Political leaders in the SYMLOG space: Perceptions of right and left wing leaders by right and left wing constituencies

Political leaders in the SYMLOG space: Perceptions of right and left wing leaders by right and left wing constituencies

POLITICAL LEADERS IN THE SYMLOG SPACE: PERCEPTIONS OF RIGHT AND LEFT WING LEADERS BY RIGHT AND LEFT WING CONSTITUENCIES Shmuel Ellis* Arie Nadler Amos...

2MB Sizes 0 Downloads 34 Views

POLITICAL LEADERS IN THE SYMLOG SPACE: PERCEPTIONS OF RIGHT AND LEFT WING LEADERS BY RIGHT AND LEFT WING CONSTITUENCIES Shmuel Ellis* Arie Nadler Amos Rabin Tel Aviv University

Systematic Multiple Level Observation of Groups (SYMLOG) was applied to studying images of political leaders. Three left wing and three right wing Israeli leaders were evaluated by left and right wing voters along the SYMLOG’s three dimensions-friendliness, task orientation, and dominance. It was found that right wing voters rated right wing leaders higher on friendliness and task orientation than left wing leaders. In a similar fashion, left wing voters viewed left wing leaders as more friendly and task-oriented than right wing leaders. Leaders of the right were viewed as more dominant than leaders of the left by left and right wing voters alike. Right and left wing voters assigned different SYMLOG types to right and left wing leaders. The dimension that distinguished between the various images was the dimension of friendliness. Whereas for right wing voters, all images of right wing leaders included a component of friendliness, none of these three images included an accented component of friendliness when rated by left wing voters. In a similar fashion, left wing voters saw two out of the three left wing leaders as including a component of friendliness, whereas none of the three left wing images included this component for right wing voters. Respondents from the right and the left rated their “ideal leader” as reflecting values of “dominance,” “friendliness,” and “task-orientation.” However, respondents saw leaders affiliated with their own political camp as closer to their image of the ideal leader on friendliness and task-orientation than leaders of the opposite political camp. Both right and left wing voters saw all political leaders as equally similar to their image of “ideal leader” on the “dominance” dimension regardless of that leader’s political affiliation.

Since their emergence more than two decades ago, the concepts “transformational” or “visionary” leadership (Bass, 1985; Bums, * Direct all correspondence to: Shmuel Ellis, Recanati Graduate University, Ramat Aviv, Tel Aviv 69978, Israel. Leadership Quarterly, 7(4), 507-526. Copyright 0 1996 by JAI Press Inc. All rights

of reproduction

ISSN: 1048.9843

in any form reserved.

of “charismatic,” 1978; Conger &

School of Business Administration,

Tel Aviv

508

LEADERSHIP

QUARTERLY

Vol. 7 No. 4 1996

Kanungo, 1987; House, 1977; Sashkin, 1988; Shamir, House, & Arthur, 1993) have yielded many theoretical and empirical works. House, Spangler and Woycke (1991) and Shamir, House and Arthur (1993), for example, reported at least 35 empirical studies which investigated the extraordinary effects of charismatic leaders on their followers and the social systems within which they function. Research in this field has been focused during the last 20 years on two issues: the effects of leaders on followers’ emotions, motivations and attitudes, and followers’ perceptions of the leader’s qualities and behaviors (Conger & Canungo, 1987; Shamir, 1994). Efforts have been made on both issues to “. . strip the aura of mysticism from charisma and to deal with it strictly as a behavioral science” (Conger & Kanungo, 1987, p. 639). While significant theoretical progress has been made with respect to the first issue (Bass, 1985; Bums, 1978; House, 1977; House, Spangler, & Woycke, 1991; Shamir, House, & Arthur, 1993), there is a deficiency in the systematic theorizing of the second issue. Much still needs to be done in order to figure out how leaders are perceived by their followers and how these perceptions affect leaders’ ability to harness their followers’ needs, motivation, values, and aspirations for the sake of the collective or the leaders’ own interest. Scholars of political science have devoted considerable attention to people’s perceptions of leaders and how these perceptions are generated (Conver, 198 1; Dow, 1969; Forgas & Kagan, 1977; Kinder, 1985; Marcus, 1961; Miller, Warrenberg, & Malanchuk, 1986; Pancer, 1989; Willner, 1984). Political scientists assume that people establish their attitudes toward the political leadership and form their voting preferences not only according to the ideological position of the leader, but also are influenced by the leader’s charisma (Shamir, 1992). More specifically, people follow him or her not only in order to accomplish their goals, but because of that extraordinary quality of the leader (Dow, 1969; Willner, 1984). It should be noted, however, that in spite of political scientists’ efforts to study the issue of perceived leadership, research in this domain has been often criticized for suffering from an atheoretical orientation which relies on a variety of dimensions that are derived empirically from an assortment of scales (Herman, 1986; Nygren & Jones, 1977). Among organizational scientists, after years of relatively slow theoretical development in the field of charismatic leadership qualities, Conger and Kanungo (1987) presented a behavioral framework for studying the concept of charismatic leadership. They argued that charisma should be viewed as an attribution made by followers on the basis of the observed behavior of the leader in the organizational context. Conger and Kanungo identified a set of 12 behaviors that comprise the database for the followers’ attributional inferences on the leader’s charisma. Unfortunately, although this constellation of behaviors provides important information on why people perceive their leaders as charismatic, it gives no information on their perceptions of the enduring personal orientations of the leaders, nor on the way in which these perceptions are related to followers’ willingness to follow them. Another group of researchers dealt with the effects of leaders’ motive profile (need for achievement, need for affiliation, and need for power) on effective leadership (e.g., see House, Spangler, & Woycke, 1991; McClelland & Boyatzis, 1982; Winter, 1973, 1978). However, these studies did not look at perceived leadership traits but at measured qualities. Further, they did not constitute a theoretical framework that incorporated leaders’ traits or motivations into the dynamic relations between leaders and their fohowers. Relations between attributed charisma and followers’ characteristics were first studied by Meindl (1990) and Shamir (1992, 1994). Meindl found, for example, a substantial

Political Leaders in the SYMLOG

Space

509

correlation between the strength of individuals’ beliefs about the significance of leadership and leaders and their perception of charisma in political and organizational leaders. Shamir found that followers attributed charisma even to a fictional leader. Shamir (1994) also found that voters’ ideological positions were strongly related to leaders’ perceived charisma. Furthermore, Shamir demonstrated that the combination of voters’ ideological position and the leader’s perceived charisma predicted voting preferences very accurately. These findings indicate that a different approach is needed to explain effects of perceived leadership on followers’ behavior. Such an approach should build on concepts and empirical findings that might increase our understanding of the dynamic relations between leaders and followers and how these relations affect followers’ perceptions of their leader. In the present study we suggest using the SYMLOG theory developed by Bales and his colleagues (Bales, 1970, 1985; Bales & Cohen, 1979; Hare, 1992; Weick, 1985). Since this theory describes interpersonal relations in terms of perceived images, it might be useful to analyze the dynamic mutual relations between leaders and followers and any other factor that might be relevant to understanding these relations. Now we shall describe the basic concepts of the theory and how it can enrich our understanding the charismatic leadership perception. SYMLOG:

BASIC CONCEPTS

In their SYMLOG theory Bales and Cohen (1979) suggested that interpersonal and within-group interactions can be most parsimoniuosly described by three interpersonal orientations which manifest themselves in group and interpersonal behaviors. The three dimensions are captured in value orientations and interpersonal behavior. They are: Upward vs. Downward orientation, Positive vs. Negative orientation, and Backward vs. Forward orientation (i.e., U-D, P-N, and F-B dimensions, respectively; see Figure 1). The individual’s position on each of these dimensions is expressed in a person’s behavior and in his or her more durable value-orientation. This conceptualization of group processes emerged from three decades of research by Bales and his colleagues on observation and classification of group and interpersonal interaction (Bales, 1950, 1970; Hare et al., 1965; Hare, 1976) and is rooted in meta-theoretical concepts from the psychoanalytic, socialpsychological, and sociological traditions (cf. Bales & Cohen, 1979, Chapters 1 and 2). The three dimensions refer to an individual’s image as “dominant,” “friendly” and “task oriented” as perceived by self or others. The “dominance” dimension is labeled the U-D dimension; it refers to dominant versus submissive behaviors. This dimension reflects values favoring material success and power vs. values favoring self-denial. The “friendliness” dimension, reflecting friendly versus unfriendly behaviors, is labeled the PN, and is expressed in values of equality and sharing versus egoism and self-interest. Task orientation is labeled the F-B dimension. It is reflected in task-oriented and role-prescribed behaviors versus emotional and out-of-role behaviors, and finds expression in values favoring “set ways of doing things” toward achieving the group’s goal, as opposed to values expressing lack of concern for the “prevailing rules,” and the group’s goals. The person’s position on each of these three dimensions can be either high, medium, or low. A high position reflects the ‘positive’ end of the continuum (i.e., U - Dominant, P Friendly and F - orientation towards group task). A low position reflects the ‘negative’ side of that continuum (i.e., D - submissive, N - unfriendly and B - moving backwards from group’s task). A medium position reflects neutrality on that dimension (i.e., the

LEADERSHIP

510

QUARTERLY

Vol. 7 No. 4 1996

individual’s unaccented behavior or tendency on that dimension). Thus, for example, whereas one person’s orientation may be described as “friendly” (i.e., high on the P- N dimension), “dominant” (i.e., high on the U-D dimension), and “task oriented” (i.e., high on the F-B dimension), another may be characterized as “unfriendly” (i.e., low on the P-N “submissive” (i.e., low on the U-D dimension), and with no accented dimension), orientation toward either conforming or not conforming (i.e., medium, or neutral, on the FB dimension). The intersection of the three levels (i.e., high medium or low) of each of three dimensions, with each other results in 27 cells which together makes up the SYMLOG space. These cells represent, in effect, 27 distinct inte~ersonal o~entations which find expression in a group members’ actual behavior as seen in the group or interpersonal context; and the cells also reflect individuals’ basic value orientation in the social world. The placement of an individual into one of these cells is based on his/her which is filled out by individuals scores on an especially designed questionnaire, themselves or by observers.

UPB

UB N

-

N =Negative N = Untiendly

PB

NB i

4 DNB

/

~

/

DB

DPB

B

B = Backward B = Emotionally Expressive

Figure 1.

D = Downward D = Submissive

The SYMLOG Three-Dimensional

Space

Political Leaders in the SYMLOG

Space

511

In a group setting, the three-dimensional space, described above, makes up the individual’s perception of his or her \‘social field’ (Bales, 1985). The contents of this field are social stimuli that are the focus of the person’s attention, and his/her position within the field is based on the perceiver’s images and evaluations. The aggregate of all participants’ fields is the emerging pattern of the group’s SYMLOG space. It is a constellation of the group’s communicated images and is based on the members’ perceptions of each other. Finally, the theory itself and the ensuing conceptual and empirical developments have focused on group behavior, and the images in the SYMLOG space typically refer to other group members. Yet, the use of SYMLOG as a theoretical and methodological tool has not been confined to the study of small group behavior. SYMLOG has been also used to study managerial performance, personality assessment, and the measurement of different brands in the context of marketing research (see Bachman, 1988; Bales & Isenberg, 1982; Hogan, 1988; Jesuino, 1985; Polley, Hare, & Stone, 1988).

SYMLOG AND POLITI~L

ELDERSHIP

The present study has two aims. The first aim is to explore how SYMLGG, as a theoretical and methodologic~ tool, can shed light on the process by which voters form perceptions of political leaders. This is especially important in view of the deficiency of theory and empirical research on voters’ perceptions of enduring orientations of leaders, the role of the interpersonal dynamics between leaders and followers, and the way in which people’s perceptions are related to their political preferences. The second aim of this study is to demons~ate that the pola~zation-uni~~ation process within the SYMLOG field exists not only within the realm of group behavior, but can also be applied in the context of studying political leaders. Uni~cation and polarization are key SYMLOG concepts that were inspired by Newcomb’s (1953) congruity theory. This theory holds that individuals tend to react to each other in terms of the similarities and differences they see between themselves and others in orientation toward third person(s) or object(s) to which they are co-oriented. Consequently, group members who have favorable attitudes toward each other will tend to share similar opinions, and group members who share similar opinions will tend to like each other (cf. also Byme, 1971). Such a sub-group of similar images will be viewed as more cohesive than a sub-group whose members are perceived as holding conflicting opinions on issues that are relevant to the group. In SYMLOYG terms, the theory holds that images of group members that share similar opinions will tend to unify along the three SYMLOG dimensions, In the course of the unification process, similar images tend to be perceived as more similar to each other on the three SYMLOC dimensions than they actually are. In contrast, the images of a sub-group whose members hold opinions that are opposite to those of the perceiver, will be located at the other extreme of the SYMLOG space. These images will be perceived as clustered together and viewed as more dissimilar on the three SYMLOG dimensions from the perceiver’s own “unified” group, than they actually are. This process is referred to as pol~zatjon. Unification and polarization are said to characterize perceptions and interaction in any group, and cause “good” (favorable) images to become “better,” and “‘bad” (unfavorable) ones to become “worse,” than they actually are. The end result of the polarizationuni~cation process is a pattern whereby group members’ perceptual and social interaction

512

LEADERSHIP

QUARTERLY

Vol. 7 No. 4 1996

fields are characterized by very small variance within the “favorable” and “unfavorable” sub-groups (i.e., ‘own’ and ‘opposing’ sub-groups, respectively), but large variance between the “favorable” and “unfavorable” sub-groups. Leadership is defined in SYMLOG terms as the ability to have consistent and significant intluence on the development and structure of the social interaction field and the power to shape the unification-polarization dynamics according to the leader’s own interests and goals (Bales & Cohen, 1979; Bales & Isenberg, 1982). Studying the perceptions of political leadership from this perspective has implications for research in the arena of political leadership, as well as for the research in the context of SYMLOG theory. Finally, advantages of the SYMLOG theory are that it is a parsimonious theory, that it has been validated in the study of political leaders (i.e., Isenberg & Ennis, 198 l), and that it has been shown to be consistent over time, culture, and situations (Bales & Isenberg, 1982; Polley, et al., 1988). The study was conducted in Israel in the summer of 1991, about one year prior to the 1992 national elections. Since voters’ identification with a particular political party has been identified by political scientists as the main factor affecting their perceptions of political leadership (Shikiar, 1974; Shikiar, Wiggins, & Fishbein, 1976), we expected right wing and left wing voters to place right and left wing leaders differently in the SYMLOG space. In line with Arian’s (1985) emphasis on the role of party identification in Israeli political life, we expected that identification with political party (as opposed to political ideology) would be an important predictor of individuals’ perceptions of their political leaders. In light of this, two hypotheses were formulated, the first of which reflects the polarization process, and the second, the process of unification: and right-wing-voters differ in evaluating Hypothesis 1. Left-wing-voters the two groups of leaders along the three SYMLOG dimensions.

