PSY 6002

Organisational Psychology B

Assessment item 1 Leadership instruments. (Word count = 3130)

Ian Johnson Student # 433402

Lecturer: Brioni Thompson

Measurement of Leadership
The systematic social study of leadership continues to evolve, allowing progressively better understanding of what a leader is supposed to have or be or do, and how to measure the multiple facets of this phenomenon. There have been numerous definitions and theories of leadership, especially by researchers and theorists during the systematic and scientific approach since the 1930's. During the earlier period of that systematic study, from 1930's to 1950, leadership was widely regarded as a 'gift' (Marland, 1972), that required certain traits. However, replication of studies supporting trait theory were rare, with little empirical support for personality trait theories, so the early leadership measures based on traits suffered poor psychometric properties (Stogdill, 1948). Subsequently, trait theory lost support, encouraging the development of alternative theories. To test the various leadership theories, researchers have used (and continue to use) questionnaires, various forms of election, nomination or ranking, observed leadership behaviour, and past leadership behaviour (Edmunds, 1998). The paper and pencil instruments to measure leadership tend to examine the leader or the followers or both, in terms of traits (Zaccaro, Foti, & Kenny, 1991; Fiedler, 1961), leader behaviour (Yetton, 1984), follower behaviour, situations (Yukl, 1988), interactions between leader and follower (Dansereau, Graen, & Haga, 1975), and charisma (House, 1977; Conger & Kanungo, 1987). A multitude of leadership theories followed trait theory, but far from superceding trait theories, have seemingly led back to include trait qualities. Revived interest in traits follows empirical support for a small number of traits (Bem & Allen, 1974), and particular traits seem to suit certain situations (Mischel, 1973). Trait predictability is now regarded as being a trait by itself (Bem & Allen, 1974). Interestingly, and unlike many other areas of psychology where theories compete, understanding of leadership has been notably evolutionary, with many theories building upon, (elaborating or correcting) previous ones. The most recent class of theories, referred to as 'New 2

Leadership Theories' (Bryman, 1993) includes such theories as Transformational Leadership (Burns cited in (House & Aditya, 1997) and Value Based Leadership (House cited in (House & Aditya, 1997). A syncretical model of the new leadership theories attempts to expose common threads of leadership theories, operationalise key constructs, and offer testable hypotheses (Behling & McFillen, 1996). However, no single model or theory has yet succeeded in encapsulating leadership, and can not do so until its definition has been agreed upon. To illustrate the difficulty faced in defining leadership, consider the different leadership responsibilities, behaviours and qualities between political and organisational leadership, and also between executive leadership, managerial leadership and supervisory leadership (House & Aditya, 1997). In the absence of an accepted and tested universal theory of leadership, there can be no generic 'leadership' instrument. The design of each existing leadership instrument depends upon the specific leadership theory on which it is based, and will target traits, behaviours, interactions, charisma or other dimensions accordingly. In other words, different leadership instruments measure different leadership facets. It follows that in measuring leadership, the person requiring the measure must understand exactly what is to be measured, and select the appropriate instrument. House & Aditya (1997) claim that nearly all leadership studies to date have been about supervisory leadership, and therefore nearly all currently available leadership instruments are limited in application to supervisors and managers. The optimal instrument for any organisation seeking to measure internal leadership, will depend upon exactly what information is required and what hierarchical level is to be measured, so a range of instruments covering varied uses and levels will be discussed in this paper. Starting at the very beginning of a leadership career, the first instrument described will be used to measure potential or emergent leaders, rather than existing leaders. The remaining instruments will attempt to cover the range of requirements from supervisory to national 3

