Santa Clara University

Scholar Commons
Management Leavey School of Business

7-2004

A Leadership Development Instrument for

Students: Updated
Barry Z. Posner
Santa Clara University, bposner@scu.edu

Follow this and additional works at: http://scholarcommons.scu.edu/mgmt

Part of the Business Administration, Management, and Operations Commons

Recommended Citation
Posner, Barry Z. "A Leadership Development Instrument for Students: Updated." Journal of College Student Development 45.4
(2004): 443-56.

Copyright © 2004 The Johns Hopkins University Press. This article first appeared in Journal of College Student Development 45:4 (2004), 443-456.
Reprinted with permission by Johns Hopkins University Press.
http://dx.doi.org/10.1353/csd.2004.0051

This Article is brought to you for free and open access by the Leavey School of Business at Scholar Commons. It has been accepted for inclusion in
Management by an authorized administrator of Scholar Commons. For more information, please contact rscroggin@scu.edu.
A Leadership Development Instrument for
Students: Updated
Barry Z. Posner

This paper updates the research literature public-sector organizations (Freeman, Knott,
on the Student Leadership Practices Inven- & Schwartz, 1994). Serious questions have
tory, which is one of the few leadership been raised about whether such models are
development instruments targeted for college applicable to college students and collegiate
students. The psychometric properties of a environments, which differ considerably
revised version of the instrument are also from the environments in which managers
provided, along with a discussion of develop- and corporations operate. One way to
mental issues pertinent to developing and address this issue has been the development,
enhancing leadership capabilities in college over the past ten years, of a number of new
students. textbooks aimed at college students (e.g.,
Bratton, Grint, & Nelson, 2004; Daft, 2005;
Leadership development is now an integral Komives et al., 1998; Wren, 1995). Still, the
part of the educational program of college typical personal assessment techniques
students, with courses and activities scattered supporting these initiatives continue to be
throughout the co-curricular experience. borrowed from settings other than collegiate
Komives and her colleagues argue that environments. Brodsky’s (1988) observation
leadership, like any other skill, needs to be of more than 15 years ago is still generally
learned and practiced (Komives, Lucas, & applicable today: “Valid instruments de-
McMahon, 1998). Scholars like Wren (1995) signed specifically for college students to
assert that leadership is central to the human measure their leadership development do not
condition—timeless and current, not a exist” (p. 23). Consider, for example, that
passing fad—and that leadership should be while the 8th Edition of Leadership Re-
understood and practiced by all. Kouzes and sources: A Guide to Training and Devel-
Posner (2002) posit “leadership is everyone’s opment Tools (Schwartz & Gimbel, 2000)
business” (p. 383). Astin (1993) argues that lists 68 instruments “that are supported by
it is important to develop young men and technical data” (p. 195) to measure a variety
women during their college years to become of leadership skills and styles yet only two
future leaders. This is because leadership indicate a direct application to student
development encompassing various activi- populations. One of these is the student
ties, perspectives, and experiences enhances version of the Leadership Practices Inventory
the ability to make a meaningful difference. (Kouzes & Posner, 1998).
Many of the leadership development The Student Leadership Practices Inven-
programs designed for college students are tory (Student LPI) identifies specific be-
based upon studies and models that were haviors and actions that students report using
developed with managers in business and when they are at “their personal best as

Barry Z. Posner is Dean and Professor of Leadership at the Leavey School of Business, Santa Clara
University.

JULY/AUGUST 2004 ◆ VOL 45 NO 4 443

 Posner

leaders.” These behaviors are categorized school (junior or senior) and gender (male
into five leadership practices: Modeling the or female) to participate in this stage of the
Way, Inspiring a Shared Vision, Challenging research project. The students were asked to
the Process, Enabling Others to Act, and think about their personal-best leadership
Encouraging the Heart. Identified as prac- experience and to make notes about the
tices common to successful leaders, these behaviors they believed were most critical
leadership practices correspond well to the to the success of their endeavors.
developmental issues of importance for One week later, in a structured-interview
college students. format, each student responded to specific
This article explains the original devel- questions based on the personal-best survey
opment of the Student LPI and updates the reported in The Leadership Challenge
research literature relevant to its continuing (Kouzes & Posner, 1987). The interviews
reliability and validity. In addition, infor- lasted between thirty and ninety minutes;
mation is provided about a revised version each was tape-recorded with the respon-
of the Student LPI. Finally, several con- dent’s consent. The student interviews were
clusions are drawn about student leadership content analyzed for themes (sentences or
development and continuing challenges for phrases) about leadership actions and
both scholars and educators. behaviors. These themes were coded and
tabulated into the five leadership categories
DEVELOPMENT OF THE that had been originally proposed from
STUDENT LPI private-sector and public-sector managers.
These findings indicated that college student
In developing the original version of the leaders did engage in these leadership
Leadership Practices Inventory, Kouzes and practices and that this conceptual framework
Posner (1987) collected case studies from was relevant to the college students’ leader-
over 1,200 managers about their personal- ship experiences. A recent study by Arendt
best experiences as leaders. Content analyses (2004) followed a similar process for
of these case studies suggested a pattern of validating the appropriateness of the personal
behaviors used by people when they were best leadership case study methodology and
most effective as leaders. The development Student LPI for use with college students.
of a student version of the instrument She conducted in-depth, open-ended inter-
followed the same case-study approach to views with eight students about their per-
investigate whether the leadership behaviors sonal experiences that might typify the five
of college students were comparable with leadership practices. These interviews, she
those of managers (Brodsky, 1988; Posner concluded, “established the existence of
& Brodsky, 1992). leadership behaviors in hospitality manage-
The initial student group consisted of ment and dietetics undergraduate students as
outstanding student leaders at a large urban students described leadership behaviors in
state university campus, as demonstrated by each practice” (p. 26).
their nomination for Leadership America, a Each statement on the original LPI was
nationally prominent leadership develop- assessed in terms of its congruence with the
ment experience for college students. Four themes derived from case studies of students’
students were randomly selected by year in personal-best leadership experiences. The

