minute filler task, participants then completed self ratings, on a 9 point scale between extremely unlike me (1) to extremely like me (9) for each of the 24 words. Patte rns in the data supported the hypotheses such that participants responded the fastest to traits that were the most descriptive of the self and slowest to the least self descriptive traits ; this was termed the descriptiveness effect. Kuiper's (1981) study

PAGE 15

8 incorporated a co gnitive aspect in studying the s elf, and provided essential methodological potential for future research concerning the s elf as flexible and inclusive. Meuller, Thompson, and Dugan (1986) followed up with Kuiper's (1981) stu dy by not only investigating trait similarities, but also by focusing more on trait distinctivene ss and its meaning Participants were instructed to rate the descriptiveness of 120 trait adjectives for themselves in a computer based task, much like Kuiper' s study (1981). For each trait, participants indicated their response as, "describes you" or "describes most students;" the response choices were on the screen throughout the duration of the task. They rated these traits in a randomized order two separate times, with a different order of traits presented each time. Participants then did a filler recall task, and were asked afterwards to rate the distinctiveness for each trait that the participant judged as self descriptive ("describes you") The study opera tionalized distinctiveness as the extent to which the trait described the individual, on a scale of low (1) to high (7). Results supported Kuiper's (1981) descriptiveness effect, such that participants responded faster to the traits that they rated at the extremes of descriptiveness ("describes you") than to traits that were not self descriptive ("describes most students") The results from this study also a dded a distinctivenes s effect to the pattern of data. T raits that were rated lower on the scale from the second task were distinctive traits. In the computer task, participants recognized the distinctive traits as self descriptive more slowly than the traits that participants thought described most students. Meuller et al. (1986) made a point to redefine distinctiveness as having two aspects: not only how much the trait is true of the individual, but also how many other people share the trait with the individual.

PAGE 16

9 Together, the descriptiveness and distinctiveness ef fects provided a methodological basis for testing the IOS concept Ar on, Aron, Tudor, and Nelson (1991) provided evidence for the Self Expansion model and the IOS through a reaction time experiment, similar to previous methods (Kuiper, 1981; Meuller et al. 1986). Aron and colleagues (1991) proposed an overlap of self and other mental representations, or a cognitive significance of being in a close relationship. They formed their argument through three perspectives of closeness, and subsequently three exper iment s. C loseness was viewed as a changed resource allocation strategy, a changed actor/observer perspective, and as vicariously sharing other's characteristics. As a whole, these ideas lend to the overall theory that regards a close relationship as fallin g on a continuum. Aron and his colleagues (1991) primarily sought to consider the close relationship as the inclusion of an individual's significant other in his or her sense of se lf, and to demonstrate the Self E xpansion model through diverse methodologie s. The first experiment hypothesized that because people in a close relationship should theoretically perceive the other's resources as their own, that they should make less of a distinction between the self and other when allocating money between the t wo. A 3 (other : close friend, friendly acquaintance, stranger) x 2 ( choice: other would know choice, other would not know choice) within subjects design was used; the results indicated a strong main effect for type of other, such that the smallest differen ce in allocation amount was between the self and close friend Several follow up experiments also confirmed their hypothesis that the close relationship lies on a continuum. The second experiment involved both figure/ground and actor/observer perspectives; Aron et al. (1991) reasoned that the other s hould be more like the ground (s elf). In this

PAGE 17

10 experiment, participants were presented with 60 concrete nouns on a screen. After each noun appeared, they were requested to picture themselves or a target person (i n this case, their mother or Cher) interacting with what the noun was referring to. Participants were then asked to write down the image on a sheet of paper, and after all of the trials were completed, the sheets were removed and participants were then ins tructed to recall as many nouns as possible. Because participants recalled more nouns with the unknown other (Cher), it was reasoned that the close other was indeed more like the ground than figure, and therefore less perceptually distinct and more like th e self. The third and final experiment was grounded on the descriptiveness and distinctiveness effects (Kuiper, 1981; Meuller et al. 1986). Participants, from a sample of married couples, were instructed to rate the descriptiveness from a set of 90 trait adjectives of themselves and their significant other in a questionnaire. Then, participants performed a second self descriptive task on a computer, in which they were asked to respond yes or no if a trait that appeared on the screen (from the same set of 90 traits) described themselves Aron e t al. (1991) proposed that if there is a cognitive overlap between the self and spouse, there should more of a distinction in responses with traits that mis match between the self and spouse, that is traits that are true of the self but not the spouse. If a trait that describes the spouse mismatches the trait participants are focusing on (the same self descriptive trait), mental processing becomes inhibited, a nd therefore participants are slower to respond. However, if the traits match one another, mental processing is facilitated, which yields faster reaction times. As a control, they included the unknown target other (Bill Cosby) and results indeed supported their hypotheses. The follow up study then included the IOS scale which allowed for

PAGE 18

11 correlations between self/spouse and similar/dissimilar differences in reaction time with perceived closeness. The closer the couples were, there was a greater overlap of mental representations between the self and spouse. These experiments as a whole paved the way for future research on overlapping self representations. The Self Expansion m odel posits that individuals strive to include their close others in their sense of selves in order to acquire the other's resources (Aron et al., 2005). The IOS measures relationship closeness, or the degree to which a close other is included in an individual's sense of self (Aron et al., 1992). Through a series of reaction time experiments the Self Expansion m odel not only gained credibility and serious consideration as a way for explaining close relationships, but more importantly provided concrete empirical potential for various future self overlap exp eriments (Kuiper, 1981; Meuller et al., 1986; Aron et al., 1991). The multi faceted nature of the self was applied to the study of close relationships, but the central idea of overlapping self representations yields more possibilities for the expansive sel f. From Other to Ingroup After the conception of the self as complex and inclusive was applied to the study of close r elationships, research ers focused on group relations If the self includes a close individual, it could also include an identified c lose group. The Self Expansion m odel inspired research on social identification, or the degree to which individuals include their identified ingroup in their sense of self. Therefore, for the purposes of this study, ingroup is conceptualized more abstractly in terms of social or ingroup identification.

PAGE 19

12 Social Identification In order to understand the process of ingroup identification, the structure of the concept must first be examined There are numerous models that conceptualize social identification as either a solitary idea or multidimensional construct. Perhaps the most influential theory of identification is Social Identity Theory (Turner, 1986). The model proposes that an individual's self concept is derive d from membership in a categorical group. The construct of the self concept is viewed as enduring, but also multi faceted and varying with group contexts ; it's a mediator between an individual's environment and behavior. According to Turner (1986), a socia l group is defined as two or more individuals who perceive themselves to be members of the same social category. In order for a group to be well defined and stable, there must be a satisfaction of needs, attainment of goals, and validation of attitudes. Tu rner (1986) argued for a shift in the definition of a social group to be concep tualized as cognitive, and viewed social identity as the structural cognitive aspect of an individual's self concept, from which self images then emerge. Social categorization i ncludes both inductive and deductive processes. When an individual is assigned traits based on his or her membership, that is the deduction aspect of social categorization, while an individual's identification as a member would be induction. Stereotypic pe rceptions are also inherent in social categorization in the group, and this increases the perceived intergroup similarity. While social cohesion is not always necessary for group formation and behavior, social identities are essential; individuals must be cognizant of their categorical membership within the group. Positive distinctiveness and self esteem within the group is also essential for the survival of the

PAGE 20

13 group as a whole. Group behavior is ultimately composed of collective positive social identities and the various self concepts from which they originate (Turner, 1986) Social Identity T heory triggered both the development of varying models of social identification and the research that attempted to evaluate these models and narrow the structure. Br own, Condor, Mathews, Wade, and Williams (1986) evaluated the utility of three theories of intergroup relations : Realistic Conflict T heory Contact Hypothesis, and Social I de ntity Theory. According to the Realistic Conflict T heory, intergroup behavior, ant agonistic or conciliatory, reflect s the state of t he group's goal relations. The Contact H ypothesis proposes that increased contact of individuals with others in the group lessens intergroup tension and hostility. Social Identity T heory posits that individual s social identities develop from their group memberships through intergroup comparison (Turner, 1986) Because of the previous uncertainties pertaining to group identification measures used in past research, one of the main aims of this study wa s to create a more robust measure that can be used with a larger sample. Data were collected through interviews, and m any of the results indicated that relationships between variables of intergroup differentiation, group identification, perceived group c onfli ct, structure, and meaning of group identification were moderate to weak. Group identification in particular was a weak pre dictor of group differentiation. H owever the meaning and process of group id entification, as measured by a scale developed in t he study, can be different depending on the context of the group. This ultimately led to the conclusion that group identification could be regarded as multidimensional, just like the self (Brown et al., 1986). Through a series of five studies, Cameron (20 04) sought to develop a new multi

