JOURNAL OF SPORT PSYCHOLOGY, 1980,2,34-45

Dimensions of Leader Behavior in Sports: Development of a Leadership Scale P. Chelladurai University of Western Ontario

S. D. Saleh University of Waterloo Three different samples (total N = 485) participated in the development and refinement of the Leadership Scale for Sports (LSS). A five-factor solution with 40 items describing the most salient dimensions of coaching behavior was selected as the most meaningful. These factors were named Training and Instruction, Democratic Behavior, Autocratic Behavior, Social Support, and Positive Feedback. Internal consistency estimates ranged from .45 to .93 and the test-retest reliability coefficients ranged from .71 to .82. The relative stability of the factor structure across the different samples confirmed the factorial validity of the scale. The interpretation of the factors established the content validity of the scale. Finally, possible uses of the LSS were pointed out.

Although the concept of leadership has been discussed frequently and various leadership theories have been casually referred to in the sport literature, there has been a lack of consistent thrust in the study of leadership in sports (Loy, McPherson, & Kenyon, 1978). This is both unfortunate and surprising. The relevance of leadership theory to sports becomes readily apparent when sports teams are viewed in a formal organizational context. Ball (1975), for example, noted how sports teams fit the general description of formal organizations. More specifically, he pointed out that sports teams are characterized by (a) an unequivocal identity, (b) an exact roster of members including a roster of positions or statuses, (c) a planned program of activity and a division of labor to achieve specified goals, and (d) procedures for replacing members and for transfer of members from one position to another. The authors wish to thank Dr. Albert V. Carron and Dr. Ian Spence for their helpful comments in the preparation of this paper. This study could not have been completed without the kind co-operation of the coaches and the athletes of the various varsity teams, and the generous assistance of the local contacts. We express our sincere gratitude to them all. Requests for reprints and/ or copies of the LSS should be addressed to P. Chelladurai, Faculty of Physical Education, University of Western Ontario, London, Canada.

LEADERSHIP SCALE

35

If sports teams are viewed as formal organizations, then, the coach's role may be equated to that of management (see Sage, 1974). The many and varied managerial functions of the coach include planning, organizing, budgeting, scheduling, recruiting, public relations, leadership, etc. Of these, leadership, defined by Barrow (1977) as "the behavioral process of influencing individuals and groups toward set goals" (p. 232), is the most significant because other functions are performed away from the actual coaching context and can be performed by other individuals in the organization. Because the leadership provided by the coach is mainly instrumental in enhancing the motivational state of the group (House, 1971) and because, in turn, the motivational state of the group is the ultimate basis of performance effectiveness, research in leadership process in sports may significantly contribute to the understanding of sport performance. To date, most studies of leadership in sports have focused on the personality of the coach (e.g., Sage, 1975), or the coach's decision style-autocratic versus democratic (e.g., Lenk, 1977). Some studies have examined the application of Fiedler's (1967) Contingency Model of Leadership Effectiveness to the sport situation (e.g., Bird, 1977). And, finally, a recent and promising approach has as its focus the analysis of the varying behaviors of the coach which are appropriate to the different athletic situations (Chelladurai & Carron, 1978; Chelladurai & Saleh, 1978). In fact, Chelladurai (1978) has proposed a Multidimensional Model of Leadership which specifies that the effectiveness of leader behavior (i.e., the coaching behavior) is contingent on its congruence with the preferences of the members as well as the dictates of the situational characteristics. Although theoretical attempts to develop leadership models for sports are worthwhile, the clarity and cogency of any theory cannot be established without adequate leadership instruments. In fact, the elaboration of any theory entails an obligation to measure its constructs or to specify behavioral manifestations which can be adequately measured. Otherwise, theoretical formulations only yield a proliferation in terminology, instead of fulfilling a promise of empirical advance. Although the need for systematic research to identify and describe the dimensions of leader behavior in coaching and the need to develop valid scales to assess coaching behavior have been stressed (e.g., Cratty, 1973), there has been no attempt in this direction. There have been, of course, a number of leadership behavior instruments used in previous research in other organizational settings-the Leader Behavior Description Questionnaire (Halpin, 1957), the Supervisory Behavior Description Questionnaire (Fleishman, 1957a),the Leadership Opinion Questionnaire (Fleishman, 1957b), and the revised LBDQ-Form XI1 (Stogdill, 1963). However, it is also necessary to verify whether the dimensions identified in these instrumentsinitiating structure and consideration-as well as the specific items are relevant t o the sports context. The dimensions may not be pertinent because sport-as-anorganization is unique in the following ways: First, in athletics the organizational members spend a disproportionate number of hours in training for a competition of about 1 hour. This is in contrast t o the industrial and business situation where