each of

Hypothesis 2. The variance within the group of “good” images (i.e., those belonging to respondents’ own political ‘camp’), and the variance within the group of “bad” images (i.e., those belonging to the ‘opposite’ political camp) are relatively smaller than the variance between the two groups. An additional goal of the present study is to investigate images held by different political groups of the ideal leader. (In the language of SYMLOG theory, an ideal image is labeled the Wish image.) Bales (1986) argued that UPF types (i.e., people who are perceived as dominant - U, positive and friendly - P, and task oriented and forward moving - F) are seen as the best leaders across situations. This argument has been supported by several studies that applied the SYMLOG method to leadership research (Hogan, 1988; Kriger & Barnes, 1988). In the present study, we were interested in extending these findings by demonstrating that ideal political leaders would also be perceived as UPF (dominant, task oriented, friendly), and that according to polarization-unification processes, the “wish image” (i.e., the ideal leader) of a group would tend to unify with the images of the group’s actual leaders. Thus, the following hypotheses were formulated:

Political Leaders in the SYMLOG

Hy~th~~

3.

Domin~t,

513

Space

groups perceive the ideal leader as having the profile, UPF Positive (i.e., friendly) and Forward moving (i.e., task oriented).

Both

Hypothesis 4.

Left wing voters perceive left wing leaders as being more similar to their ideal leader than are right wing leaders, and right wing voters perceive right wing leaders as being more similar to their ideal leader than are left wing leaders.

METHOD Participants One hundred twenty-four individuals (56 women and 67 men, 1 unidentified) participated in the present study. Subsequent analyses revealed that there were no significant differences between men and women in their rating of political leaders on any of the three SYMLOG dimensions. Participants’ age ranged from 18 to 67 years, and the median age was 29. Respondents were first-line and middle management at a large food manufa~tu~ng organization in Israel. P~icipation was voluntary. Most p~icip~ts had a high school education, and a minority held an academic degree.

Measures Subjects were asked to participate in a study on political life in Israel, and most those approached agreed to take part in the study. Because some of the questions were sensitive (e.g., asking for voting preferences), respondents were assured of complete ~onymity. Respondents completed a questionnaire which consisted of two parts. Identification

with Political

Party

Three questions were designed to assess identification with the political right or the political left: (a) “If elections were conducted today, for which party would you vote?’ (b) “For which party did you vote in the last (1988) elections?’ (c) “In general, with which of the two groups of parties do you identify: the left-wing parties or the right-wing parties?’ Answers to present voting intentions (question 1) matched perfectly with answers to the question about affiliation with the political left or right (question 3). All those who intended to vote for a ‘right wing’ party affiliated themselves with the political right, and all those who intended to vote for a left wing party affiliated themselves with the political left. A nearly perfect match was obtained between current voting intentions (question 1) and past voting (question 2). Only 8.9% indicated that they would change their vote (about 10% from left to right, and 9% from right to left). Given this agreement between the questions, subjects’ voting intentions {answers to question 1) were used as the index of af~liation with the political right or political left. SYMLOG Social Values Rating Scales

Respondents were asked to rate six prominent Israeli political leaders on a 26-item SYMLOG value questionnaire. Three of these were prominent political figures who served as ministers in the then right-wing Israeli government (i.e., Ariel Sharon, David Levi, Itzhak Shamir). The other three were prominent leaders on the Israeli left, who at the time

514

LEADERSHIP

QUARTERLY

Vol. 7 No. 4 1996

were in the opposition (i.e., Yosi Sarid, Shimon Peres, and Itzhak Rabin). In addition, respondents were asked to rate two more images: the ideal leader and their self. In all, respondents completed eight identical SYMLOG value questionnaires-one for each image. The questionnaire used was a SYMLOG Individual and Social Values Rating Form (see Polley et al., 1988, p. 350). This questionnaire was translated from English into Hebrew by the second author; it had been used successfully in previous research in Israel. The 26 items tapped the 26 types of social values which make up the SYMLOG space (3 levels of U-D by 3 levels of P-N by 3 levels of F-B). For each of the 26 items, subjects were asked to rate on a three-point-scale whether a particular value statement applied to the image (O=very rarely; I=sometimes; 2=frequently). Cronbach’s alpha coefficient was computed separately for each of the six ends of the dimensions (i.e., for the items measuring P, N, D, U, F and B; each of these value orientations was assessed by 9 items), and for each of the 3 dimensions (18 items assessing the P-N, U-D, and F-B dimensions, respectively). Table 1 shows the alpha coefficients calculated for the present study and those reported by Bales and Cohen (1979). This procedure resulted in an 18-point scale for each of the three dimensions. Each scale ranged between the maximal ‘positive’ score of + 9.0 (i.e., representing maximum P, F, and U orientations) and the maximal ‘negative’ score of -9.0 (i.e., representing maximum N, B, and U orientations.) Analysis The research hypotheses stated above address the links between participants’ political af~liation and their perception of right wing and left wing leaders. Two-way split-plot ANOVAs were conducted. In each of the analyses the voter’s political affihation (with 2 levels, right wing voters vs. left wing voters) comprised the between-subjects independent variable, and their perceptions of the political leaders (with 2 levels, right vs. left wing

Table 1 Reliability Coeftkients of the Original and the Hebrew Versions of the SYMLOG Scales _ ~_____.. _ ..---_ -.. HebrewV<~r.vion 02,c$shVersion

--_.--

Norr:

*Reliability

(Prcwt~t -.

Study) -

(B&r nnd Cohen. 1979) ._______ _---

U

55

D

.67

.I4

P

.83

.95

N

.I5

.86

F

.80

.66

B

.ss

.6S

U-D

.lY”

.Y5*

P-N

.x1*

.so*

F-B

Al*

.-i4*

coefficlcnt~

.~rcponed by Bales and Cvhcn (19’?Y-) are after deletion of the 3 worct itcmh.

3 items were also deleted (itcms 7. IS and 24).

.-.

SC)

-_In the present study.

Political Leaders in the SYMLOG

515

Space

leaders) comprised the wi~in-subjects inde~ndent variable. ANOVAs were performed on ratings of each of the three SYMLGG dimensions. A regression model was used to partition the sum of squares. Each effect was adjusted for all other effects in the model. The model terms were tested against the pooled within-cells and residual sum of squares.