PAGE 21

14 faceted model of social identification, which consisted o f ingroup ties, cognitive centrality, and ingroup affect Centrality was defined as the enduring salience of an individual's group membership and the subjective impo rtance of the group to an individual's self definition. Ingroup affect was conceptualized as the emotional consequence of group membership, while ingroup ties were operationalized as the extent to which individuals feel connected to the group. In the fir st s tudy, Australian undergraduate u niversity students were given a questionnaire that included personality and attitudinal measures, as well as an 11 item measure of social identification. T he second study also recruited u niversity students, but the survey utilized gender related attitude measures instead of general attitudinal measures and a gender derived social identification measure composed of 15 items. Participants were administered the questionnaire in mixed sex grou ps of 10 to 15 people as well. The third study used a measure of national identification and the questionnaire was administered to participants in several different areas of Australia. In the fourth study, Cameron (2004) assessed the gender derived social identification measure, using 12 items adapted from the third study. Las tly, the fifth study recruited u niversity students, as in study 2, and the questionnaire that participants received employed the 12 item measure used in the third and fourth study and additional measures of social identity. Participants returned one week later to complete a second questionnaire containing the 12 item social identity measure. Confirmatory factor analyses revealed that the items on the social identity measure fit the da ta from the five studies significantly better than the previous models of social identification (Brown et al., 1986). Because ingroup affect and ties are the most

PAGE 22

15 emotional aspects of identity, these factors showed consistently the strongest relationships in all of the samples. The three factor model of social identity showed overall internal consistency, test retest reliability, convergent validity, and discriminant validity (Cameron, 2004) While this research examined the structure of social identity and the significance of various group memberships to the self concept, it ultimately served as the foundation in defining ingroup identification. Bringing Models to a Measure Previous models of social or ingroup identification focused on the structure of the construct rather than the process, or how an individual identifies with an ingroup. Cameron (2004) defined this process as the merging of mental represe ntations of self and group. While the groups that individuals socially identify with vary between one a nother, the process of ingroup identification remains the same. Assimilating ideas from the Self Expansion m odel and Social Identity T heory, Tropp and Wright (2001) created an inclusive definition and measure for ingroup identification: the degree to whi ch ingroups are included in an individual's sense of self. Researchers also developed a measure of ingroup identification, or the Inclusion of Ingroup in Self Scale (IIS) (Tropp & Wright, 2001). The measure itself was explicitly based on Aron et al. 's (199 2) IOS scale such that it utilized seven Venn diagram pictures with labels of Self and Group next to the overlapping circles Each pair of circles overlapped slightly more than the previous pair (see Figure 2). The ir first study examined the cognitive s ignificance between the self and ingroup by evaluating differences in reaction time for descriptive words associated with self and ingroup, in relation to IIS scores The logic was similar to Aron et al. 's (1991) such that

PAGE 23

16 participants should respond slowe r to traits that mismatch between self and ingroup than to matching traits. These reaction times should also vary in accordance with IIS scores, such that those who have high IIS scores should have quicker reaction times because their sense of ingroup and self are more closely connected. Results from the first study confirmed IIS as a significant moderator between the ingroup and self through the relation of the reaction time data and the IIS scores While the first study provided construct validation for t he IIS, the second study focused on the concurrent and discriminant validity of the IIS by associating it with other measures of ingroup identity. Tropp and Wright (2001) hypothesized they would find positive relationships between the IIS and other conce ptually related measures, such as the Collective Self Esteem Scale, Aspects of Identity Questionnaire, Group Relative Deprivation, and Support for Collective Action. All of the hypotheses concerning the construct validity of the IIS were supported However, in the third study, Tropp and Wright (2001) replicated and modified the second study by adding a separate 3 item measure of ingroup identification to investigate how individuals feel connected to their ingroups. Results for this replication stud y were similar to the previous one, such that the IIS was strongly positively correlated with the additional measure of identification. Lastly, the fourth study considered the test retest reliability of the IIS. Participants were given the IIS at two times one to three weeks apart, and were expected to generate similar scores for both testing sessions. Analyses revealed the reliability of the IIS, as the scores were highly positively correlated with one another. This study not only provided a comprehensiv e way of conceptualizing and measuring ingroup identification as a whole, but also highlighted the research potential of the IIS and self expansive mental representations

PAGE 24

17 Reaction Time Revisited In order to further evaluate the construct validity of the IOS and IIS, reaction time methodologies were employed. Tho ugh briefly mentioned in Tropp and Wright's (2001) study, more concrete reaction time studies assessed the idea of overlapping self and ingroup menta l representations, just as Aron et al. 's (1991) study examined overlapping self and other mental representations. Drawing upon ingroup identification and ultimately the more experimenta l side of the Self Expansion m odel, Smith and Henry (1996) provided response time evidence for the i nclusion of the i ngroup in the self. Because most of the e vidence behind Social Identity T heory used judgmental measures, Smith and Henry (1996) directly m odeled their method after Aron et al. 's (1991) study concerning reaction time evidence showing the self and other as l inked in a single mental representation. In this procedure pa rticipants who were liberal arts or engineering majors and either of Greek or non Greek groups rated themselves and their target groups (liberal arts/engineering majors and Greeks/non Greeks) from extremely like (1) to extremely unlike (7) on the same lis t of 90 traits utilized in Aron et al. (1991). Then, participants performed a second self descriptive task on a computer, in which they were asked to respond yes or no if a trait (from the same set of 90 traits) that appeared on the screen described themsel ves or not. Using the same logic as Aron et al.'s (1991) self and spouse reaction time task, it was predicted that matching traits between the ingroup and self should facilitate reaction times whereas mismatching traits should inhibit reaction times. Because participants may po rtray themselves in a more socially desirable and positive way, it was expected that

PAGE 25

18 participants should be quicker to respond to positive traits and slower to respond to negative traits The primary reaction time predictions were supported; participants were faster at responding to traits that matched between themselves and their identified ingroups than to mismatching traits (Smith & Henry, 1996) This pattern was obse rved in all of the target groups, and for no self descriptive ness responses (responding no to the same trait for both self and ingroup descriptiveness) as well as yes responses (responding yes to the same trait for both self and ingroup descriptiveness). P articipants were also faster at accepting positive traits for themselves and rejecting negative ones, which accounted for social desirability in the sense that participants viewed themselves in a more positive and in a socially desirable way. The results f rom this study show a clear relationship between the self and ingroup; they are certainly not separate and distinct cognitive structures, but rather a connected working mental representation. Smith and Henry (1996) provided a solid methodology for investig ating the cognitive implications of the extended nature of the self. This idea of overlapping self representations and the related reaction time methodology was then applied to attitudes. Coats, Smith, Claypool, and Banner (2000) extended Smith and Henry 's (1996) findings by adding attitudinal variables and correlating the reaction time results with various measures of group identification. In addition, they aimed to connect self and ingroup reaction time evidence to various group identification measures, like the Group Attachment Questionnaire and the Social Identity Scale (Brown et al., 1986).

PAGE 26

19 Coats and colleagues (2000) used a between subjects design such that half of the participants w ere rating the descriptiveness of traits between themselves and their ingroups (trait condition), while the other half of the participants were asked questions regarding their likes and dislikes for objects that they ascribe to their self and ingroup (atti tude condition). In the trait condition, Smith and Henry's (1996) method was replicated, while in the attitude condition, participants were instead presented with a list of 90 relevant attitude objects for the college sample (ex. parties, tattoos, science, beer), and asked to rate the extent to which they thought themselves and groups (Greeks and non Greeks) liked or disliked the objects. For the computer task, participants were shown the same 90 attitude objects on the screen and asked to respond as quickl y as possible to choose "yes" if they liked the object and "no" if they did not It was hypothesized that there would be a faster reaction time effect for attitude objects on liking judgments that match those of participants' ingroups than those that misma tch While t his extension explicitly differed from previous research because of its incorporation of attitudes, it still conceptually supported the findings for self and ingroup overlapping mental representations. Results from the study indicated that th e more that participants identified with their group, based on the identification measures, a larger reaction time effect between responses that match and mismatch their group was observed (Coats et al., 2000) Moreover the reaction time effect for the tr ait condition self and ingroup matching traits was just as strong for the attitude condition. These findings support the extending nature of the self in that the relationship between the self and ingroup can be replicated and applied to attitudes.

PAGE 27

20 Smi th, Coats, and Walling (1999) replicated previous reaction time studies concerning overlapping self and ingroup mental representations. The hypotheses in this study were the same as those in Smith and Henry's (1996), except they also replicate d Aron et al. 's (1991) study concerning react ion time evidence for the Self E xpansion model, and investigate d whether the IOS scale is a significant moderator of relationship closeness. Smith et al. (1999) instructed 87 participants to complete the same paper based descriptive tasks with a list of 90 trait words from Aron et al. (1991) to describe themselves, their groups (Greek or Non Greek) their current romantic partners, and a stranger (Bill Cosby) (Smith & Henry, 1996) Participants were asked what their group affiliation was, and then were subsequently given the IOS scale. Then, participants were aske d in two separate blocks to do a computerized reaction time task in which they were asked to respond yes or no if a trait (from the same se t of 90 traits) that appeared on the screen described themsel ves or not. The predicted interaction between the IOS, self ratings, and p artner responses was supported participants responded faster to matching traits between themselves and their partners t han to mismatching traits The self and ingroup interaction was also supported in that participants were also significantly faster at responding to matching traits than to mismatching traits Overview Ingroup identification research elaborated on the id ea of the multi faceted self. Through additional studies, the original reaction time methodology used by Aron et al. (1991) was strengthened, refined, and ultimately viewed as an appropriate way to study self inclusive relationships (Smith & Henry, 1999; C oats et al., 2000; Smith et al., 1999).