36

CHELLADURAI AND SALEH

the training period is much shorter in duration. A second distinguishing characteristic of team sports is that the organizational rewards, namely winning, is denied to at least one of the contestants. Thus, the members of a team continuously strive for a reward with the realization that they may be deprived of it either through superior performance of the opponents or pure chance. Finally, the relatively brief existence for a team is yet a third distinguishing characteristic. Members of an athletic team are assembled for only about 3 to 6 months. There have been two studies conducted to identify the salient dimensions of leader behavior in sports. Danielson, Zelhart, and Drake (1975) administered a questionnaire containing 140 items modified from the original Leader Behavior Description Questionnaire (Hemphill & Coons, 1957) to 160 junior and senior high school hockey players. The results of the factor analysis and multidimensional scaling led the authors to conclude that "when perceived coaching behaviors are grouped on the basis of co-occurrence, the number of underlying dimensions is between 8 and 20" (Danielson et al., p. 332). Chelladurai and Saleh (1978), using 160 students enrolled in physical education degree programs at a Canadian University, derived and used a leadership scale consisting of five factors of coaching behavior. Although these two studies do represent a promising start, unfortunately, neither of the studies presented evidence of the reliability and validity of the scales used. Moreover, the samples were limited and therefore may not give an adequate indication of coaching behavior in different types of sports. The present paper is an attempt to deal with the above mentioned problems by refining the Leadership Scale for Sports (LSS) used by Chelladurai and Saleh (1978). This refinement consisted of determining the reliability and validity of this revised LSS scale.

Method In the first stage of the development of the LSS (Chelladurai & Saleh, 1978), 160 students (males = 80, females = 80) enrolled in physical education degree programs at a Canadian university responded to a questionnaire containing 99 items chosen and modified from existing leadership scales: Leader Behavior Description Questionnaire (Halpin, 1957), Supervisory Behavior Description Questionnaire (Fleishman, 1957a), Leadership Opinion Questionnaire (Fleishman, 1957b), and Leader Behavior Description Questionnaire-Form XI1 (Stogdill, 1963). Each item in the LSS was preceded with the phrase, "The coach should . . ."; and five response categories were provided: always, often, occasionally, seldom, and never. These data were factor analyzed using the principal factoring with iteration and varimax rotation as outlined by Nie, Hull, Jenkins, Steinbrenner, and Bent (1975). Of the several solutions extracting different numbers of factors, a fivefactor solution was found to be the most meaningful. The factors were labeled Training, Democratic Behavior, Autocratic Behavior, Social Support, and Rewarding Behavior. Items ( N = 37) were selected to represent these five dimen-

LEADERSHIP SCALE

37

sions of leader behavior on the basis of their high loading on one factor (.40 or more) and low loadings on any other factor (.30 or less). Subsequently, it was noted that none of the 99 items in the original pool tapped the behavior of the coach in teaching the skills and strategies of the sport. Although the term "training" connotes a certain degree of teaching, none of the items in the dimension of Training Behavior reflected the teaching behavior of the coach. Hence, in the second stage of development of the LSS, seven more items were included to tap the "instruction" behavior of a coach. Further, because Social Support is a dimension of leader behavior that is close to the traditional "Consideration" (Halpin & Winer, 1957)and because Social Support is a direct correlate of leader's orientation toward interpersonal relations (Fiedler, 1967) six more social support items were included. With the addition of these 13 items, the revised scale was increased to 50 items. In the revised version, the response categories were quantified in order to present a common frame of reference to all respondents. Accordingly, the response categories of "often," "occasionally," and "seldom" were anchored at 75%, 50%, and 25% of the time, respectively. The revised questionnaire was then administered to a different sample of 102 physical education students (males = 45, females = 57) and to a male sample of 223 varsity athletes (81 basketball players, 62 wrestlers, 57 track and field athletes, and 23 oarsmen) from different Canadian universities. The physical education students responded to the questionnaire in a class setting. For the athletes, a member of the faculty of physical education (or the athletic department) was contacted and requested to help in the collection of data. The set of questionnaires was sent to the contact who administered, collected, and returned them to the investigator. In most instances the administration was in a group setting, although in the case of a few wrestlers, the questionnaire was directly mailed to their home address. The physical education students were asked to indicate their preference for specific leader behavior in relation to their favorite sport. The athletes were asked to express their preference in relation to the sport in which they were currently competing. In addition, the athletes responded to a version of the scale in which they recorded their perceptions of the actual behavior of their current coaches. For the preference version, the items were preceded by "I prefer my coach to . . ."; and for the second version, the items were preceded by "My coach. . ."The three sets of data from physical education students' preference, athletes' preference, and athletes' perception were factor analyzed separately. The technique of principal factoring with iteration (Nie et al., 1975)which was used in the original version of the LSS was also used in analyzing the revised scales. I