RESULTS The Polarization Hypothesis The first hypothesis was that, compared with left-wing voters, right wing voters would rate right-wing political leaders as higher on the SYMLOG dimensions (i.e., view them as relatively more Upward and dominant, Positive and friendly and more Forward moving and task-oriented). Conversely, it was predicted that left wing voters would rate left wing political leaders higher on the SYMLOG dimensions than would right wing voters. Two statistical analyses were performed. differences

in Percep~ians of the Six Po~i~jcai leaders

Simple F tests between right wing voters and left wing voters on each leader on each of the three SYMLOG dimensions were computed. Table 2 shows that each of the six political leaders was perceived differently by the two groups on at least two of the three SYMLOG dimensions. Perceptions of three of the six leaders differed along the U-D dimension; five were rated differently along the P-N dimension, and four were perceived differently on the F-B dimension. In general, the pattern of the means indicates that respondents tended to rate political leaders from their own “camp” as being friendlier, more domin~t, and more task-oriented than did voters from the opposing political affiliation. This pattern of results indicates that the three SYMLOG dimensions distinguished successfully between the two groups of voters. Perceptions

of Right Versus tefi Wing ieaders by Right and fefi Wing Voters on Each

of the Three SYMLOG Dimensions

To examine the links between an individual’s political affiliation and his or her perception of right wing and left wing leaders, collapsed across leaders in each of these two categories, two-way split-plot ANOVAs were conducted. In each analysis the voter’s political affiliation (with 2 levels, right wing voters vs. left wing voters) comprised the between-subjects independent variable, and their perceptions of the political leaders (with 2 levels-the average score for right vs. left wing leaders) comprised the within-subjects independent variable. ANOVAs were performed on ratings of each of the three SYMLOG dimensions. In these analyses, the average score of the three right wing leaders on each of the three SYMLOG dimensions and that of left wing leaders served as dependent measures. Ratings of leaders’

Dominance

(U-D Dimension)

A 2 (right vs. left wing respondents) X 2 (right revealed a significant main effect for the respondents’ 11.46, p < .OOl, and for the leaders’ affiliation, F respondents’ affiliation main effect indicates that left

vs. left wing leaders) ANOVA political affiliation, F (1, 122) = (1, 122) = 6.41, p c .05. The wing voters perceived political

LEADERSHIP

516

QUARTERLY

Vol. 7 No. 4 1996

Table 2 Perceptions Held by Left Wing and Right Wing Voters of Left and Right Wing Leaders: Mean Ratings, Standard Deviations, F Scores and Significance Levels Left Wing Vote-s MUUI

SD

Right Wing Voters Mron

SD

_.__._

F

I’< _-.__ _____..___

Left Wing Leaders Peres

Rabin

Sarid

U-D

2.86

3.41

2.20

3.77

I .03

0.3 12

P-N

1.17

4.43

-1.41

4.63

9.44

0.003

F-B

4.17

4.14

1.36

4.16

11.90

0.001

U-D

3.55

3.75

1.94

3.67

5.58

0.02

P-N

0.25

S.16

-0.42

4.76

0.5s

0.46

F-B

4.24

3.84

2.50

4.50

5.40

0.022

U-D

-0.44

3.00

0.46

3.60

1.73

0.19

P-N

4.96

4.18

1.09

6.90

13.24

0.000

F-B

2.24

4.10

0.43

3.90

5.83

0.017

U-D

-0.47

3.86

0.5 1

3.73

2.15

P-N

-2.66

4.42

3.36

4.80

55.67

0.000

Right Wing Leaders Shamir

Levi

Sharon

0.14

F-B

4.04

3.97

6.47

3.00

16.23

0.000

U-D

4.83

3.74

2.86

3.17

10.28

0.002

P-N

0.23

3.20

4.21

4.89

29.64

0.000

F-B

2.56

4.08

3.14

4.5 I

0.50

0.49

U-D

5.97

3.53

4.09

3.97

7.76

0.006

P-N

-4.50

4.60

2.19

1.69

S7.67

0.000

F-B

-1.39

4.69

2.63

3.97

25.71

0.000

leaders in general as more dominant than did right wing voters (means are 3.1 and 2.1, respectively). The leaders’ affiliation main effect indicates that right wing political leaders were seen as more dominant than left wing political leaders (means are 2.95 and 2.25, respectively). Ratings of Leaders’ Friendliness (P-N Dimension) A 2 (right vs. left wing respondents) X 2 (right vs left wing leaders) ANOVA revealed a

significant two way interaction, F (1, 122) = 53.53, p < .OOOl. Orthogonal comparisons revealed that this interaction effect arose because right wing respondents perceived right wing leaders to be friendlier than left wing leaders, F (1,122) = 25.9 1, p < .OOO 1, and left wing respondents perceived left wing leaders as friendlier than right wing leaders, F (1, 122) = 38.97, p < .OOOl. (F-B A 2 (right vs. left wing respondents) X 2 a significant two way interaction, F (1,120) performed on the relevant means showed leaders as more task-oriented and forward

Ratings of Leaders’ Task Orientation

Dimension)

(right vs. left wing leaders) ANOVA revealed = 46.1, p < .OOO1. The orthogonal comparisons that right wing respondents perceived their moving than left wing leaders, F (1, 122) =

Political Leaders in the SYMLOG

517

Space

34.45, p < .OOOl, and that left wing respondents perceived their leaders as more taskoriented and forward moving than the right wing leaders, F (1, 122) = 16.47, p < .OOO1. The main effects were not statistically significant. Taken together, these data indicate that respondents rated political leaders from their ‘own’ political camp as friendlier and forward moving, and that right wing political leaders were perceived as more dominant than left wing political leaders by both left wing and right wing voters. Finally, it seems that left wing respondents perceived political leaders as generally more dominant than did right wing voters.

The Polarization-Unification

Hypothesis

It was expected that both groups of voters would perceive more similarity between leaders in the same party than between the two groups of leaders (left wing vs. right wing leaders). More specifically, it was hypothesized that in the right wing group, the pooled variance between the three “good” images (Shamir, Levi, Sharon), and the variance between the three “bad’ images (Rabin, Peres, Sarid) would be smaller than the variance between the two groups of images. In the same way for left wing voters, the pooled variance between the three “bad’ images (Shamir, Levi, Sharon) and the variance between their three “good’ images (Rabin, Peres, Sarid) would be smaller than the variance between the two groups of images. Contrary to our expectation, this hypothesis was not supported. The data did not, in fact, reflect greater difference between the two groups of leaders than between the three leaders within each category (left wing leaders and right wing leaders). Image of the ‘Ideal Political leader’

The third hypothesis pertained to the perception of the ideal leader. To characterize the ideal SYMLOG image of the two groups of voters, a one-sample-t-test was conducted for each of the three SYMLOG dimensions across groups. The purpose of conducting these tests was to determine how significantly different from zero (the center of the three scales) were the left wing and right wing voters’ mean scores on each of the three ideal leader scales. Table 3 presents the mean ratings and the t values of the ideal leader of left wing and right wing voters along the three SYMLOG dimensions. The results demonstrate that voters, both left wing and right wing, share the same UPF ideal leader profile. The means of these three characteristics were significantly different from zero. Furthermore, the one-

Table 3 Mean Ratings, Standard Deviations and t Values of Ideal Leader L& Wing Voters Mem

Righi Wing Voters

SD

t(561

SD

t(66)

F-B

4.84

3.15

Io.66*

5.25

3.15

13.81*

5.02

3.14

17.92*

P-N

5.06

3.70

10.32*

4.68

3.91

9.75s

4.89

3.79

14.38*

U-D

2.48

2.93

6.36*

1.54

2.96

4.21*

2.0s

2.91

7.m*

Note:

*

,’

<

ml31

M&In

All Voter.7 MNUI

SD

t( 123)

LEADERSHIP

518

QUARTERLY

Vol. 7 No. 4 1996

way MANOVA as well as the three univariate F tests yielded a nonsignificant group effect. This indicates that the two groups held similar perceptions of the ideal leader. Ideal Leader and ‘Right’ and ‘Left’ Wing Political leaders The fourth hypothesis suggested that right wing voters would perceive right wing leaders as closer to the SYMLOG profile of the ideal leader than they would left wing leaders and that left wing voters would perceive left wing leaders as closer to their ratings of the ideal leader than they would right wing leaders. To test this hypothesis, difference scores between a respondent’s perception of ideal leader, as rated on the three SYMLOG dimensions, and his or her ratings of right and left wing leaders were computed. The means of respondents’ ratings of the ‘ideal leader,’ the right wing and the left wing leaders (collapsed across the 3 political leaders in each group), and the difference scores that are pertinent to hypothesis 4 (i.e., differences between ratings of ideal leader and right and left wing leaders) are displayed in Table 4. These analyses revealed a significant interaction for the ‘friendliness’ dimension (i.e., PN), F (1, 122) = 21.66, p < .OOOl, and the ‘task-orientation’ dimension (i.e., F-B) , F (I, 122) = 22.43, p < .OOOl. The analysis on difference scores of ‘perceived dominance’ (i.e., U-D) did not yield statistically significant effects. Orthogonal comparisons showed that in line with the predictions, compared to their perceptions of right wing leaders, left wing voters perceived left wing leaders as closer than right wing leaders to their image of the ideal leader on the dimension of friendliness (i.e., P-N dimension), F (1, 122) = 17.78, p <

Table 4 Mean Ratings and Standard Deviations of Leftists’ and Rightists’ Images, and Gap Scores Between Left Wing Leaders, Right WingLeaders and Ideal Leaders

Leftists

F-B

P-N U-D Rightists

F-B P-N

U-D

M

3.54

1.71

4.83

0.9 I

3.06

4.38

4.58

4.42

SD

2.88

3.14

3.15

4.89

2.34

3.33

3.31

3.48

M

I .92

-2.07

5.06

3.70

4.54

.24

3.63

6.50

SD

3.21

2.88

3.70

4.02

2.93

4.13

2.94

3.93

M

2.84

3.40

2.48

0.27

2.68

2.83

3.48

4.04

SD

2.27

2.15

2.93

3.72

2.27

2.46

3.38

3.16

M

1.54

4.16

5.25

2.37

4.32

2.88

3.86

3.23

SD

2.74

2.65

3.15

4.20

2.68

2.13

3.24

2.62

M

-0.16

2.99

4.68

4.30

6.07

4.49

5.64

3.97

SD

3.77

4.33

3.91

3.44

5.03

3.36

4.23

2.60

M

I .73

2.49

I.54

I.83

2.92

2.58

2.87

3.01

SD

2.24

2.13

2.96

3.1 I

2.33

1.84

2.17

2.13

Political Leaders in the SYMLOG Space

519

.OOOl (means are 4.54 and 7.24, respectively). Similarly, compared to their perceptions of left wing leaders, right wing voters perceived right wing leaders as closer to their image of the ideal leader on the dimension of ~endliness, F (1, 122) = 6.08, p < .Ol (means are 4.49 and 6.07, respectively). In the same way, left wing voters perceived their leaders as closer to their ideal leader on the task-orientation dimension (i.e., F-B dimension) than they did right wing leaders, F (1, 122) = 10.87, p c .Ol (means are 3.06 and 4.38, respectively), and right wing voters perceived the right wing leaders as closer to their image of the ideal leader than the left-wing leaders, F (1, 122) = 12.96, p -c .Ol (means are 2.88 and 4.32, respectively). The differences on the dimension of perceived dominance (i.e., U-D) were not statistically significant. Self Ratings on the SYMLOC Dimensions Although no specific hypotheses regarding differences between right wing and left wing voters’ self-ratings were made, the differences between right wing and left wing participants’ perceptions of their self image along the three SYMLOG dimensions aiso were examined (the relevant means and standard deviations appear in Table 4). The statistical analysis of the differences in ratings of self indicates that right wing voters rated themselves as more dominant than did left wing voters, F (1, 118) = 6.05 p -c .OS, (means are 1.83 and 0.27, respectively). No differences between the groups were observed on the dimension of friendliness (i.e., P-N; means were 3.7 and 3.4 for left and right wing voters, respectively). Although right wing voters tended to rate themselves as more forward moving and task-oriented than left wing voters (i.e., F-3; means were 2.37 and 0.9, respectively), this difference failed to exceed a conventional level of significance, F (1, 118) = 3.0, p < .lO. Thus, it seems that self-definitions of right and left wing voters are distinguishable on the U-R dimension in that right wing voters rate themselves as having a value orientation favoring dominance (e.g., striving for power and status) more than do left wing voters. It is of interest to note that only on the dimension of P-N (i.e., friendliness) were self-ratings of left wing voters significantly different from zero (i.e., the point of “neutmlity” in the SYMLOG space). They rated themselves as signi~c~tly friendly and caring in this value orientation, f (63) = 7.5, p < .Ol. Comparing the ratings of self by right wing voters to zero (i.e., the point of “neutrality”) revealed that they saw themselves as dominant, t (53) = 4.3, p < .Ol, friendly, t (53) = 9.45, p x .Ol, and task oriented, t (53) = 4.2, p < .Ol. Thus in the language of SYMLOG, whereas left wing voters see themselves as P types (characterized by high level of friendliness and caring), right wing voters see themselves as UPF types. Finally, we decided to examine whether the similarity between the self-rating and the rating of the leader can be used as an indicator for followers’ identi~cation with their leader. More specifically, we examined whether right wing voters perceived themselves as closer to the SYMLOG profile of the right wing leaders than did the left wing voters, and whether left wing voters perceived themselves as closer to the SYMLOG profile of the left wing leaders than did the right wing voters. To test this hypothesis, difference scores between respondents’ perception of themselves, as rated on the three SYMLOG dimensions, and their ratings of right and left wing leaders were computed. The means and standard deviations of the absolute difference scores are displayed in Table 4.

520

LEADERSHIP

QUARTERLY

Vol.