PAGE 28

21 The complex nature of the self was applied to the study of both close relationships and ingroup identification and the fluidity of the self was reflected in the relationships between the self and various other socia l entities, like a close other or ingroup. W hen studying self overlap mental representations, one such relationship accounts for context even more than the relationship between the self and ingroup. This is the relationship between self and community. From Ingroup to Community Ingroup identification is distinguished from community in two primary ways Much of the literature on these self inclusive topics view ingroup more abstractly through social or ingroup identification (Cameron, 2004; Tropp & Wright, 2001) whereas community is represented through different physical contexts or concrete, tangible and realistic groups. Additionally, even though measures of ingroup identification and psychological sense of c ommunity are similar, what the scales actually measure is significantly different. Community must be defined first, in order to distinguish it from ingroup. This distinction merits further research on the connection between self and community. Psychological Sense of Community In one of the m ost influential art icles in community psychology, McMillan and Chavis (1986) developed a cohesive theoretical framework for defining sense of community. Researchers examined group cohesiveness literature and subsequently developed the various elements that contribute to an individual's overall sense of community. The most important concepts that were repeated in the literature include neighboring, length of residency, home owner ship, and community satisfaction.

PAGE 29

22 However, the most important recurring theme wa s the emotional implication of sense of community. While previous research simply verified the existence of a sense of community, McMillan and Chavis (1986) proposed an integrative definition that consists of four elements: membership, influence, integrati on and fulfillment of needs, and shared emotional connection. Membership includes the sense of belonging within the community and influence pertains to the community members individually making a difference within the group. I ntegration and fulfillment of needs relates to reinforcement such that the community reinforces status and competence of the individual while also meeting the various needs of its members The last element shared emotional connection, wa s arguably the most important for sense of commu nity because it is what ultimately holds a group together ; members can mutually understand the group's history, common places, and past experiences (McMillan & Chavis, 1986 ) When a group gives individuals a p sychological sense of community (PSOC), or a sense of membership, influence, integration and fulfillment of needs, and shared emotional connection this group may be considered a community. Identification as a D is tinction B etween Ingroup and Community Obst, Smith, and Zinkiewicz (2002) evaluated the structure of PSOC and also began investigating PSOC in relation to ingroup identification. Their study aimed to provide more empirical support for McMillan and Chavis' (1986) hypothesized PSOC dimensions and to elaborate a new dimension of Conscious Ident ification. It was hypothesized that this new i dentification dimension would emerge as distinct from other PSOC dimensions. Obst et al. (2002) recognized the importance of context, and studied residents of rural, regional, and urban geographical communities Participants hailed from

PAGE 30

23 Australia; there were 344 in urban areas, 201 in regional areas, and 122 in rural areas. The questionnaire utilized in the study inclu ded a large battery of measures that assessed demographics, dimensions of PSOC, and identificat ion. Hierarchic al multiple regression analyses examined how PSOC dimensions predicted overall sense of community. Results supported their hypotheses in that all PSOC dimensions were significant predictors of sense of community. Additionally, Conscious Iden tification emerged as predictive of overall sense of community, but a significantly distinct factor in comparison with the PSOC factors. Obst et al. (2002) provided additional support for McMillan and Chavis' (1986) model of sense of community as well as r evealing identification as a separate construct from PSOC Even though research that focuses on both ingroup and community acknowledge s identification as a quintessential aspect of PSOC, specific to members hip to the community, measures of ingroup identification are no t included in the studies. Obst and White (2005) addressed this empirical discrepancy while also providing further support in the explicit differentiation between ingroup and community. Obst and White (2005) assessed three different community contexts and explored how situational salience of the communi ties or the cognitive accessibility of an individual's membership in the group might influence the results Before the participants were given a survey, they were presented with instr uctional information regarding their membership in their local neighborhood, student community, and a self chosen interest group. The questionnaire i tself included measures of PSOC, social identification and situational salience. PSOC dimensions were base d on McMillan and Chavis 's (1986) model, while the three dimensions of social identification were based on Cameron's

PAGE 31

24 (2004) model of social identification Hierarchical regression analyses revealed that salience and all three subscales of identification we re significant predictors of PSOC. Ingroup ties were the strongest predictor of PSOC for each group and greatly overlapped with membersh ip and emotional connection; this demonstrated the relationship between identification and PSOC as a whole. However, ing roup affect and c entrality were more distinct, such that they were less closely associated with PSOC, and more specific to identification. These findings were interpreted as evidence for the relationship between ingroup (identification) and community (PSOC ), in that certain aspects of each con struct were indeed very similar, but significantly distinct (Obst & White, 2005) Another study that distinguished ingroup and community was Obst and White's (2007) investigation of how choice affects identification a nd sense of community, in three differe nt contexts : neighborhood, university student, and self chosen interest group s It was predicted that levels of social identification would increase as the degree of choice over group membership increases, and that le vels of PSOC would function in the same manner. The neighborhood, university student, and self chose n interest groups were first analyzed in a pilot study, and were confirmed to reflect the appropriate levels of choice low, medium, and high, respective ly (Obst & White, 2007) In the main study, participants were presented with a questionnaire, which included measures of PSOC and identification, for each group. To control for salience, participants were primed by having them write down the name of the group before answering questions about the group, as well as including items at the end of the survey that served as salience checks. Results indicated that all of the predicted patterns were significant; participant's degree of choice

PAGE 32

25 indeed influenced th eir level of P SOC and social identification, such that participants showed the highest levels of each variable with their self chosen interest group. Though PSOC and identification varied in similar ways and were certainly conceptually related in this stud y, they were still treated as two separate empirical entities Importance of Context Because ingroup may be more abstract and community may be more concrete, the relationship between self and community requires the most attention to the physical context o f the group The following studies are examples of the importance of context when studying community. They also treat community as the more physical and concrete group under study, which supports the distinction between ingroup and community. Prezza and C ostantini (1998) examined the association between individuals' sense of community and overall life satisfaction in three different territorial contexts in Italy. I t was hypothesized that a greater sense of community is positively associated with a good qua lity of life perceived by the self through life satisfaction. Three separate samples were utilized in this study participants were from a small town, small city, or a city. The survey included measures of sense of community, life satisfaction, perceived social support, efficiency of comm unity services, and self esteem along with a questionnaire that contained questions pertaining to demographics and involvement in the community All of the scales were translated into Italian and showed good reliability and validity measures. Analyses on the three different groups showed that sense of community, perceived social support, and life satisfaction were the highest in the small town, as well as the judgment of efficient services ( Prezza & Costantini 1998) C oncerning self

PAGE 33

26 esteem there were no significant differences between the groups. Additional correlational analyses revealed positive associations between sense of community and evaluation of services, life satisfaction, perceived social support, and self e steem in the small town, which supports the previous findings. Only life satisfaction and evaluation of efficiency of services were positively correlated with sense of community in the small city group, and only efficiency of services was correlated positi vely with sense of community in the city. The original hypothesis was only strongly confirmed with the small town territory participants in this group displayed the most sense of community and subsequent life satisfaction. Prezza and Costantini (1998) no t only highlighted the importance of contextual differences within groups, but also discovered a positive association between sense of community and life satisfaction. Another study that assessed the importance of context investigated the relationship bet ween social participation, sense of community, and well being with American, Italian, and Iran ian university students (Cicognani et al ., 2008). Social participation was defined as individuals being involved in making decisions, developing programs, and hav ing a sense of control over the environments that affect them. Community identification, as a facet of sense of community, is important to this study because this variable ties sense of commun ity with social participation. While the primary aim of this stu dy was to examine the relationship between social participation and sense of community, and how this relationship might affect well being, the second aim was to assess the generalizability of this relationship across three different contexts: USA, Italy, a nd Iran.

PAGE 34

27 Cicognani and colleagues (2008) hypothesized that levels of social participation, sense of community, and social well being would be the highest among American students and that the former two variables would be positively related to social well being, in all samples. Some specific differences that were hypothesized between the nations included levels of political engagement being low among Italian students, a higher volunteer involvement for Americans, and religious participation high er among Iranian students. Participants were administered questionnaires with items assessing social participation, sense of community, and social well being. The results supported the hypotheses predicting American students having higher levels of sense o f community and social well being. Social participation was correlated positively with sense of community, which was then positively correlated with social well being. Hierarchical regression analyses assessed whether sense of community and identification with the community mediated the effects of social participation on social well being. With American students, sense of community had a stronger mediating effect, whereas community identification was stronger in both the Italian and Iranian samples. While t his research evaluated relationships between social participation and sense of community on well being, it ultimately highlighted the importance of accounting for contextual differences among different communities because of the differing results between the nations Lounsbury and DeNeui (1996) focused on physical contextual differences based on school size, and developed an internally consistent and reliable measure of PSOC specifically for colleges and universities. It was predicted that collegiate PSO C would be inversely related to the size of the college, and also that there would be a relationship between their measure of PSOC and extroversion (from the "Big Five").