Results and Discussion Because the second stage of the study was an attempt to confirm the factor structure predicted by the earlier factor analysis, five factors were extracted from each set of data. It must be pointed out that the five factors account for only

38

CHELLADURAI AND SALEH

limited amount of variance in each data set: 41.2% for physical education students' preferences, 39.3% for athletes' preferences, and 55.8% for athletes' perceptions. From the three solutions certain items were selected to constitute each factor. The criteria for selection of these items were (a) the item should have its highest loading on the same factor in all three solutions, and (b) its loading should be higher than .3 in at least two of the solutions. This procedure resulted in a total of 40 items: 13 items for training and instruction, 9 items for democratic behavior, 5 items for autocratic behavior, 8 items for social support, and 5 items for positive feedback (see Table 1). Schriesheim and Kerr (1977), in their critical analysis of the existing leadership scales, noted the difficulty of comparing data derived through the four different Ohio State scales even though they are purported to measure the same two dimensions of leader behavior-Initiating Structure and Consideration. This difficulty is caused by the fact that these Ohio State scales contain different sets of items. Our present procedure of selecting the items with the highest loading in all three solutions alleviates this problem and allows for the comparability of data collected from different groups (e.g., physical education majors and athletes) and between preferred and perceived leader behavior.

Internal Consistency and Reliability As an index of internal consistency, Cronbach's alpha was calculated for each subscale. These coefficients are presented in Table 2 and are considered acceptable. Though the coefficient was lower for Autocratic Behavior preferred by the athletes (.45) it was adequate for the other two sets of data. It should also be noted that the other coefficients are generally higher in the case of the athletes' perception of leader behavior than in the other two data sets. For the estimation of test-retest reliability, 53 of the physical education majors responded to the revised questionnaire a second time after an interval of 4 weeks. The composite factor scores (i.e., the sum of the selected items in each factor) were used to calculate the reliability coefficients. These reliability coefficients were adequate and ranged from .71 (Social Support) to .82 (Democratic Behavior).

Factorial Validity One method of establishing validity of a scale is to administer the scale to different samples and to verify if the factor structure remains stable across these samples (Kerlinger, 1973). Thus, the expectation was that five factors similar to those of the first stage (Chelladurai & Saleh, 1978) would be extracted from the data of the second sample of physical education students and from the two versions of the scale administered to the athletes. Table 1 shows that the factor structure is very similar across the three analyses-a finding which lends support to the suggestion that the factor structure is stable. The items and factors were also similar to those extracted in the earlier study (Chelladurai & Saleh, 1978).

LEADERSHIP SCALE

39

Table 1-Items and Factor Loadings in Three Solutions

Training and instruction See to it that every athlete is working to his capacity. Explain to each athlete the techniques and tactics of the sport. Pay special attention to correcting athlete's mistakes. Make sure that his part in the team is understood by all the athletes. Instruct every athlete individually in the skills of the sport. Figure ahead on what should be done. Explain to every athlete what he should and what he should not do. Expect every athlete to carry out his assignment to the last detail. Point out each athlete's strengths and weaknesses. Give specific instructions to each athlete as to what he should do in every situation. See to it that the efforts are coordinated. Explain how each athlete's contribution fits into the total picture. Specify in detail what is expected of each athlete. Democratic Behavior Ask for the opinionof the athletes on strategies for specific competitions. Get group approval on important matters before going ahead. Let his athletes share in decision making. Encourage athletes to make suggestions for ways of conducting practices. Let the group set its own goals. Let the athletes try their own way even if they make mistakes. Ask for the opinion of the athletes on important coaching matters. Let athletes work at their own speed. Let the athletes decide on the plays to be used in a game.