leadership, and from stable unchanging organisations to those in distress or experiencing rapid change. In this way, the instruments attempt to cover the range from pre-job to top-job. Leadership Skills Inventory (LSI)(Karnes &Chauvin, 1985). Schools have a strong interest in assessing the leadership abilities of their students, as do organisations canvassing graduates for potential leadership positions. For schools the interest stems from offering constructive career guidance and training to students, and also to the link between leadership skills and self-actualisation (Karnes et al., 1993). Some instruments, in order to identify potential leaders and quantify their leadership skills, rely upon 'self' and 'other' instruments (student and teacher) and congruency between them (Anderson & Wanberg, 1991). If child and teacher both agree about the ownership of certain skills by the student, then they probably exist. When the Leadership Skills Inventory was tested in this way, student responses showed congruency with the teacher, but only if the teacher had leadership skills (Karnes & D'ilio cited in (Edmunds, 1998). With congruent validity demonstrated, only the student is now required to complete this instrument, making it simple for schools to use, and suitable for organisational recruitment purposes where the teacher is not available. The Leadership Skills Inventory is a 125 item Likert-style instrument covering nine categories of leadership. Those categories are: fundamentals of leadership; written communication; speech communication skills; 'values' clarification; decision-making skills; group dynamic skills; problem solving skills; personal development skills, and planning skills. This instrument exposes the leadership strengths and weaknesses of a student, enabling clear training plans to be tailor-made for that individual. A supporting study of the instrument provides empirical support of concurrent, content, and construct validity (Edmunds, 1998). Concurrent validity was supported by correlation (r=.34, p<.01) between LSI score and observed leadership behaviour. Content validity was 4

gained through expert jury agreement on content applicability of each question. Finally, a single factor was evenly and heavily loaded by each component (.636 to .876), supporting the construct validity, but not supporting the nine categories as separate components. The categories had much in common, with correlations between them ranging from r=.295 to .761, p<.01. The authors recommend that this instrument be used to guide students, but used in combination with other measures to comprehensively identify leadership ability. Comment: This is to test members of a group to predict emergent leadership or to assign group leadership, or for using on school children to test for leadership, or to monitor the effectiveness of leadership training. The information is of use to teachers, guidance officers, trainers, and the person tested. Least preferred co-worker (LPC) This contingency model for leadership assumes that the leadership style is fixed within each leader, and that each situation with its task structure and different level of position power requires a certain style of leadership that is a combination of task and relationship-orientated styles. The aim is to match leadership style with the situation requirement (Fielder cited in (Robbins et al., 1994). The instrument from this theory comprises eighteen 8 point bipolar adjective scales which are summed, and is unusual in that instead of asking about the leader, or the subordinate, it asks about the person with whom the leader (being tested) has found it most difficult to work. In other words, they describe their 'least preferred co-worker'. A high score indicates that the leader places importance on good personal relations, while a low score shows emphasis on productivity and tasks. Field studies have generally supported the theory and the instrument, while laboratory studies have been less supportive (Peters et al., 1985). Reliability has been supported at alpha . 64 and average internal consistency of coefficients of .88 (Rice, 1978). Other researchers, 5

however, find some cause to question Fiedler's interpretation of the LPC, with their results from the instrument suggesting a closer fit with a democratic - autocratic continuum (Edwards et al., 1990). Comments: Although updated by Fiedler to become Cognitive Resource Theory, the LPC still enjoys support. The participants chosen by researchers, and the theoretical concept, suggest that LPC is applicable to smaller organisations, or larger ones with varied environments, where low to middle managers can be matched with prevailing local leadership situations. Literature implies its usefulness where there are placements or selection choices to be made, and where there is no requirement or facility or intention to help, guide, train, or monitor leadership improvements. Because this theory assumes leadership style is fixed, the organisation most suited to this instrument is one that does not train or develop leaders, is not under stress or changing, and needs a best fit from the start. Leader member exchange (LMX-7) LMX theory is essentially about the dyadic relationship between an individual member and his or her supervisor, and describes reciprocal fulfillment of expectations (Graen et al., 1982). An important ingredient of the dyadic theory is the differentiated relationships yielding different qualities of exchange, in turn creating 'in' groups and 'out' groups of subordinates (Dansereau et al., 1975). LMX theory, based on social exchange theory (Sparrowe & Liden, 1997), maintains that the work group under a leader should not be considered a single entity (Dansereau et al., 1975). Instead, it distinguishes between the formal contract of 'management' and the socially influencing behaviour of 'leadership' (Jacobs cited in (Bhal & Ansari, 1996). LMX has been shown to positively relate to performance and satisfaction (Graen et al., 1986), locus of control and time-based pressure (Kinicki & Vecchio, 1994), delegation (Schriesheim et al., 1998), job attitudes and performance evaluations (Dienesch & Liden, 6