PAGE 35

28 A preliminary PSOC scale w as developed in the first study, which sampled college st udents from 23 schools. The second study, which sampled 1,121 students from 27 colleges, involved the implementation of the co llegiate PSOC 14 item measure. The size of the colleges was coded into the following groups: less than 2,000, 2,000 9,999, 10,000 19,999, and 20,000 or more students. The collegiate PSOC scale was found to be internally co nsistent and reliable and the results supported the hypothesis that smaller colleges foster higher levels of collegiate PSOC. Additionally, students who lived on ca mpus reported higher levels of PSOC for all size groups except the medium to large (10,000 19,999) college group. To account for these findings, Lounsbury and DeNeui (1996) offered the explanation that living at the college might provide more opportunities and activities, which might emphasize feelings of belonging and shared identity. Additionally, participants in the smaller schools had a slight tendency for higher extroversion levels. Because both the size of the schools and whether or not students lived on campus significantly affected the results, this research shows the importance of using different contexts when studying community, especially in a college sample Inclusion of Community in the Self Both the Self Expansion Model and ingroup identifica tion research conceived of the self as multi faceted and inclusive in nature (Aron et al., 1991; Tropp & Wright, 2001). The s e self overlap ping mental representation s account for the contextual nature of the self, but only to a degree. Directly based on the IOS and II S, the connection between self and community was made through the development of the Inclusion of Community in Self (ICS) Scale (Mas hek et al. 2007). This measure was designed to assess community connectedness, or the degree to which individuals include their communities

PAGE 36

29 in their sense of self. Previous research on the IIS was viewed as an essential step in the research, making it possible for new resea rch to expand from the ingroup literature and connect it to community based studies Just like previous work on the inclusive self, it was predicted that individuals include their community in their sense of selves B oth college students and adult incarc erated offenders were sampled in this study (Mashek et al., 2007) Participants were instructed to choose which of the six Venn diagram s on the ICS scale best portrayed their own relations hip with the community at large The degree of overlap between the s elf and community circles increased from not overlapping at all to almost completely overlapping (see Figure 3). It was hypothesized that the ICS scores would map onto the community related constructs of the PSOC, and not the constructs associated with clo se relationship. Additionally, it was hypothesized that ICS scores would be positively related to community helping and negatively related to community hurting. Lastly, researchers aimed to test the reliability of the ICS over a two week period. All of the ir hypotheses were supported except for the negative relationship between community hurting and ICS scores. In study 2, Mashek and colleagues (2007) aimed to expand the generalizability of the ICS, account for social desirability, and to clarify the relationship between community helping and hurting. Results from the second study confirmed generalizability of the ICS while controlling for social desirability. The results from study 2 wer e essentially the same as study 1, except community hurting was significantly related to lower ICS scores in the second study. Overall, this study looked at the efficacy of the ICS scale, and provided evidence toward its reliability and validity as a measu re

PAGE 37

30 for community connectedness to the self. However the theoretical aspect of this measure has yet to be empirically tested. The connection between the self and community accounts for context the most of any self inclusive relationship because community is the more physically contextually sensitive construct than ingroup or close other Even though the relationships between the self with close others and ingroups have been cognitively tested (Aron, Aron, Tudor, & Nelson, 1991; Smith et al., 1999) reacti on time evidence in support of the ICS has yet to be examined. This gap necessitates further research. The Current Study The goal of this study was to provide evidence for the inclusion of the community in the self. The self is contextual and extensive in nature; personal identity has shifted to a more social identity, and the self concept is more complex and multi faceted ( Greenwald & Pratkanis, 1984). The descriptiveness and distinctiveness effects assert that individuals respond faster when responding to traits that describe themselves and slower when responding to traits which are distinctive from the mselves (Kuiper, 1981 ; Meuller et al. 1986). Researchers have applied these principles when relating the self to various other social targets namely a close other and an ingroup. Previous research has shown reaction time evidence for both the inclusion of the other and ingr oup in the self, and ultimately the overlap of mental representations between the self and other and ingroup (Aron et al., 1991; Smith & Henry, 1996; Coats et al., 2000; Smith et al., 1999). Based on both the Inclusion of Other in Self scale ( IOS ) and the Inclusion of Ingroup in Self scale ( IIS ) an additional measure was created: the Inclusion of Community in Self scale (ICS) (Mashek et al., 2007). Exploring the community construct was viewed as empirically

PAGE 38

31 delving deeper into the issue of the contextual s elf and its various other representations, even though researchers viewed ingroup identification research as essential and distinct. However, behavioral evidence supporting the ICS concept has yet to be investigated. As a whole, this study wa s designed to answer the following question: is there a cognitive overlap of self and community mental representations? Previous reaction time methodologies that address self and other representations were modified for the present study (Aron et al., 1991; Smith et al. 1999). The sample consisted of college students. Participants were given a questionnaire in which they evaluated the descriptiveness of 90 traits for themselves and either their school community or local city community. Afterwards, they participated in a computerized self descriptive ness reaction time task, using the same 90 traits. The repeated measures independent variables were Self Survey, as the dichotomized version of the self descriptive ness rating from the survey, and Self Computer, as the yes or no response recorded by the computer program in the second self descriptive ness task. Community, presented as the school community or local city community, was the independent variable and the only between subjects variable. Evaluation was a continuous var iable that served as a stimulus check (S mith et al., 1999). The main dependent variable in the study was reaction time in the self descriptive ness computer task An addition dependent variable was Community Connectedness, measured by the ICS (Mashek et al. 2007). This method addressed the following specific questions: (a) When students describe their community, will traits that match with themselves be associated with reaction time facilitation? (b) Will the traits that mismatch with themselves be associ ated

PAGE 39

32 with reaction time inhibition? Concerning the trait analyses, based on Aron's (1991) and Smith et al. 's (1999) research, it was hypothesized that: H $% : Reaction time responses for traits that match between the community descript ive and self descriptive tasks should be faste r than the traits that mismatch. In previous research utilizing this method, Aron et al. (1991) affirmed that individuals respond ed faster to yes than to no when deciding if the trait was self descriptive Similar rese arch found that participants responded faster to positive traits and slower to negative traits (Smith & Henry, 1996; Smith et al., 1999). Additional findings indicated a response error effect such that the traits that mismatched between the self descriptiv e ness survey and computer tasks were more likely to mismatch between the close other and self descriptive ness tasks (Aron et al., 1991). Concerning the contextual aspect of community, research that has suggested that college students have a greater PSOC no t only for small college campuses than for larger ones, but also if they live on campus than off campus (Lounsbury & DeNeui, 1996). Based on these collective findings, it was hypothesized that: H $& : Participants should attribute more positive traits to t h emselves than negative traits. 1 H $' : There will be more response errors between the questionnaire and computer self descriptive responses, for traits on which self and community mismatch, than for the traits that match between the self descriptive nes s tasks. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!! 1 For the purposes of this study, the survey self descriptive responses were viewed as more accurate than the computer self descriptive responses, because participants were not pressured to respond as quickly as possible.

PAGE 40

33 H $( : Participants will be more connected to the school community than the local city community. Method Participants Sixty underg raduate students from a small liberal arts college in Southwest Florida participated in the study. The sample was composed of 53% women, 45% men, and 2% other The age of the p articipants ranged from 18 to 22 ( M = 20.1 SD = 1. 31 ). Ninety percent of participants r eported living on campus while 10 % reported living off campus. Five percent of participant s reported growing up in the local area. Participants were recruited through the student email listserv and after classes. Materials Questionnaire The survey used a list of 90 positive and negative traits adopted from Aron et al. (1991). For survey descriptiveness tasks, the list of traits was applied to each target (self, community, stranger). As in prior research (Aron et al., 1991; Smith & Walling, 1999), the stranger target was Bill Cosby, and was included as a filler, to obscure the focus of the research. Example traits included creative domineering cheerful shy and ambitious The descriptiveness rating for each trait was on a 7 point scale from extremely unlike (1) to extremely like (7) The list of traits was rand omized, so that each participant was presented with a different randomized order of traits for each target. The three targets (self, community, and Bill Cosby) were presented on three separate pages. The page order of the three targets was counterbalanced within the questionnaire packet. Depending on the condition, the specified community was either participants' school community or the surrounding local city community. The school

PAGE 41

34 community (New College community) was defined as including fellow students, p rofessors, employees, activities on campus, and any events that are exclusive to the school that occur on campus. The local city community (Sarasota/Bradenton community) was defined as including local events off campus, non campus students, off campus jobs, and all of the local people and ac tivities not affiliated with the c ollege. The next page of the questionnaire included a modified version of the ICS (Mashek et al., 2007). Depending on the community group participants were pl aced in, the measure used either the school community or the local city community. Both items incorporated a replication of the ICS's six pairs of overlapping circles, except Community at large' was replaced according ly for each item. The degree of overla p between the self and community circles increased from not overlapping at all to almost completely overlapping. The last four items on the survey were demographic questions. Participants were i nstructed to write in their age, circle their sex (male, fema le), and select their housing situation for most of the current school year (on campus, off campus). They were also instructed to answer "yes" or "no" if they grew up in the Sarasota/Bradenton area. The survey was comprised of 275 total items (see Appendix for full questionnaire). Reaction time task. The computer software utilized was Superlab 4.0, which is a stimulus presentation program The presentation designed for this study utilized the same 90 traits from the survey (Aron et al., 1991). Each trait was presented at the center of the computer screen, one at a time, and in a different randomized order for each participant. Traits were presented in a 14 size Arial font, in the color black on a white background Participants were equipped with a response pad to press either a "yes" or "no" as