Physical education students

Athletes Preferred

Perceived

26

44

70

23

54

71

37

55

69

49

58

66

26 28

50 33

66 55

43

60

71

48

63

55

47

46

64

59

47

66

63

54

69

49

50

58

70

57

66

CHELLADURAI AND SALEH

Table 1 (cont'd.)

Autocratic Behavior Work relatively independent of the athletes. Not explain his action. Refuse to compromise a point. Keep to himself. Speak in a manner not to be questioned.

Physical education students

Athletes Preferred

Perceived

50 47 36 69

37 52 32 33

63 66 69

51

38

52

60

63

68

71

79

74

64

59

64

74

75 64

66 69

70

Social Support Help the athletes with their personal problems. Help members of the group settle their conflicts. Look out for the personal welfare of the athletes. Do personal favors t o the athletes. Express affection he feels for his athletes. Encourage the athlete to confide in him. Encourage close and informal relations with athletes. Invite athletes to his home. Positive Feedback Compliment an athlete for his performance in front of others. Tell an athlete when he does a particularly good job. See that an athlete is rewarded for a good performance. Express appreciation when an athlete performs well. Give credit when credit is due.

55

I t should be pointed out that the eigen values and variance of each factor were different among the three solutions. However, the difference in the eigen values and associated variances are t o be expected and should not be considered as evidence o f instability o f factor structure across situations. Since 40 o f the 50 items had the highest loadings in the same factors across the three samples, it can be concluded that the factor structure i s replicable. Further Cattell's (1966) Scree test lends support for the present decision t o extract five factors. Although the Scree test is somewhat outdated, it provides a rough estimate o f the maximum number o f factors t o be extracted. In general,

LEADERSHIP SCALE

Table 2-Internal

Subscale Training and Instruction Democratic Behavior Autocratic Behavior Social Support Positive Feedback

41

Consistency Estimates (Cronbach's Alpha)

Physical education Students ( N = 102) Preferred

Athletes (N = 223)

Preferred

Perceived

this test specifies that extraction of factors need not proceed beyond the point where the percentage of variance explained by successive factors levels off. This leveling off occurred after the sixth factor in each of the three solutions as illustrated in Figure 1. The five-factor solution was preferred since it was more meaningful than the six-factor solution. The eigen values of the factors in each solution exceeded the customary value of 1. It must be noted that factor analysis is a powerful though not the only method of establishing construct validity (Kerlinger, 1973, p. 468).

Content Validity Based on Factor Interpretation Not only was the five-factor solution stable but it was also found to be meaningful. Each of these factors is described and interpreted below. Factor I-Training and Instruction. The first factor, Training and Instruction, includes 13 items. It reflects one of the important functions of a coach-to improve the performance level of the athlete. The coach trains and instructs the athletes to help them reach their maximum physical potential. He or she is also expected to instruct them in how to acquire the necessary skills and to teach them the techniques and the tactics of the sports. In addition, in the case of team sports, the coach coordinates the activities of the team members. The Training and Instruction factor in the context of sport is similar to the Competitive Training factor identified by Danielson et al. (1975). It is also similar to the Instrumental Leadership dimension outlined by House and Dessler (1974) which essentially consists of role clarification, coaching, and coordination. Factor 2-Democratic Behavior. The factor of Democratic Behavior is composed of nine items. It reflects the extent to which the coach permits participation by the athletes in decision making. These decisions may relate to the setting of group goals and/or the ways in which these goals are to be attained. Factor 3-Autocratic Behavior. Autocratic Behavior, a factor with five items, indicates the extent to which a coach keeps apart from the athletes and stresses his or her authority in dealing with them. In such situations, it is expected that the coach would demand strict compliance with his or her decisions.