1986), and commitment and citizenship behaviour. A positive relationship that has the potential to confound the use of LMX measures, is between LMX and 'liking' of the member. Some organisational variables that demonstrated a negative relationship with LMX were turnover (Graen et al., 1982)(good for the organisation) and performance (Vecchio & Gobdel, 1984)(bad for the organisation). Literature showed more studies with positive LMXperformance relationships than negative, but there was considerable variation. The LMX-7 instrument contains 7 items that examine the quality of exchange between leader and subordinate, and is completed by the subordinates of the supervisor/manager/leader. Bhal & Ansari (1996) criticised existing LMX measures, including the LMX-7, for lack of psychometric rigour in their development. There is, however, considerable literature using the 7-item LMX measure from (Scandura & Graen, 1984), with most users reporting high alphas such as .91 (Klein & Kim, 1998), .86 (Scandura & Schriesheim, 1994) and .9 (Wayne et al., 1997). A refined version (LMX-6) has been produced from the same team (Schriesheim et al., 1992), and although empirically validated, has not yet enjoyed published replication or use. During development of LMX6, the tests for convergent and discriminative validity included the LMX7, providing the missing psychometric rigour. Additional psychometric validation for LMX7 occurred during the development and validation of a multidimensional measure of leader-member exchange (Liden & Maslyn, 1993) The LMX7 instrument can be used to examine the quality of social relationships perceived by subordinates to exist between themselves and their immediate leader. The testing of the instrument has been confined to organisational settings and is aimed at middle to lower management. The fact that it has not had meaningful exposure to higher levels of leadership, suggests that it should be limited to management levels under the executive layer. Comment: The LMX seems more useful in hierarchical organisation with managers and supervisors, rather than organisations with a flat structure. LMX does not apply at 7

executive level because high levels do not have the same co-worker/superior social support relationships. LMX also does not suit very high stress environments because there is a tendency for 'directing' behaviours to replace 'social influence' behaviours by supervisors under stress. (Staw et al., 1981) report that decisions made under stressful circumstances are, unlike LMX behaviour, likely to exhibit symptoms of the threat rigidity syndrome, meaning decision will be characterized by restriction of information-processing, constriction of control, and rigidity of response. Multifactor leadership questionnaire (MLQ) (Bass & Avolio, 1985) Charismatic leadership theory (House, 1977), having evolved from Path-goal theory (House, 1971), presents charismatic leaders as exceptionally self-confident, strongly motivated to influence, and with strong convictions of the morality of their beliefs. This theory applies to all levels of leadership within or external to organisations. A later version of the Path-goal theory (House, 1996) has been supported in a wide range of situations (Yukl, 1993). Over the same general period that charismatic theory has been developed, others such as transformational leadership (Bass & Avolio, 1985), attributional theory of leadership (Conger & Kanungo, 1987) and the value based leadership theory (House & Aditya, 1997) also emerged. This 'class' of theories became known as neocharismatic theories or 'the New Leadership Theory' (Den Hartog et al., 1997). The interest in this class of theories results from their generalisability across levels, genders, and cultures (summarised in (House & Aditya, 1997), and the high level of follower commitment and performance, especially when the organisation is under stress (Adamson, 1997) However, these theories have been critisised for limitations regarding measures of effects on followers, the single lumping of charismatic behaviours as a single undifferentiated syndrome, and inattention to important group characteristics (Shamir et al., 1998). Also, there are two factors hindering the successful study of charismatic/transformational leadership. One 8