PAGE 42

35 indicated on the keys as their response to the stimulus. This program recorded participants' responses and latencies. Procedure Volunteers were recruited for this study on a convenience basis through studen t listserv emails. The researcher gave each participant a consent form before administering the questionnaire They were informed that they were going to first make a series of judgments on paper and then make similar judgments on a computer. Participants were then notified that their responses we re anonymous, and to increase anonymity even further, participants were asked to seal the completed survey in a blank white envelope when they were finished Participants were randomly assigned to either the school community condition or the local city condition and were given either the school community questionnaire packet or the local city questionnaire packet accordingly The packets were numbered to correspond to computer files, in order to be matched for the analyses. The beginning of the survey included the following statement, constructed by the author for the study: "This survey investigates how New College of Florida students view their social relationships with the community." Directions at the beginnin g of each page in the packet instructed participants to rate the descriptiveness of themselves, either their school community or local city community, and Bill Cosby As in previous research (Smith et al. 1999) the following note was also included after the instructions for the community descriptions and the ICS scales : "Whereas not all members of any community are exactly alike, community members tend to be similar on many traits however, you should provide your personal opinion about the general chara cteristics of the

PAGE 43

36 community. For each target, participants rated the descriptiveness of each of the 90 traits on a 7 point scale from extremely unlike (1) t o extremely like (7) Next, participants completed the modified ICS. In the school community condition, the ICS instructed participants to "circle the picture that best describes your relationship with the New College Community." For the local city condition, the measure used the same instructions as the firs t, however "New College Community" was replaced with "Sar asota/Bradenton Community Each pair of the six circles was labeled appropriately, with "Self" next to the left circle and either "New College Community" or "Sarasota/Bradenton Community" next to th e right circle. After completing the survey, participants were seated at a computer for the second self descriptive task. As a cover story for the task, participants were informed that the researcher was comparing trait responses between participants who had previously completed the questionnaire and those who had not (Smith & Henry, 1996). The on screen instructions for the computer task that were constructed by the author were the following: You are about to be presented with some words. Press the "yes" key if you think the word describes you. Press the "no" key if you do NOT think the word describes you. Please respond as quickly and accurately as possible. Press any key to begin the task. As in earlier studies (Aron et al., 1991; Smith & Henry, 1996; Sm ith et al. 1999), the same 90 traits that appeared in the questionnaire were displayed on the computer screen, in a randomized order for each participant. Following the completion of the computer

PAGE 44

37 task, participants were thanked for their time and particip ation, debriefed, and compensated. The primary trait analyses in this study compared participants' community descriptiveness responses from the survey with the self descriptiveness responses on the computer task. For the subsequent error analyses, self descriptiveness survey responses were compared with self descriptiveness computer analyses, to evaluate whether partic ipants made errors by evaluating a trait as self descriptive on t he survey but not on the computer task, or vice versa. For all analyses, descriptiveness ratings from the survey were dichotomized such that responses ranging from 1 3 were coded as "No" and responses from 5 7 were coded as "yes." Responses of 4 (the midpo int in the scale) were discarded, as in previous research (Smith & Henry, 1996; Smith et al. 1999). Therefore, the number of matching and mismatching traits varied for each participant. For the trait evaluation variable, eight separate randomly selected p articipants from the same school rated the same 90 traits used in the study as negative or positive. A 17 point scale was used, which ranged from Negative ( 8) to Positive (+8) Mean evaluation scores for each trait were coded such that means ranging from 8 to 1 were coded as "neg ative" and means ranging from 1 to 8 were coded as "positive." Mean scores of 0 were possible, and these traits would have been excluded from the evaluation analysis. However mean scores of 0 did not occur. Responses for the ICS measure were treated as a continuous variable, from the least overlapping pair of circles (1) to the most overlapping pair (6) (Mashek et al. 2007) For the reaction time analyses, responses less than 300 ms and greater than 5,000 ms were excluded (Smith & Henry, 1996).

PAGE 45

38 Results Trait Analyses In order to determine whether there was a relationship between matching or mismatching trait responses and community type ( H $% ), a 2 (trait : match/mismatch ) x 2 (community : school/local city ) mixed ANOVA with repeated measures on the first factor was calculated The dependent variable was reaction time There was a main effect of trait, such that there was significant difference between matching ( M = 1423.25, SD = 321.56) and mismatching ( M = 1 547.68, SD = 355.84 ) reaction times F (1, 58) = 14.42, p < .001, = .20 While there was no main effect of community, F (1, 58) = 1.12, p = .29 = 02 the interaction betwee n community and trait was significant F (1, 58) = 4.38, p = .04, = .17 These results are shown in Figure 4. The t test calculated between matching ( M = 1346.17, SD = 300.46) and mismatching ( M = 1539.17, SD = 357.11) reaction times for the school community indicated a significant difference t (29) = 3.87, p = .001. An additional t test between matching ( M = 1500.33, SD = 328.32) and mismatching ( M = 1556.20, SD = 360.47) reaction times for the local/city community revealed insignificant results, t (29) = 1.31, p = .20. Participants were significantly faster at responding to matching traits than mismatching traits in only the school community. Because reaction time responses are not normally distributed, participants' responses wer e transformed using a logarithmic function An additional repeated measures ANOVA using the same design, on the transformed scores indicated a significant main effect of trait, such that matching scores ( M = 3.14 SD = 0.10 ) were significantly different from mismatching scores ( M = 3.18 SD = 0.11) F (1, 58) = 15.92,

PAGE 46

39 p < .001, = .22 There was no main effect of community, F (1, 58) = .95 p = .34 = .02 The interaction between trait and community was also significant, F (1, 58) = 6.06 p = .02 = .10 A t test on matching ( M = 3.12 SD = .09 ) and mismatching ( M = 3.18 SD = .10 ) scores for the school community was significant t (29) = 4.37 p < .001. However, the t test between matching ( M = 3.17 SD = .10 ) and mismatching ( M = 3.18 SD = .12 ) reaction times for the local/city community revealed insignificant results, t (29) = 1.08 p = .29 The results for the transformed scores were similar to the results of the original reaction time scores; though participants responded faster to matching traits than mismatching traits, this effect was only significant in the schoo l community In order to investigate the number of matching traits and number of mismatching traits that participants reported the proportions of the reported matching or mismatching traits to the total possible number of matching or mismatching traits were calculated. Thus, the matching and mismatching proportion for each participant added up to 1. A t test between the proportions of reported matching traits in the school community ( M = .68 SD = .11 ) and the local/cit y community ( M = .52 SD = .11) yielded a significant difference, t (29 ) = 5.45 p < .001. An additional t test between the proportions of reported mismatching traits in the school community ( M = .32 SD = .11 ) and the local/city community ( M = .48 SD = .11 ) was also significant t (29) = 5.46 p < .001 These results showed that participants reported more matching traits in the school community than the local/city community, and fewer mismatching traits in the school community that the local/city commun ity. Evaluation

PAGE 47

40 The second hypothesis was developed primarily to assess social desirability, such that participants should attribute more positive traits to themselves than negative traits across community Mean percentages were calculated for the propo rtion of the number of times participants attributed a positive or negative trait to themselves over the total number of positive or negative traits. A 2 ( evaluation: pos itive /neg ative ) x 2 (community : school/local city ) mixed ANOVA with repeated measures on the first factor was analyzed The mean percentage was the dependent variable The mean proportions and standard deviations are shown in Table 1 As expected, there was a main effect of evaluation, such that there was a significan t difference between the proportion of positive traits participants attributed to themselves ( M = .61 SD = .15) and the proportion of negative traits participant s attributed to themselves ( M = .24 SD = .18) F (1, 58) = 142.78, p < .001, = .71 There was no main effect of community, F (1, 58) = 1.57, p = .22 = 03 and no interaction of response and community, F (1, 58) = .271, p = .27 = 02 To further assess social desirability, a 2 ( response: yes / no ) x 2 (community : school/local cit y) mixed ANOVA with repeated measures on the first factor was calculated to determine if there was a relationship between the yes/no computer responses and community on reaction time, regardless of traits. The mean response times and standard deviations are shown in Table 2 As predicted, t here was a main effect of response, such that there was significant difference between yes ( M = 1419.77, SD = 310.25) and no ( M = 1571.78, SD = 396.21) response reaction times, F (1, 58) = 19.32, p < .001, = .25 There was no main effect of community, F (1, 58) = 1.72 p = .69 =

PAGE 48

41 03 and no interaction of response and community, F (1, 58) = .16 p = .19 = .003 Participants were faster at responding yes than responding no regardless of community. Self De scriptive Errors H $' posited that participants should make more self descriptive errors on mismatching traits than matching traits across community Self errors were defined as a disagreement between the dichotomous self descriptive computer response (yes/no) and the coded dichotomous self descriptive survey response (yes/no) A 2 ( self error: match/mismatch ) x 2 (community : school/local city ) mixed ANOVA with repeated measures on the first factor was calculated, with the proportion of self errors par ticipants made on either matching or mismatching traits as the dependent variable Mean proportions and standard deviations are shown in Table 3. As expected, t here was a main effect of self error such that there were significant ly more self errors on mismatching ( M = .10 SD = .09) traits than on matching ( M = .07 SD = .06 ) traits F (1, 58) = 6.56 p = .01 = .10 There was no main effect of community, F (1, 58) = 1.17 p = .28 = 02 and the interaction betwee n community and trait was also insignificant F (1, 58) = 2.62 p = .11 = .04 Community Connectedness Concerning H $( )! an independent sample t test was used to evaluate differences between community connectedness scores from the ICS The school community scores ( M = 3.73, SD = .94 ) and the local outside community scores ( M = 2.83, SD = .99 ) were compared, and the analysis indicated a significant difference between the scores, t (59) = 3.6 1, p = .001. Participants were more connected to their school community than th eir local city community. Additionally, a one tailed Pearson product moment correlation