CHELLADURAI AND SALEH

P.E. MAJORS

1

2

3

4

5

6

7

8

9

10

FACTOR

Figure 1-Percent variance explained by successive factors.

It is of interest to note that the dimensions of Democratic and Autocratic Behavior refer to the decision style adopted by the leader whereas the other dimensions refer to the substance of the behavior. The emergence of the dimensions of Democratic and Autocratic Behavior is consistent with the distinction made in organizational behavior research between the decision style adopted by the leader and the content of his or her decisions (e.g., House & Dessler, 1974). Factot 4-Social Support. Whereas the training and instructing behavior is task oriented and the democratic and autocratic behaviors reflect the decision making approaches adopted by the coach, the Social Support factor (eight items) refers to the extent to which the coach is involved in satisfying the interpersonal

LEADERSHIP SCALE

43

needs of the athletes. The coach's behavior may directly satisfy such needs or the coach may create a climate in which the members mutually satisfy their interpersonal needs. It should be noted that social support is provided independently of member performance. This factor is similar to the Support factor in other leadership scales (Bowers & Seashore, 1966; House & Dessler, 1974). Danielson et al. (1975) in their study of coaching behavior identified a similar dimension which described socially oriented behavior outside the athletic situation. Factor 5-Positive Feedback. The factor of Positive Feedback includes five items. Athletic competitions, in general, are zero-sum games in which victory is attainable by only one of the contestants. An athlete, or team may perform at the maximum potential and yet lose a competition. Further, in team sports, contributions by certain positions may go unnoticed and unrecognized. It is therefore important for the coach to express appreciation and to compliment the athletes for their performance and contribution. Positive feedback from the coach is crucial in maintaining the motivational level of the athletes. Oldham (1976) found that it was one of the motivational strategies that predicted leader effectiveness. The path-goal theory (House & Dessler, 1974) posits that the effectiveness of leadership is related to the extent that the leader provides "rewards necessary for effective and satisfying performance that would otherwise be lacking in the environment" (p. 3 1). This factor was originally labeled as "Rewarding Behavior" (Chelladurai & Saleh, 1978). Although Social Support and Positive Feedback are two aspects of the traditional dimension of "Consideration," a distinction must be made between them. Social support behavior is not contingent on individual performance and is provided outside of the athletic context (Danielson et al., 1975) whereas positive feedback behavior depends on the athlete's performance. Positive feedback can be motivational only if it is contingent on performance. In summary, the LSS consists of one direct task factor (Training and Instruction), two decision-style factors (Democratic and Autocratic Behavior), and two motivational factors (Social Support and Positive Feedback). Although the five selected factors do not explain all of the total variance in the three sets of data,' the LSS provides the researcher with a valuable tool that has advantages over other proposed factor structures (e.g., Danielson et al., 1975). These dimensions are consistent with the path-goal theory of leadership (House & Dessler, 1974), they are conceptually distinct categories of coaching behavior, and each of the dimensions is relatively reliable. The LSS could be used profitably in the analysis of coaching behavior and its effectiveness. For instance, leadership theory and research suggest that leader behavior should be varied according t o the situation and the needs of the individual. The appropriateness of specific coaching behaviors could be assessed by administering the questionnaire to participants in 'An obvious limitation restricting the amount of explained variance in studies of this type is that the number and type of item statements included in the initial pool may not have been completely adequate.

44

CHELLADURAI AND SALEH

different sports (Chelladurai & Saleh, 1978). The differences in the preferences (or perceptions) expressed by athletes of different teams in a sport could be related to criteria of interest like satisfaction and performance (Chelladurai, 1978). The distinction between the decision style of the leader and the substance of his or her behavior holds promise for the matching of a decision style to specific sports situations (Chelladurai & Haggerty, 1978). Because scales similar to the LSS are used in other fields, it would facilitate comparison with and extensions of research findings from those fields.