is the fact that there is not a commonly accepted conceptual framework that can be used as a guide for research and practice, the second being the paucity of initiatives to operationalise major constructs (Shamir et al., 1998). In exposing these issues, a model has been offered as a solution (Behling & McFillen, 1996), but a literature search did not find subsequent support for the model by others. The instrument used most frequently for this form of leadership is the MLQ (Bass & Avolio, 1985). While there are several revisions of the MLQ, they all comprise subscales and use a 5 point Likert type scale. Being based on neocharismatic leadership, MLQ looks at traits and at the effects of leaders on followers, all measured through the eyes of the followers (Shamir et al., 1998). The MLQ attempts to measure transactional, transformational, and laissez-faire leadership in a leader's performance, where transactional and transformational leadership styles each comprise several sub-scale dimensions and accompanying behaviours (Bass & Avolio, 1990). The transformational subscales are charisma, intellectual stimulation, and individualised consideration, with alphas .91, .93, and .92 respectively. In a personal communication, Bass (cited in (House & Aditya, 1997) reported that MLQ charisma scale correlated above .90 with a longer descriptive (un-named) scale. Discriminant validity tests of the MLQ transformational measure supported a single order construct, with high inter-subscale correlations averaging .93 (Carless, 1998). The three-factor model of this leadership theory remains a valid concept for training purposes, but the high correlation suggests caution when interpreting sub-scales. More recently, the MLQ is being used to not just measure the transformational leadership, but the ratio of the three styles as perceived by the followers. In their study, (Den Hartog et al., 1997) found a one factor solution (leadership), a two factor solution (active and passive leadership), and a three factor solution that fitted the intended design. Their transformational result of alpha .95 correlated .99 with an inspirational instrument. The authors 9

reported an alternative grouping of questions from the MLQ gave better transactional results (alpha .79) than did the original (alpha .60). The internal consistency of the original MLQ groupings of questions did not return adequate internal consistency, but the three factors that emerged did fit the model, and improved the internal consistency. This and other research (Tracey & Hinkin, 1998) supports the underlying constructs. Comment: The MLQ is broadly applicable to a wide range of organisational, political, or other leadership situations, and at any level where there is responsibility to influence the actions of others. It can apply to managers, supervisors, or CEO's, and can indicate the extent of transactional, transformational, or laissez-faire leadership experienced by followers. It is probably wasted in a strict rule-driven bureaucratic environment where there is little opportunity for doing things differently or for personal growth. LPI Leadership Practices Inventory (Posner & Kouzes, 1993). This instrument grew not from a model, but from a desire to measure actual successful organisational leadership behaviours, and to shift away from psychological characteristics (Kouzes & Posner cited in (Posner & Kouzes, 1993). It moved from 'what they are like' to 'what do they do'. Surveys and interviews of over 1,100 leaders about their best personal leadership experiences yielded data about the ideal behaviours of leaders. A process of testing on another 2,100 managers and narrowing down the data, resulted in two 5 point Likert-style instrument of 30 questions covering 5 leadership practices. The 5 practices are: challenging the process, inspiring a shared vision, enabling others to act, modeling the way, and encouraging the heart. One of the instruments (LPI Observer) is for subordinates to report on their manager. The other (LPI Self), is for the manager to self-report. Reliability and validity tests involved another 2,876 managers and their subordinates, representing both genders, and 4 foreign countries. Internal consistency of the 5 sub-scales ranged from .70 to .91, with test-retest reliability .93 and above. 10

Further psychometric investigation (Posner & Kouzes, 1993) using data from 5,298 managers and 30,913 subordinates confirmed 5 factors corresponding to the hypothesised dimensions. The authors also reported small gender differences with females reportedly engaging more in modeling and encouraging. Female subordinates reported their managers as more challenging and higher modeling, regardless of manager gender, and LPI Self scores tend to be higher than LPI Observer. In addressing the initial restriction of the LPI to managerial populations, the LPI was modified to suit non-managers and individuals (Posner & Kouzes, 1994). Called PLI-IC:Self and LPC-IC:Observer, the new instrument returned alphas of .68 to .85 for the 'self' instrument, and .82 to .91 for the 'observer' instrument. Test-retest reliability returned .90 to .92. Once again, self-scores tend to be higher than observer scores. This instrument suits the emerging understanding that leadership is not about a position in the organisation, but is a skill exercised by any individual. LPI scores have been shown to relate to organisational effectiveness, work group vitality, job satisfaction and organisational commitment (summary cited in (Posner & Kouzes, 1993). LPI has differentiated between transactional and transformational styles (Fields & Herold, 1997), where challenging the process and inspiring a shared vision related to transformational leadership, while enabling others related to transactional leadership. Encouraging the heart and modeling the way related to both styles. Comment: Between the LPI and the LPI-IC, leaders throughout an organisation can be assessed. Not only can the LPI quantify the five sub-scale qualities of leadership practice, but by combining scores appropriately it can distinguish between transformational and transactional leadership style. Such a comprehensive range of information from a 30-item instrument may be useful to assess adjustments during periods of organisational change,

11

regardless of cultures or genders involved. This instrument is restricted to intra-organisational leadership.