7 No. 4 1996

Only the analysis of the difference score of the friendliness dimension (P-N) yielded a significant interaction, F (1, 122) = 31.5, p < .OOOl. Scheffe aposteriori comparisons showed that compared to their perceptions of right wing leaders, left wing voters perceived left wing leaders as closer than right wing leaders to their self-image on the dimension of friendliness (p < .02). Similarly, compared to their perceptions of left wing leaders, right wing voters perceived right wing leaders as closer to their self-image on the dimension of friendliness (p -Z ,000 1). DISCUSSION In general, the findings support the predictions which were derived from the SYMLOG theory. Respondents viewed leaders of their own political camp more favorably on the PN (i.e., friendliness) and F-B (i.e., task orientation and forward movement) dimensions than leaders from the opposing political camp. More specifically, right wing voters evaluated right wing leaders as more friendly and more task oriented than they evaluated leaders affiliated with the left side of the political map. In a similar fashion, left wing voters viewed left wing leaders as more friendly and task oriented than they viewed right wing leaders. This finding provides partial support for the study’s first hypothesis, but perceptions on the U-D dimension (dominance) did not show this pattern. Rather, both left wing and right wing respondents viewed right wing leaders as more dominant than left wing leaders. This may be a result of the right wing’s political ideology and the verbal and nonverbal behavior of the right wing leaders. In matters involving external affairs, the political right in Israel, and perhaps elsewhere, puts an emphasis on the country’s military self-reliance and on a tough stance in negotiations. These values are reflected in the leaders’ public speeches and policy decisions. The political left, on the other hand, puts a premium on negotiated resolution of conflicts, and the necessity of compromise in order to achieve negotiated settlements. Thus, the greater perceived dominance of the right wing leaders found by this study may be attributed to these two different actual political emphases. A second explanation for the greater perceived dominance of the right wing leaders may be attributable to the fact that during the time of data collection, the right wing Likud-led coalition was in power. If this second explanation is valid, then data that would have been collected when the political left is in power would reveal that left wing leaders are perceived as more dominant than right wing leaders. If, however, the first explanation is valid, then regardless of which party is in power during data collection, right wing leaders will be perceived as more dominant than left wing leaders because of the inherent differences between a right wing and a left wing political ideology. Interestingly, Shamir (1994) conducted his study on the perceived charisma of the Israeli right and left wing leaders about the same time as the present study. Shamir found, for example, that voting preferences were strongly associated with ideological position, and that there is a positive correlation between charisma perceptions of leaders belonging to the same party and a negative correlation between charisma perceptions of leaders belonging to opposite parties. Although Shamir did not use the SYMLOG scales, the rationale for his hypotheses as well as and his findings can easily be explained according to the polarization and unification principles.

Political Leaders in the SYMLOC

Space

521

The differences between the two political groups also were demonstrated in the examination of the SYMLOG types associated with the different political leaders. A SYMLOG type is determined by examining the salient position of a particular image on each of the three dimensions. Thus, for example, an image which is significantly different from zero on the P-N dimension in the ‘Negative’ direction (i.e., unfriendly), is not significantly different from zero on the F-B dimension, and is significantly different from zero on the U-D dimension in the ‘Upward’ direction (i.e., high dominance), will be said to be a UN type (i.e., dominant and unfriendly, and ‘neutral’ on the task-orientation dimension). If another image is significantly different from zero in the ‘Downward’ (i.e., submissive), the ‘Positive’ (i.e., friendly), and the ‘Forward’ (i.e., task oriented) directions, it will be defined as a DPF type. When we examined the images of the political leaders in the present study using this procedure, differences between right and left wing voters were evident. Regarding right wing leaders, Itzhak Shamir (the then-prime minister) was viewed as PF and NF (i.e., friendly and task oriented vs. unfriendly and task oriented) by right wing and left wing respondents, respectively. Ariel Sharon (the then-minister of housing) was viewed as UPF and UNB (i.e., dominant, friendly, and task oriented vs. dominant, unfriendly, and moving back from the task) by right and left wing respondents, respectively. David Levi (the then-foreign minister) was viewed as UPF and UF (i.e., dominant, friendly, and forward moving vs. dominant and forward moving) by right and left wing respondents, respectively. Regarding left wing leaders, Shimon Peres was viewed as UPF and UNF (i.e., dominant, friendly, and forward moving vs. dominant, unfriendly, and forward moving) by left and right wing respondents. Yosi Sarid was viewed as UPF and U (i.e., dominant, friendly, and forward moving vs dominant) by left and right wing respondents, respectively. Itzhak Rabin’s image was seen as UF (i.e., dominant and forward moving) by both right and left wing respondents. Examination of these images reveals that, in line with the hypotheses, right and left wing respondent assigned different SYMLOG types to right and left wing leaders. Further, it seems that the dimension that distinguishes between the various images is P-N (friendliness). More specifically, whereas the images of right wing leaders included a component of friendliness (i.e., P) for all right wing voters, none of the three images of right wing leaders included an accented component of friendliness when rated by left wing voters. In a similar fashion, left wing voters saw two out of the three left wing leaders (i.e., Peres and Sarid) as including a component of P, but none of the three left wing images included a component of P for right wing voters. In further support of this, right wing voters saw the aggregate group of right wing leaders as a UPF image, whereas left wing voters saw the same group of right wing leaders as a UNF image. In a similar fashion, left wing voters saw the aggregate group of left wing leaders as a UPF image, whereas the same group of leaders was viewed by right wing voters as a UNF image. Thus, it seems that perceptions held by right and left wing voters were differentiated by the affective component viewing the image of the leader as friendly or unfriendly (i.e., the P-N dimension) rather than the instrumental component viewing the leader as forward or backward moving (i.e., the F-B dimension) or the dominance component (i.e., the U-D dimension). The difference in perceptions of the self-leader gap along the friendliness dimension gives further support to this conclusion.

522

LEADERSHIP

QUARTERLY

Vol. 7 No. 4 1996

It is of interest to note that the only political image which was perceived similarly by right and left wing voters was that of Itzhak Rabin who was elected prime minister about 1 year after the data were collected. Although the suggestion that ratings on a SYMLOG space are predictive of upcoming election results is speculative, the consensual perception of Rabin agrees with commentators who attributed Rabin’s 1992 electoral victory to his ability to appeal to voters on the political left and right. These results are in accord with Shamir’s (1994) findings regarding the perceived charisma of Rabin. Shamir found that there was no significant correlation between voters’ perception of Rabin’s charisma and their ideological position. What probably contributed to Rabin’s success in the Israeli politics was neither his charisma (he did not come out as the most charismatic leader in Shamir’s study) nor his SYMLOG profile (he was not rated as an ideal (UPF) leader). Rather, the two studies showed that Rabin was not clearly identified along the left wing-right wing dimension and that he was perceived similarly by voters belonging to the two opposing ends of the political continuum. One may argue that his ability to manage “business-like” relations without eliciting extreme emotions from people of either right or left wing orientations (P score close to zero) engendered the consensus around his leadership in Israel. (Izhak Rabin was assassinated by an Israeli citizen on November 4, 1995.) Both right and left wing voters saw leaders of their own bloc as dominant, friendly, and task oriented/forward moving individuals (i.e., UPF types). This finding supports the third hypothesis of this study-that regardless of their political affiliation, respondents would see the UPF type as their ‘ideal leader.’ A more direct test of this hypothesis showed that respondents from both the right and left rated their “ideal” leader as reflecting values of “dominance” (U), “friendliness” (P), and “task-orientation” (F). The hypothesis that respondents would see leaders affiliated with their own political camp as being closer to their image of the ideal leader than they would rate leaders of the opposite political camp was supported for the P-N and F-B dimensions (i.e., ‘friendliness’ and ‘task orientation’). Right wing voters saw right wing leaders as closer to their ratings of the ideal leader on the P and F dimensions than left wing leaders, and left wing voters saw left wing leaders as closer to their image of the ideal lead on these dimensions than right wing leaders. One may say that followers shape their perceptions of their leader’s charisma according to their schema of the ideal leader. No difference was observed on the “dominance” dimension. Both right and left wing voters saw apolitical leader as equally similar to their image of ideal leadership on the “dominance” (i.e., U) dimension regardless of that leader’s political affiliation. It seems that being defined as a leader is enough to generate the expectation that one expresses values favoring “dominance” (e.g., seeking power and success), and this is true irrespective of the specific political affiliation of that leader. It should be noted that the use of difference scores has been criticized for their unreliability (e.g., Cronbach & Gleser, 1953; Campbell & Stanley, 1963; Edwards, 1993). However, so far, the mere alternative for difference scores between two variable is the use of one variable as a covariate in ANOVA where the other variable is the dependent variable. In the present research, this could have distorted the meaning of the results, since the hypothesis that was examined pertained to the effect of political association on the difference between the image of the ideal leader and the self along the three SYMLOG dimensions. Furthermore, the problem with using difference scores is that they have a low reliability and therefore may have low validity. Obviously, this problem did not occur in the current research: The difference scores variable was valid (it confirmed the hypotheses.)