PAGE 49

42 analysis determined a significant positive correlation between school and local city community connectedness scores, r (58) = .77, p < .001. Discussion The objective of this study was to investigate overlapping mental representations of self and community. It was hypothesized that participants would respond faster to matching traits than mismatching traits. Participants were also expected to attribute more positive tra its to themselves than negative ones, and to exhibit more self errors on mismatching traits than on matching traits. Lastly, it was hypothesized that participants would be more connected to the school community than the local/city community. All of the hyp otheses were supported. The se results complemented previous research that investigated the relationship between the self with a close other or ingroup (Aron et al., 1991; Smith & Henry, 1996; Smith et al., 1999). The current study utilized a modified versi on of the previous methods and found similar significant results for the relationship between the self and community. Inclusion of Community in the Self The first hypothesis that assessed whether there was a significant difference between matching and mi smatching trait reaction time responses was confirmed: participants were quicker to respond to traits that matched between themselves and their community than to the traits that mismatched. The interaction between trait and community indicated that this ef fect was only significant in the school community conditio n. Participants responded significantly faster to matching traits than mismatching traits in the school community condition, whereas there was no difference between matching and mismatching reaction times in the local/city community condition. T hese

PAGE 50

43 reaction time results complement previous studies that found evidence for the overlap of self with other and ingroup representations (Aron et al., 1991; Smith & Henry, 1996; Smith et al., 1999). An additional analysis on the proportion of matching and mismatching traits indicated that the proportion of reported matching traits in the school community was greater than the proportion of matching traits in the local/city community T he proportion of rep orted mismatching traits was smaller in the school community than the proportion of mismatching traits in the local/city community. Though previous studies did not perform this analysis, it offered an additional perspective on the primary hypotheses and co nceptually supported the main results. Concerning trait evaluation ( H $& ) participants attributed more positive traits to themselves than negative traits, which accounted for social desirability and supported the second hypothesis This finding is in lin e with previous research (Aron et al., 1991; Smith & Henry, 1996; Smith et al., 1999). Participants were also faster responding yes than responding no on the self descriptive reaction time task, which also affirms previous findings (Kuiper, 1981; Mueller e t al., 1986; Aron et al., 1991; Smith & Henry, 1996). Responding yes to a self descriptive trait may yield a quicker response than no because these traits might be more salient, self relevant, and easily accessible for participants, rather than a trait tha t is distinctive from the self. For the error analysis ( H $' ), it was hypothesized that pa rticipants should make more self descriptive errors between the computer and survey self descriptive tasks, on mismatching traits than matching traits between self and community. As expected, participants made more self errors on mismatching traits than matching traits. This result

PAGE 51

44 was also found in previous studies using similar methodologies (Aron et al., 1991; Smith & Henry, 1996). B ecause the self a nd community are closely cognitively integrated, the mismatching traits may lead to self descriptive confusions. Participants were more likely to make self errors on mismatching traits because the self descriptive confusions may have delayed reaction time response; with the pressure of having to respond as quickly as possible, a self descriptive error was more likely to occur. The last hypothesis ( H $( ) determined whether participants felt differently about their school community and local outside communit y, in terms of connectedness and relation to their selves. There was a significant difference betw een the two community condition ICS scores, such that participants reported higher ICS scores in the school community condition, and lower ICS scores in the l ocal/city community condition. T he mean school community connectedness score was just above the midpoint in the scale, and the mean local/city community connectedness score was just below the midpoint ; participants viewed themselves as more a part of their school community This finding relates to Lounsbury and DeNeui's (1996) study on the relationship between PSOC and college campuses settings. Students who reported living on campus in the small school group also reported greater levels of PSOC. Because 90 % of the participants in the current study reported living on campus, they probably had a higher PSOC and thus felt more connected to their school community. The collective trait and community connectedness analyses converge on the same fundamental idea: the inclusion of community in the self. Participants are more connected to the school community than their local/city community. Therefore, participants report ed a greater proportion of matching traits and a smaller proportion of mismatching traits in

PAGE 52

45 the school community than in the local/city community ; this shows there was significantly more congruence or overlap between participants' self concept and the sch ool community they felt connected to Additionally, participants responded significantly quicker to matching traits than mismatching trait s only in the school community. Self relevant traits were more salient and easily accessible for participants, so they responded faster. When these traits matched with the community that participants felt most connected to (school community), these matching traits yielded facilitation of reaction time responses, as predicted. In sum, these results point to the overlap of self and community mental representations. This study contributes to the growing research of the complex self concept, and the connections between the self and other individuals or groups. The Connectionist Perspective Smith et al. (1999) argued that a connectionist model is a succinct and promising explanation for the overlap of self and other mental representations. Smith (1996) discussed connectionist models in relation to social psychology, primarily focusing on the distinction between symbolic and c onnectionist models. According to Smith (1996), almost all social psychological models are symbolically based on their theoretical assumptions they consist of concrete representations of ideas. Symbolic models assume the ideas of representation construct ion (cognitive representations form from simpler concepts) discrete representations, representation process distinction (processes are distinct from the content) and rule governed processes (Smith, 1996) The mind is conceptualized as exclusively processing symbols but connectionist models rely on a very different framework. Fixed symbolic models that explain the relationship between

PAGE 53

46 the self and other mental representations are inappropriate because the self i s complex and context sensitive (Smit h et al., 1999). Connectionist models appropriately characterize the flexible nature of mental representations. Basic components of a connectionist model are its units, links, and activation (Smith, 1996) Units, which perform simple tasks, are all intertwined and contain several weighted links between one an other. Connectionist models assume that a mental representation is not a symbol, but rather a pattern of activation that is formed between connections of simple units. This distributed representation accounts for prototype extraction. When similar instances are thought of as a series of connections, the central similarities between the instances should be highlighted and reinforced, thus creating a prot otypical representation. Because connectionist models focus on the overall activation pattern as creating a mental representation as opposed to static symbols, Smith (1996) argues that these models are more realistic because people are anything but static unchanging beings. Smith et al. (1999) developed a simple connectionist model as a possible way to explain the overlap of self and other mental representations. The Interactive Activation and Competition (IAC) model, proposed by McClelland and Rumelhart (1981), is a localist connectionist model of memory and retrieval. Concepts are represented with units, and the activation between different units varies between 0.2 to +1.0. Smith et al. (1999) used this model, for simplicity and ease of interpretation, as a means to explain the overlap of self and other mental representations. In their model, units represented self and partner (other), and were connected to several units representing traits (see Figure 5). It is assumed that this network is created throu gh associative processes, namely relating

PAGE 54

47 the self to certain traits and the partner to some of the same traits. According to the model, the activation for trait 1 is the greatest (.90) because it's associated both with the self and partner unit ; this trai t is expected to elicit a shorter response time than trait 2. The network predicted the pattern of results in the study, such that the closer the activation level between the Self and trait nodes was to the endpoints of the scale ( 0.2 to +1.0), the faster the response was than if the activation was closer to the midpoint of the scale. Though this proposed connectionist model was clearly in it s beginning stages, it offers a possible explanation for the results from the current study Smith et al.'s (1999) model was appropriate for the relationship between self and close other, and because the data from the present study were similar to their study, the connectionist model may also be appropriate for the relationship between self and community. B ecause conn ectionist networks account for context sensitivity, they may be a promis ing way of further studying the complex self concept Limitations and Implications The central limitation to this study was the participant distribution. Previous research has indicated housing status as significantly related to PSOC on college campuses, such that students who live on campus report greater levels of PSOC (Lounsbury & DeNeui, 1996). Ninety percent of participants in the current study reported living on campus, an d b ecause the college that participants attended is mostly residential and the sampling method did not specify the need for off campus participants, the comparison between the school and local city communities might not have been as reliable considering 90 % of the participants reported living on campus. This also limited the external validity of the study and the possible comparisons of ICS scores between on

PAGE 55

48 and off campus participants. Additionally, only 5% of participants reported growing up in the local/ city community. Of these participants only one was in the local/city condition. This participant indicated the highest community connectedness score on the scale (most overlapping circles). Lastly, b ecause community was a between subjects variable, it was not possible to compare how participants felt about both communities. Future research should address these issues and establish a more robust comparison between groups, such as comparing college students at a commuter college and a residential college, while controlling for possible confounds like motivation and needs for off campus living. Additional research could also incorporate personality measures, as they have been indicated to affect PSOC on college campuses (Lounsbury & DeNuei, 1996). If this st udy were replicated, a multiple regression analysis could be used to assess the data. This could allow for more control of the variables, as well as predicting if ICS scores moderate reaction times. The area of research distinguishing ingroup and community is scarce, which also necessitates further studies on empirically distinguishing between the two constructs. Ultimately, this research should be continued within a connectionist framework. The connectionist model posited by Smith et al. (1999) was in the beginning stages of development B ecause the self has been implicated with close others, ingroups, and community, these relationships necessitate an integrative model to explain the fundamental mental processes underlying the overlap of these various menta l representations. The self concept is extremely complex and context dependent. Research has shown the connection between the self with close other and ingroup, and the present study provided preliminary evidence for the overlapping mental representation s of self and