Summary and recommendations Until the time that there is a comprehensive definition of leadership and a companion instrument, the instruments that enjoy high and consistent empirical or replicated support are to be preferred over others, no matter how convincing the theory or argument in their favour. All instruments mentioned in this report are supported, but use of any instrument must be on condition that it is appropriate to the intended application.

12

Reference List Adamson, T. (1997). An examination of gender differences in ideal leadership style using the Multifactor Leadership Questionnaire. Unpublished doctoral dissertation, Griffith University, Griffith. Anderson, S., & Wanberg, K. (1991). A congruent validity model of emergent leadership in groups. Small Group Research, 22(3). Bass, B., & Avolio, B. (1985). The multifactor leadership questionnaire. Palo Alto: Consulting Psychologists Press. Bass, B., & Avolio, B. (1990). The implications of transactional and transformational leadership for individual, team, and organizational development. Research in Organizational Change and Development, 4 , 231-272. Behling, O., & McFillen, J. (1996). A syncratical model of charasmatic / transformational leadership. Group and Organizational Management, 21(2), 163. Bem, D., & Allen, A. (1974). On predicting some of the people some of the time: The search for cross-situational consistencies in behaviour. Psychological Review, 81, 506-520. Bhal, K., & Ansari, M. (1996). Measuring quality of interaction between leaders and members. Journal of Applied Social Science, 26(11), 945. Bryman, A. (1993). Charismatic leadsership in business organisations: Some neglected issues. Leadership Quarterly, 4(3/4), 280-304. Carless, S. (1998). Assessing the discriminant validity of transformational leader behavoir as measured by the MLQ. Journal of Occupational and Organizational Psychology, 71(4), 353-358. Conger, J., & Kanungo, R. (1987). Toward a behavioural theory of charasmatic leadership in organisational settings. Academy of Management Journal, 12, 637-647. Dansereau, F., Graen, G., & Haga, W. (1975). A vertical diad linkage approach to leadership within formal organisations. Organizational Behaviour and Hmuan Performance, 13, 46-78. Den Hartog, D., Van Muijen, J., & Koopman, P. (1997). Transactional versus transformational leadership: An analysis of the MLQ. Journal of Occupational and Organisational Psychology, 70(1), 19(16). Dienesch, R., & Liden, R. (1986). Leader-exchange exchange model of leadership: A critique and further development. Academy of Management Journal, 11, 618-634. Edmunds, A. (1998). Content, concurrent and construct validity of the leadership skills inventory. Roeper Review, 20(4), 281-285. Edwards, J., Rode, L., & Ayman, R. (1990). The construct validity of scales from four leadership questionnaires. The Journal of General Psychiatry, 116(2), 171-181. 13

Fiedler, F. (1961). Leadership and leadership effectiveness traits: A reconceptualisation of the leadership trait problem. D. A. P. Porter (Ed), Studies in Organisational Behaviour and Management . Pennsylvania: International Textbook. Fields, D., & Herold, D. (1997). Using the Leadership Practices Inventory to measure transformational and transactional leadership. Educational and Psychological Measurement, 57(4), 569-580. Graen, G., Novak, M., & Sommerkamp, P. (1982). The effects of leader-member exchange and job design on productivity and job satisfaction: Testing a dual attachment model. Organizational Behavior and Human Performance, 30, 109-131. Graen, G., Scandura, T., & Graen, M. (1986). A field experiment test of the moderating effects of growth need on productivity. Journal of Applied Psychology, 71, 484-491. House, R. (1971). A path goal theory of leader effectiveness. Administrative Quarterly, 16, 321-338. House, R. (1977). A 1976 theory of charasmatic leadership. J. L. L. Hunt (Eds), Leadership: The cutting edge . Illinois: Illinois University Press. House, R. (1996). Path-goal theory of leadership: Lessons, legacy and a reformulated theory. The Leadership Quarterly, 7(3), 23-58. House, R., & Aditya, R. (1997). The social scientific study of leadership: quo vadis? Journal of Management, 23(3), 409-474. Karnes, F., Deason, D., & D'ilio, V. (1993). Leadership skills and self-actualization of schoolage children. Psychological Reports, 73(3), 861. Kinicki, A., & Vecchio, R. (1994). Influences on the quality of supervisor-suborinate relations: The role of time-pressure, organisational commitment, and locus of control. Journal of Organizational Behavior, 15, 75-82. Klein, H., & Kim, J. (1998). A field study of the influence of situational constraints, leadermember exchange, and goal commitment on performance. Academy of Management Journal, 41(1), 88-95. Liden, R., & Maslyn, J. (1993). LMX-MDM: Scale development for a multidimensional measure of leader-member exchange. National meeting of Academy of Management Chicago: University of Illinois. Mischel, W. (1973). Towards a cognitive social learning reconceptualisation of personality. Psychological Review, 80, 252-283. Peters, L., Hartke, D., & Pohlmann, J. (1985). Fiedler's contingency theory of leadership: An application of meta-analysis procedures of Schmidt and Hunter. Psychological Bulletin, March, 274-285. Posner, B., & Kouzes, J. (1993). Psychometric properties of the Leadership Practices Inventory - updated. Educational and Psychological Measurement, 53, 191-198. 14