Political Leaders in the SYMLOC

Space

523

Finally, it is of interest to note that the self-ratings of left and right wing voters differed significantly from each other. Right wing voters rated themselves significantly higher on the dimension of dominance than did left wing voters. An examination of this valueorientation indicates that it agrees well with what might be expected of supporters of a right wing outlook in social life: “... strong, assertive, powerful, . , . superior, of high social status, rich, self-confident.. .” (Bales & Cohen, 1979, p. 355). Furthermore, the self-ratings of left wing voters were significantly different from zero only on the P dimension. Here again, an examination of this dimension suggests that putting a premium on the value of Positive orientation towards others agrees well with what we might expect to be the essence of a left wing political ideology. The P value orientation in society is described by Bales and Cohen (1979) as reflecting values supporting: ‘I.. . the vote of all elements in society, the ideal of equal opportunities for all antidiscrimination organizations, . . . . equal rights under the law, freedom of speech, freedom of press.. .” ( p. 367). These findings are important in that they lend support to the theory’s contention that the scales measure basic and enduring value orientations and not simply a preference for a particular kind of behavior in a specific social context. The present findings are relevant to research on political leadership and to research on SYMLOG. Regarding the research context of SYMLOG theory, the present results support the theoretical contention that individuals tend to differentiate the social field into ‘bad’ and ‘good’ images. Here, political affiliation served as the basis on which such a division occurred. In general, respondents saw the images of leaders affiliated with their own political tendencies as more favorable on the SYMLOG space than their perceptions of leaders from the other side of the political arena. This distinction was reflected in direct comparisons between the two groups of leaders and their distance from the image of the ‘ideal leader.’ The difference seems to have been most apparent on the dimension of ‘friendliness.’ In fact, although leaders of the opposite side may have been viewed as dominant, their image was consistently ‘unfriendly.’ This finding is congruent with earlier studies, which found the unification-polarization tendencies to be most prevalent on the P-N dimension (Polley et al., 1988). This suggests that respondents’ affiliation with political leadership has a strong affective flavor. Further, the study supports the theoretical suggestion that regardless of the specific context (e.g., business context, educational context, or political context), individuals see the ideal form of leadership as embodying value orientations of Dominance, Friendliness, and Forward movement towards group goals. In other words, regardless of their own political stand or value-orientations, individuals seem to see the UPF SYMLOG type as the ideal type of leader. The support for the study’s hypotheses should be viewed in light of the particular set of political circumstances which existed in the Israeli society when the data were collecteda few months after the Gulf War, when the Israeli society, which had undergone the traumatic experience of Scud missile attacks, was relatively united. Thus, even stronger results might have been obtained in times of greater political friction and division. It should be noted, however, that the prediction of significant variance between the two groups of leaders would be greater than the variance between the leaders within each group was not supported. An explanation for this lack of support is that the complexity of actual political life renders such an expectation unrealistic. In fact, in a politically dynamic society such as Israel, there are many differences between the various leaders within a particular political bloc. Thus, for example, on the left side of the political spectrum, Itzhak Rabin was viewed

524

LEADERSHIP

QUARTERLY

Vol. 7 No. 4

1996

as closer in his overall political convictions (e.g., on matters of security) to many political leaders on the right, than to Yosi Sarid, a prominent political leader on the left. Thus, although an overall distinction exists between the ‘positive’ and ‘negative’ images, some images seem to be “sitting on the fence” between the two political. In line with SYMLOG theory, such political leaders are in a position to ‘draw the two camps together’ and in electoral terms they stand the chance of drawing support from both sides of the political arena. These findings serve as an important source of validity for the use of SYMLOG theory in the context of political leadership. They lend support to the theoretical contention that individuals tend to polarize their social field into a group of ‘bad’ and ‘good’ images, and this is reflected in the images on the SYMLOG space. In the present study, this distinction was based on the political affiliation of two groups of individuals who were pre-selected according to their political affiliation. As such, the present findings provide an important source of “known-group” validity for the major hypotheses regarding unification and polarization in the social field. The present research findings demonstrate the potential value of the SYMLOG theory and methodology to the study of leadership in general and political leadership in particular. The findings clearly showed that: (a) perceptions of leaders are affected by the followers’ images of themselves, their political associations, and their image of the ideal leader. Understanding of voting behavior must take into account the multiple relevant images of the voter. (b) The three SYMLOG dimensions are on one hand a parsimonious description of the leader’s personality, but on the other hand very rich and informative. It should be noted that one may find some similarity between the three SYMLOG dimensions and the three leadership motives described by Winter (1978, 1987) and House, Spangler and Woycke (1991): power motive and dominance (UD), achievement motive and task orientation (FB), and affiliation motive and friendliness (PN). However, while the motive profile is based on measured scales (TAT), the SYMLOG scales are based on perceived values or behaviors of the leaders. More research is needed in order to elucidate the relations between charismatic leadership and the three SYMLOG dimensions. Studies correlating traditional scales of perceived charisma (Bass, 1985; Shamir, 1994), leader’s personality and motives profile with the three SYMLOG scales might advance the understanding of the charisma concept and the development of a better theoretical framework for the understanding of perceived leadership. Past research into political leadership has used a number of methodologies to study perceptions of leadership. For example, the method of multi-dimensional scaling (MDS), has often been used in this context (e.g., Nygren & Jones, 1977). In this and other similar methodologies, the investigator reproduces the relevant dimensions for the study of perceptions of leadership from the empirical data. These methods give rise to variations between studies in the perceptual dimensions-which are affected by factors such as time of measurement, the specific culture in which the study is being carried out, and the investigator’s particular research orientation and preferences. On the other hand, research conducted according to the SYMLOG approach is couched within a parsimonious theoretical orientation that is less affected by time and place, and that assures (a) uniformity of research procedures, and (b) consistency of perceptual dimensions being studies. This practice allows for studies that can compare perceptions between cultures and between sub-cultures within a particular society. In all, couching studies of political

Political Leaders in the SYMLOC

Space

525

leadership within SYMLOG theory is more suitable for various theoretically driven comparisons than axe studies which derive the perceptual dimensions, each time anew, from the available empirical data.