PAGE 56

49 community. Community connectedness elicits a feeling of belonging in the sense that I' becomes us ,' and this connection serves as a solid foundation for future research in community psychology and ultimately in what defines the self

PAGE 57

50 Refe rences Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as including other in the self. Journal of Personality and Social Psychology, 60 (2), 241. Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of other in the self scale a nd the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63 (4), 596 612. Aron, A., Mashek, D., McLaughlin Volpe, T., Wright, S., Lewandowski, G., & Aron, E. N. (2005). Including close others in the cognitive structure of t he self. In M. W. Baldwin, & M. W. Baldwin (Eds.), Interpersonal cognition. (pp. 206 232). New York, NY : Guilford Press. Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, 17 475 482. Brown, R., Condor, S., Mathews, A., Wade, G., & Williams, J. (1986). Explaining intergroup differentiation in an industrial organization. Journal of Occupational Psychology, 59 273 286. Cameron, J. (2004). A three factor model of social ide ntity. Self and Identity, 3 239 262. Cicognani, E., Pirini, C., Keyes, C., Joshanloo, M., Rostami, R., & Nosratabadi, M. (2008). Social participation, sense of community and social well being: A study on American, Italtian, and Iranian university students. Social Indicators Research, 89 97 112.

PAGE 58

51 Coats, S., Smith, E. R., Claypool, H. M., & Banner, M. J. (2000). Overlapping mental representations of self and in group: Reaction time evidence and its relationship with explicit measures of group identif ication. Journal of Experimental Social Psychology, 36 304 315. Greenwald, A. G., & Pratkanis, A. R. (1984). The self. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social cognition (Vol. 3, pp. 129 178). Hillsdale, NJ: Erlbaum. James, William (1890). The Consciousness of Self. Psychology of the Self. Retrieved from http://psychclassics.yorku.ca/James/Principles/prin10.htm Kuiper, N. A. (1981). Convergent evidence for the self as a prototype: The 'inverted U RT effect' for self and other judgments. Per sonality and Social Psychology Bulletin, 7 (3), 438 443. Lounsbury, J. W., & DeNeui, D. (1996). Collegiate psychological sense of community in relation to size of college/university and extroversion. Journal of Community Psychology, 24 (4), 381 394. Mashek, D., Cannaday, L. W., & Tangney, J. P. (2007). Inclusion of community in self scale: A single item pictorial measure of community connectedness. Journal of Community Psychology, 35 (2), 257 275. McClelland, J. L., & Rumelhart, D. E. (1981). An interactive activation model of context effects in letter perception: Part 1. An account of basic findings. Psychological Review, 88, 375 407. McMillan, D. W., & Chavis, D. M. (1986). Sense of community: A definition and theory. Journal of Community Psychology, 14 (1), 6 23.

PAGE 59

52 Meuller, J. H., Thompson, W. B., & Dugan, K. (1986). Trait distinctiveness and accessibility in the self schema. Personality and Social Psychology Bulletin, 12 81 89. Obst, P., Smith, S. G., & Zinkiewicz, L. (2002). An exploration of sense of community, part 3: Dimensions and predictors of psychological sense of community in geographical communities. Journal of Community Psychology, 30 (1), 119 133. Obst, P. L., & White, K. M. (2005). An exploration of the interplay between psychological sense of community, social identification and salience. Journal of Community & Applied Social Psychology, 15 (2), 127 135. Obst, P. L., & White, K. M. (2007). Choosing to belong: The influence of choice on social identification and psychological sense of communi ty. Journal of Community Psychology, 35 (1), 77 90. Prezza, M., & Costantini, S. (1998). Sense of community and life satisfaction: Investigation in three different territorial contexts. Journal of Community & Applied Social Psychology, 8 (3), 181 194. Smit h, E. R. (1996). What do connectionism and social psychology offer each other? Journal of Personality and Social Psychology, 70 (5), 893 912. Smith, E. R., Coats, S., & Walling, D. (1999). Overlapping mental representations of self, in group, and partner: Further response time evidence and a connectionist model. Personality and Social Psychology Bulletin, 25 (7), 873 882. Smith, E. R., & Henry, S. (1996). An in group becomes part of the self: Response time evidence. Personality and Social Psychology Bulleti n, 22 (6), 635 642. Superlab (Version 4.0) [Computer software]. San Pedro, CA: Cedrus.

PAGE 60

53 Tropp, L. R., & Wright, S. C. (2001). Ingroup identification as the inclusion of ingroup in the self. Personality and Social Psychology Bulletin, 27 (5), 585 600. Turner J. C. (1986). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.). Social identity and intergoup relations (pp. 15 40) Cambridge: Cambridge University Press.

PAGE 61

54 Table 1 Mean (Standard Deviatio n) Positive and Negative Trait Evaluation Proportions Trait Community Positive Negative School .643 ( .16 ) .239 (.17) Local Outside .572 ( .14 ) .237 ( .19 )

PAGE 62

55 Table 2 Mean (Standard Deviation ) Computer Task Responses (ms) Response Community Yes No School 1370.40 ( 310.25 ) 1508.67 ( 376.20 ) Local Outside 1469.13 ( 319.20 ) 1634.90 ( 411.83 )

PAGE 63

56 Table 3 Mean (Standard Deviation) Matching and Mismatching Self Error Proportions Self Error Community Match Mismatch School 070 ( .05 ) .120 ( .10 ) Local Outside .073 ( .06 ) .085 ( .07 )

PAGE 64

57 Figure 1 The Inclusion of Other in Self (IOS) scale Adapted from Inclusion of other in the self scale and the struc ture of interpersonal closeness, by A. Aron, E. N., Aron, and D. Smollan, 1992, Journal of Personality and Social Psychology, 63 p. 597. Copyright 1992 by the American Psychological Association.

PAGE 65

58 Figure 2 The Inclusion of Ingroup in Self (IIS) scale Participants were instructed to circle the pair of circles that best represented their own level of identification with their g roup. Adapte d from Ingroup identification as the i nclusion of ingroup in the self," by L. R. Tropp, and S. C. Wright, 2001, Personality and Social Psychology Bulletin, 27 p. 587. Copyright 2001 by the Ameri can Psychological Association.

PAGE 66

59 Figure 3 The Inclusion of Community in Self (ICS) scale Adapted from Inclusion of community in self scale: A single item pictorial measure of community connectedness ," by D. Mashek, L. W. Cannaday, and J. P. Tangney, J ournal of Community Psychology 35 p. 263. Copyright 2007 by the American Psychological Association.

PAGE 67

60 Figure 4 *$$! +$$! %$$$! %&$$! %($$! %*$$! %+$$! &$$$! ,-.//0! 1/-2034567! !"#$%&'()*&+"),+-.) /'++0(&%1) 2"#()!"#$%&'()*&+")!"-3'(-"-)4$5'--)/'++0(&%1) 826-.! 859:26-.! *5#&% )

PAGE 68

61 Figure 5 Interactive activation and competition (IAC) model of self, other, and traits Solid lines represent positive links while dotted lines represent negative links. Adapted from Overlapping mental representations of self, in group, and partner: Further response time evidence and a connectionist model ," by E. R. Smith, S. Coats, and D. Walling, Personality and Social Psychology Bulletin, 25 p. 879. Copyright 1999 by the American Psychological Association.

PAGE 69

62 Appendix Please do NOT put your name anywhere on the survey. This survey investigates how New College of Florida students view the ir social relationships with the community. Your voluntary participation is greatly appreciated. Please respond to as many items as possible your responses will be anonymous. The survey contains 275 items and is 5 pages long. If you wish to stop at any time you may do so and your survey will not be included in the analysis. When you are finished with the survey, seal it in the blank envelope provided. Please do NOT write your name on the survey or the envelope. Thank you for your participation!