Posner, B., & Kouzes, J. (1994). An extenstion of the Leadership Practices Inventory to indovidual contributors. Educational and Psychological Measurement, 54(4), 959-966. Rice, R. (1978). Construct validity of least preferred co-worker score. Psychological Bulletin, 85, 1199-1237. Robbins, S., Waters-marsh, T., Cacioppe, R., & Millett, B. (1994). Organisational behaviour. Australia: Prentice Hall. Scandura, C., & Schriesheim, C. (1994). Leader-member exchange and supervisor career mentoring as complimentary constructs in leadership research. Academy of Management Journal, 37(6), 1588(15). Scandura, T., & Graen, G. (1984). Moderating effects of initital leader-member exchange status on the effects of leadership intervention. Journal of Applied Psychology, 69, 428-436. Schriesheim, C., Neider, L., & Scandura, T. (1998). Delegation and leader-member exchange: main effects, moderators, and measurement issues. Academy of Management Journal, 41(3), 298. Schriesheim, C., Scandura, T., Eisenbach, R., & Neider, L. (1992). Validation of a new leadermember exchange scale (LMX-6) using hierarchically-nested maximum likelihood confirmatory factor analysis. Educational and Psychological Measurement., 52, 983992. Shamir, B., Zakay, E., Brenin, E., & Popper, M. (1998). Correlates of charasmatic leader behaviou in military units: subordinates' attitudes, unit of characteristics, and superiors' appraisals of leader performance. Academy of Management Journal, 41(4), 387-401. Sparrowe, R., & Liden, R. (1997). Process and structure in leader-member exchange. Academy of Management Review, 22(2), 522-552. Staw, B., Sandelands, L., & Dutton, J. (1981). Threat-rigidity effects in organizational behavior: A multilevel analysis. Administrative Science Quarterly, 26, 501-524. Stogdill, R. (1948). Personal factors associated with leadership: A survey of the literature. Journal of Psychology, 25, 35-71. Tracey, B., & Hinkin, T. (1998). Transformational leadership or effective managerial practices? Group & Organization Management, 23(3), 220-236. Vecchio, R., & Gobdel, B. (1984). The vertical dyad linkage model of leadership: Problems and prospects. Organizational Behavior and Human Performance, 34, 5-20. Wayne, S., Shore, L., & Linden, R. (1997). Perceived organizational support and leadermember exchange: A social exchange perspective. Academy of Management Journal, 40(1), 82-111. Yetton, P. (1984). Leadership and supervision. J. L. L. Hint (Eds), Leadership: The cutting edge . Illinois: Illinois University Press. 15

Yukl, G. (1988). Managerial leadership: A review of theory and research. Journal of Management, 15(2), 251-289. Yukl, G. (1993). A retrospective on Robert House's 1976 theory of charismatic leadership and recent revisions. Leadership Quarterly, 4(3/4), 367-373. Zaccaro, S., Foti, R., & Kenny, D. (1991). Self-monitoring and trait-based variance in leadership: An investigation of leader flexibility across multiple group situations. Journal of Applied Psychology, 76, 308-315.

16