REFERENCES Arian, A. (1985). Politics and regime in Israel (in Hebrew). Tel Aviv: Zemora Bitan. Bachman, W. (1988). Nice guys finish first: A SYMLOG analysis of U.S. naval commands. In R.B. PoIley, A.P. Hare, & P.J. Stone (Eds.), The ~Y~~~G~~~cf~~~o~e~. New York: Praeger. Bales, R. F. (1950). interaction process analysis: A merhodfor ihe study ofsmal~ groups. Cambridge, MA: Addison-Wesley. Bales, R.F. (1970). Personality and interpersonal behavior. New York: Holt, Rinehart, & Winston. Bales, R.F. (1985). The new field theory in social psychology. International Journal of Small Group Research, I, I-18. Bales, R.F. (1986). SYMLOG and leadership theory. Paper presented at the SYMLOG Workshop, Center for Creative Leadership, North Carolina. Bales, R.F., & Cohen, S.P. (1979). SYMLOG: A system for multiple level observation of groups. New York: Free Press. Bales, R.F., & Isenberg, D.J. (1982). SYMLOG and leadership theory. In J.G. Hunt, U. Sekaran, & CA. Schriesheim (Eds.), Leadership: Beyond estabEishment views (pp. 165-195). Carbondale: Southern Illinois University Press. Bass, B.M. (1985). Leadership and performance beyond expectations, New York: The Free Press. Bums, J.M. (1978). Leadership. New York: Harper & Row. Byrne, R.F. (1971). The attraction paradigm. New York: Academic Press. Campbelll D.T., & Stanley, J.C. (1963). Experimental and quasi-experimentuE designs for research. Chicago: Rand McNally & Company. Conger, J.A., & Kanungo, R.N. (1987). Toward a behavioral theory of cha~smatic leadership in organizations. Academy of Management Review, 12,637-647. Conver, P.J. (198 1). Political cues and the perception of candidates. American Politics Quarterly, 9. 427-448. Cronbach, I.J., & Gleser, G.C. (1953). Assessing the similarity between profiles. Psychological Bulletine, SO, 456-473. Dow, T.E., Jr. (1969). The theory of charisma. Sociological Quarterly, 10,306-318. Edwards, J.E. (1993). Problems with the use of profile similarity indices in the study of congruence in organizational research. Personnel Ps.ychology, 46,64 l-665. Forgas, J.P., & Kagan, C. (1977). The cognitive representation of political personalities: A crosscultural comparison. International Journal of Psychology, 12, 19-30. Hare, A.P. (1976). Handbook of small group research. New York: Free Press. Hare, A.P. (1992). Groups, teams and social interaction. New York: Praeger. Hare, A.P., Borgotta, E.F., & Bales, R.F. (1965). Small groups, studies in social interaction, rev. ed. New York: Knopf. Herman, M.G. (Ed., 1986). A p~chological exanzjnation of polir~cu~ leaders. New York: Free Press, Hogan, D.B. (1988). The SYMLOG leadership profile as a predictor of managerial performance. In R.B. Polley, A.P. Hare, & P.J. Stone (Eds.), rl;rreSY~LOGprac~ifjoner. New York: Praeger. House, R.J. (1977). A 1976 theory of charismatic leadership. In J.G. Hunt & L.L. Larson (Eds.), Leadership: The cutting edge. Carbondale, IL: Southern Illinois University Press. House, R.J., Spangler, W.D., & Woycke, J. (1991). Personality and charisma in U.S. presidency: A psychological theory of leadership effectiveness. Administrative Science Quarterly, 36, 364396.

526

LEADERSHIP

QUARI-ERLY

Vol. 7 No. 4 1996

Isenberg, D.J., & Ennis, J.G. (1981). Perceiving group members: A comparison of derived and imposed dimensions. Jmrnal ofPersonality and Social Psychology, 42, 193-305. Jesuino, J.C. (1985). The assessment of leaders by SYMLOG. ~~te~zut~o~a~ Journal af~~al~ Croup Research, 1, 87-88. Kinder, D.R. (1985). Presidential character revisited. In R.R. Lau & D.O. Sears (Eds.), Political Cognition, Hillsdale, NJ: Laurance Erlbaum Associates. Kriger, M.P., & Barnes, L.B. (1988). Executive leadership network: Top management group dynamics in a hjgh-pe~o~ance company. In R.B. Polley, A.P. Hare, & P.J. Stone (Eds.), The SYMLOG practitioner. New York: Praeger. Marcus, J.T. (1961, March). Transcendence and charisma. Wesfern Political @tarter&, 14,236241. McClelland, DC, & Boyatzis, R.E. (1982). Leadership motive pattern and long-term success in management. Jmrnal ufApplied Psychology, 67,737-743. Meindl, J.R. (1990). On leadership: An alternative to conventional wisdom. In B.M. Straw & L.L. Cummings (Eds.), Research in Organizationat Behavior, vol. I2 (pp. 159-203). Greenwich, CT: JAI Press. Miller, A.H., Wa~enberg, M.P., & Malancbuk, 0. (1986). Schematic assessments of presidential candidates. American Political Science Review, 12,521-539. Newcomb, T.M. (1953). An approach to the study of communicative acts. PsychologicalReview, 60, 393-404. Nygren, T.E., & Jones, LE. (1977). Individual differences in perceptions and preferences for political candidates. Journal ofExperimental Social Psychology, 13, i82- 197. Pancer, SM. (1989). An application of attribution theory to the public’s perception of political leaders. Paper presented at the annual meeting of International Society of Political Psychology, Tel Aviv, Israel, June, 1989. Polley, R.B., Hare, A.P., & Stone, P.J. (1988). The SY~LOG pra~~t~i~~i~er:Applicut~~as of small group research. New York: Praeger. Sashkin, M. (1988). The visionary leader. In J.A. Conger & R.A. Kanungo (Eds.), Charismatic readership: The elusive factor in or~arz~~ationui electiveness (pp. 122- 160). San Francisco: Jossey-Bass. Shamir, B. (1992). Attribution of influence and charisma to the leader: The romance of leadership revisited. Journal of Applied Social Psychology, 22(5), 386-407. Shamir, B. (1994). Ideological position, leaders’ charisma, and voting preferences: Personal vs. partisan elections. P~?ljt~cal3e~avio~. 16(2). 265-287. Shamir, B., House, R.J., & Arthur, M.B. ( 1993). The motivational effects of charismatic leadership: A self-concept based theory. Organizational Science, 4(4), 577-594. Shikiar, R. (1974). The perception of politicians and political issues: A m~llti-dimensjonal scaling approach, M&variate Behaviorat Research, 9,461-477. Shikiar, R., Wiggins, N., & Fishbein, M. ( 1976). The prediction of political evaluation and voting preference: A multidimensional analysis. Journal of Research in Personality, 10,424-436. Weick, K.E. (1985). Systematic observational methods. In G. Lindzey & G. Aronson (Eds.), ~~~~b~~~ ofsocialps.~chology, vol. 1 (pp. 567-635.) New York: Random House. Willner, A.R. ( 1984). Tke spellbinders: Charismatic political leadership. New Haven: CT: Yale University Press. Winter, D.G. (1973). Thepou~er nzotive. New York: Free Press. Winter, D.G. (197%). Navy leadership and management competeneies: Convergence among tests, interviews and pet$ormance ratings. Boston: McBer. Winter, D.G. (1987). Leader appeal, leader performance, and the motives profiie of leaders and followers: A study of American presidents and elections. Journul of Persona& and Sociul Psyc~aiogy, 52,96-202.