PAGE 70

63 INSTRUCTIONS : Please rate the Sarasota/Bradenton Community (SBC) on each trait adjective using the 1 7 scale. Place your answer in the space next to each item. The Sarasota/Bradenton community includes local events of f campus, non campus students, off campus jobs, and all of the local people and activities not affiliated with the college. Whereas not all members of any community are exactly alike, community members tend to be similar on many traits however, you shoul d provide your personal opinion about the general characteristics of the community. 1 2 3 4 5 6 7 extremely somewhat extremely unlike SBC like SBC like SBC 1. ______ shy 2. ______ spiteful 3. ______ active 4. ______ crude 5. ______ persuasive 6. ______ tolerant 7. ______ superficial 8. ______ cowardly 9. ______ amusing 10. ______ timid 11. ______ sympathetic 12. ______ sarcastic 13. ______ sensible 14. ______ creative 15. ______ frank 16. ______ cheerful 17. ______ versatile 18. ______ serious 19. ______ trusting 20. ______ boastful 21. ______ generous 22. ______ hot headed 23. ______ spendthrift 24. ______ aggressive 25. ______ mature 26. ______ blunt 27. ______ skeptical 28. ______ independent 29. ______ scornful 30. ______ submissive 31. ______ good tempered 32. ______ opportunist 33. ______ inventive 34. ______ considerate 35. ______ envious 36. ______ deceptive

PAGE 71

64 37. ______ emotional 38. ______ congenial 39. ______ weak 40. ______ cordial 41. ______ unpredictable 42. ______ attent ive 43. ______ ambitious 44. ______ alert 45. ______ responsible 46. ______ sophisticated 47. ______ dull 48. ______ proud 49. ______ irritable 50. ______ neat 51. ______ appreciative 52. ______ tactless 53. ______ worrier 54. ______ systematic 55. ______ suspicious 56. ______ persistent 57. ______ foolhardy 58. ______ tactful 59. ______ nosey 60. ______ unreliable 61. ______ self righteous 62. ______ observant 63. ______ vain 64. ______ lazy 65. ______ antisocial 66. ______ restless 67. ______ materialistic 68. ______ naive 69. ______ prompt 70. ______ productive 71. ______ unfair 72. ______ prejudiced 73. ______ choosy 74. _____ jealous 75. ______ domineering 76. ______ ill mannered 77. ______ forward 78. ______ self reliant 79. ______ objective 80. ______ cold 81. ______ showy 82. ______ wordy 83. ______ foolish 84. ______ self centered 85. ______ argumentative 86. ______ shrewd 87. ______ nonchalant 88. ______ irrational 89. ______ methodical 90. ______ tender

PAGE 72

65 INSTRUCTIONS : Please rate the New College Community (NCC) on each trait adjective using the 1 7 scale. Place your answer in the space next to each item. The New College community includes fellow students, professors, employees, activities on campus, and any events that are exclusive to the school that occur on c ampus. Whereas not all members of any community are exactly alike, community members tend to be similar on many traits however, you should provide your personal opinion about the general characteristics of the community. 1 2 3 4 5 6 7 extremely somewhat extremely unlike NCC like NCC like NCC 1. ______ spendthrift 2. ______ tactful 3. ______ productive 4. ______ nonchalant 5. ______ showy 6. ______ hot headed 7. ______ envious 8. ______ alert 9. ______ systematic 10. ______ spiteful 11. ______ tender 12. ______ suspicious 13. ______ ill mannered 14. ______ active 15. ______ mature 16. ______ proud 17. ______ prejudiced 18. ______ amusing 19. ______ inventive 20. __ ____ dull 21. ______ appreciative 22. ______ submissive 23. ______ ambitious 24. ______ self centered 25. ______ foolhardy 26. ______ unfair 27. ______ objective 28. ______ congenial 29. ______ choosy 30. ______ skeptical 31. ______ prompt 32. ______ deceptive 33. ______ considerate 34. ______ self righteous

PAGE 73

66 35. ______ worrier 36. ______ scornful 37. ______ antisocial 38. ______ lazy 39. ______ restless 40. ______ foolish 41. ______ persuasive 42. ______ weak 43. ______ attentive 44. ______ tactless 45. ______ good tempered 46. ______ emotional 47. ______ boastful 48. ______ irrational 49. ______ self reliant 50. ______ wo rdy 51. ______ timid 52. ______ naive 53. ______ tolerant 54. ______ trusting 55. ______ nosey 56. ______ shrewd 57. ______ sympathetic 58. ______ cheerful 59. ______ cordial 60. ______ shy 61. ______ persistent 62. ______ sarcastic 63. ______ methodical 64. ______ serious 65. ______ opportunist 66. ______ jealous 67. ______ unreliable 68. ______ cowardly 69. ______ irritable 70. ______ blunt 71. ______ sophisticated 72. ______ crude 73. ______ cold 74. ______ versatile 75. ______ observant 76. ______ unpredictable 77. ______ frank 78. ______ creative 79. ______ aggressive 80. ______ materialistic 81. ______ forward 82. ______ neat 83. ______ vain 84. ______ responsible 85. ______ domineering 86. ______ argumentative 87. ______ independent 88. ______ superficial 89. ______ generous 90. ______ sensible

PAGE 74

67 INSTRUCTIONS : Please rate YOURSELF on each trait adjective using the 1 7 scale. Place your answer in the space next to each item. 1 2 3 4 5 6 7 extremely somewhat extremely unlike like like 1. ______ skeptical 2. ______ creative 3. ______ showy 4. ______ alert 5. ______ cheerful 6. ______ systematic 7. ______ crude 8. ______ frank 9. ______ cordial 10. ______ self righteous 11. ______ tactful 12. ______ proud 13. ______ naive 14. ______ cowardly 15. ______ lazy 16. ______ tolerant 17. ______ objective 18. ______ materialistic 19. ______ tactless 20. ______ vain 21. ______ methodical 22. ______ amusing 23. ______ nonchalant 24. ______ considerate 25. ______ mature 26. ______ self centered 27. ______ wordy 28. ______ restless 29. ______ ill mannered 30. ______ sarcastic 31. ______ unfair 32. ______ unreliable 33. ______ shy 34. ______ versatile 35. ______ nosey 36. ______ independent 37. ______ deceptive 38. ______ worrier 39. ______ opportunist 40. ______ envious 41. ______ good tempered 42. ______ suspicious

PAGE 75

68 43. ______ appreciative 44. ______ hot headed 45. ______ submissive 46. ______ persistent 47. ______ cold 48. ______ irrational 49. ______ attentive 50. ______ spendthrift 51. ______ neat 52. ______ observant 53. ______ blunt 54. ______ choosy 55. ______ spiteful 56. ______ generous 57. ______ sympathetic 58. ______ emotional 59. ______ productive 60. ______ prejudiced 61. ______ scornful 62. ______ self reliant 63. ______ weak 64. ______ congenial 65. ______ responsible 66. ______ foolish 67. ______ irritable 68. ______ antisocial 69. ______ argumentative 70. ______ sophisticated 71. ______ domineering 72. ______ forward 73. ______ timid 74. ______ foolhardy 75. ______ unpredictable 76. ______ boastful 77. ______ ambitious 78. ______ active 79. ______ prompt 80. ______ sensible 81. ______ tender 82. ______ inventive 83. ______ shrewd 84. ______ jealous 85. ______ persuasive 86. ______ dull 87. ______ superficial 88. ______ serious 89. ______ trusting 90. ______ aggressive

PAGE 76

69 INSTRUCTIONS : Please rate Bill Cosby on each trait adjective using the 1 7 scale. Place your answer in the space next to each item. 1 2 3 4 5 6 7 extremely somewhat extremely unlike Bill like B ill like Bill Cosby Cosby Cosby 1. ______ tactful 2. ______ sensible 3. ______ deceptive 4. ______ unpredictable 5. ______ showy 6. ______ spendthrift 7. ______ cordial 8. ______ worrier 9. ______ nosey 10. ______ forward 11. ______ unreliable 12. ______ scornful 13. ______ boastful 14. ______ amusing 15. ______ irrational 16. ______ hot headed 17. ______ persistent 18. ______ blunt 19. ______ cheerful 20. ______ opportunist 21. ______ trusting 22. ______ spiteful 23. ______ cowardly 24. ______ sarcastic 25. ______ na ive 26. ______ materialistic 27. ______ methodical 28. ______ shy 29. ______ dull 30. ______ proud 31. ______ sophisticated 32. ______ skeptical 33. ______ mature 34. ______ productive 35. ______ observant 36. ______ foolhardy 37. ______ wordy 38. ______ attentive 39. ______ independent 40. ______ objective

PAGE 77

70 41. ______ responsible 42. ______ tactless 43. ______ envious 44. ______ timid 45. ______ submissive 46. ______ suspicious 47. ______ sympathetic 48. ______ generous 49. ______ prejudiced 50. ______ active 51. ______ prompt 52. ______ serious 53. ______ self righteous 54. ______ congenial 55. ______ tender 56. ______ ambitious 57. ______ creative 58. ______ lazy 59. ______ antisocial 60. ______ superficial 61. ______ vain 62. ______ irritable 63. ______ nonchalant 64. ______ ill mannered 65. ______ crude 66. ______ cold 67. ______ systematic 68. ______ foolish 69. ______ neat 70. ______ restless 71. ______ weak 72. ______ jealous 73. _____ emotional 74. ______ argumentative 75. ______ unfair 76. ______ self centered 77. ______ inventive 78. ______ choosy 79. ______ tolerant 80. ______ alert 81. ______ versatile 82. ______ self reliant 83. ______ appreciative 84. ______ frank 85. ______ domineering 86. ______ considerate 87. ______ good tempered 88. ______ aggressive 89. ______ shrewd 90. ______ persuasive

PAGE 78

71 INSTRUCTIONS : Please circle the picture that best describes your relationship with the New College Community. This community includes fellow students, professors, New College employees, activities on campus, and any events that are exclusive to New College that occur on campus.

PAGE 79

72 INSTRUCTIONS : Please circle the picture that best describes your relationship with the Sarasota/Bradenton community. This includes local events off campus, non NCF students, off campus jobs, and all of the local people and activities not affiliated with New College.

PAGE 80

73 For the last questions, please write in or circle the answer for each question. 2. How old are you? __________ 3. What is your sex? Male Female 4. What has been your housing situation for most of the current school year (2011/2012 )? On campus Off campus 5. Did you grow up in the Sarasota/Bradenton area? Yes No Thank you!