References Ball, D. W. A note on method in the sociological study of sport. In D. W. Ball & J. W. Loy (Eds.), Sport andsocial order: Contributions to the sociology of sport. Reading, Ma.: Addison- Wesley, 1975. Barrow, J.C. The variables of leadership: A review and conceptual framework. Academy of Management Review, 1977,2,231-251. Bird, A.M. Development of a model for predicting team performance. Research Quarterly, 1977, 48,24-32. Bowers, D.G., & Seashore, S.E. Predicting organizational effectiveness with a four factor theory of leadership. Administrative Science Quarterly, 1966, 11,238-263. Cattell, R.B. The meaning and strategic use of factor analysis. In R.B. Cattell (Ed.), Handbook of multivariate experimentalpsychology. Chicago: Rand McNally, 1966. Chelladurai, P. A contingency model of leadership in athletics. Unpublished doctoral dissertation, University of Waterloo, Canada, 1978. Chelladurai, P., & Carron, A.V. Leadership. CAHPER Sociology of Sport Monograph Series, 1978. Chelladurai, P., & Haggerty, T.R. A normative model of decision styles in coaching. Athletic Administration, 1978. 13,6-9. Chelladurai, P., & Saleh, S.D. Preferred leadership in sports. Canadian Journal of Applied Sports Sciences, 1978, 3, 85-92. Cratty, B.J. Psychology in contemporary sport. Englewood Cliffs, N.J.: Prentice-Hall, 1973. Danielson, R.R., Zelhart, P.F., & Drake, D.J. Multidimensional scaling and factor analysis of coaching behavior as perceived by high school hockey players. Research Quarterly, 1975,46,323-334. Fiedler, F.E. A theory of leadership effectiveness. New York: McGraw-Hill, 1967. Fleishman, E.A. A leader behavior description for industry. In R.M. Stogdill & A.E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: The Ohio State University, Bureau of Business Research, 1957. (a) Fleishman, E.A. The Leadership Opinion Questionnaire. In R.M. Stogdill& A.E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: The Ohio State University, Bureau of Business Research, 1957. (b) Halpin, A.W. The leader behavior and effectiveness of aircraft commanders. In R.M. Stogdill & A.E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: The Ohio State University, Bureau of Business Research, 1957. Halpin, W.W., & Winer, R.J. A factorial study of the leader behavior description. In R.M. Stogdill& A.E. Coons (Eds.), Leader behavior: Itsdescription andmeasurement. Columbus: The Ohio State University, Bureau of Business Research, 1957. Hemphill, J.K., &Coons, A.E. Development of the leader behavior description questionnaire. In R.M. Stogdill & A.E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: The Ohio State University, Bureau of Business Research, 1957.

LEADERSHIP SCALE

45

House, R.J. A path-goal theory of leader effectiveness. Administrative Science Quarterly, 1971, 16, 321-38. House, R.J., & Dessler, G. The path-goal theory of leadership: Some post hoc and a priori tests. In J.G. Hunt & L.L. Larson (Eds.), Contingency approaches to leadership. Carbondale, 11.: Southern Illinois University Press, 1974. Kerlinger, F.N. Foundations of behavioral research. New York: Holt, Rinehart, & Winston, 1973. Lenk, H. Authoritarian or democratic styled coaching? In H. Lenk (Ed.), Team dynamics. Champaign, 11.: Stipes, 1977. Loy, J.W., McPherson, B.D., & Kenyon, G. Sport and social systems. Reading, Ma.: Addison-Wesley, 1978. Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K., & Bent, D.H. Statisticalpackage for the social sciences (2nd ed.). New York: McGraw-Hill, 1975. Oldham, G.R. The motivational strategies used by supervisors: Relationships to effectiveness indicators. Organizational Behavior and Human Performance, 1976, 15, 66-86. Sage, G.H. The coach as management: Organizational leadership in American sport. In G.H. Sage (Ed.), Sport and American society. Reading, Ma.: Addison-Wesley, 1974. Sage, G.H. An occupational analysis of the college coach. In D.W. Ball & J.W. Loy (Eds.), Sport andsocialorder: Contributions to thesociology of sport. Reading, Ma.: Addison-Wesley, 1975. Schriesheim, C.A., & Kerr, S. Theories and measures of leadership: A critical appraisal of current and future directions. In J.G. Hunt & L.L. Larson (Eds.), Leadership: The cutting edge. Carbondale, 11.: Southern Illinois University Press, 1977. Stogdill, R.M. Manual for the Leader Behavior Description Questionnaire-Form XII. Columbus: The Ohio State University, Bureau of Business Research, 1963. Manuscript submitted: 1/3/79 Revision received: 121191 79