444 Journal of College Student Development

Leadership Development Instrument

purpose of this coding was to determine Observer. Each form consists of 30 state-
which LPI statements accurately reflected ments—six statements to measure each of
the behavior of student leaders, thus facili- the five leadership practices. The forms differ
tating the process of identifying terminology only in terms of the individuals who com-
and concepts appropriate for use with a plete them. The Self form is completed by
college-student population. Using this data, the student leader himself or herself, and the
items were modified as necessary for use in Observer form is completed by a person who
the pilot version of the Student LPI. has directly observed the leadership be-
The pilot version of the Student LPI haviors of that student leader.
consisted of 30 descriptive statements
paralleling those found in the original LPI. EMPIRICAL STUDIES USING THE
Each of the five leadership practices was STUDENT LPI
assessed with six statements on the Student
LPI and each was measured using a five- A large number of empirical studies using
point Likert-scale (where 1 meant “rarely” the Student LPI have been conducted. Posner
and 5 meant “very frequently”). The state- and his colleagues presented many of the
ments focused on leadership behaviors and first research reports, validating the Student
on the frequency with which the individual LPI across multiple student populations and
engaged in those particular behaviors. investigating possible demographic vari-
Twenty-three members from a college ables. For example, fraternity chapter
Student Senate at a small private suburban presidents across the United States com-
college campus were asked to serve as the pleted the Student LPI-Self and had the
test group for studying the pilot version of members of their executive committees
the Student LPI. After these students com- complete the Student LPI-Observer. The
pleted the pilot version, they participated in members of the executive committees also
an item-by-item discussion to determine assessed the effectiveness of their chapter
whether any test statements were ambiguous, presidents along several dimensions: build-
confusing, or not applicable to their experi- ing team spirit, representing the chapter to
ences as student leaders. This discussion was administrators and alumni, meeting chapter
tape-recorded. Of the 30 test items, 25 (83 objectives, facilitating volunteers, and so on.
percent) were unanimously determined to be The most effective chapter presidents
clear and understandable and to consist of engaged in each of the five leadership
terminology and concepts that were within practices much more frequently than did
students’ and student leaders’ experiences. their less effective counterparts. Multiple
Ways to improve the somewhat problematic regression analyses showed that these
remaining items were also discussed and leadership practices accounted for 65 percent
determined. Five student leaders who had not of the variance in assessments of chapter
been involved with any of the earlier Student presidents’ effectiveness (Posner & Brodsky,
LPI efforts were invited to participate in a 1992).
focus-group discussion of the revised A study of sorority chapter presidents
Student LPI, and only very minor editorial from across the United States paralleled the
changes were suggested. previous study of fraternity chapter presi-
The Student LPI has two forms: Self and dents both in design and in findings (Posner

JULY/AUGUST 2004 ◆ VOL 45 NO 4 445

 Posner

& Brodsky, 1994). The most effective directors) and by their constituents. No
sorority chapter presidents engaged in each significant interaction effects between
of the five leadership practices much more gender and performance were found. This
frequently than did their less effective finding is consistent with other studies
counterparts. These leadership practices involving RAs and their residents (e.g., Levy,
accounted for 80 percent of the variance in 1995).
assessments of sorority chapter presidents’ The impact of leadership was also
effectiveness. investigated for students serving as orienta-
Together, these two studies also demon- tion advisors (Posner & Rosenberger, 1997).
strated that the practices of effective student In this study, incoming college students
leaders did not vary according to the leader’s completed both the Student LPI-Observer
gender. Effective chapter presidents, whether and a second evaluation of their orientation
male or female, engaged in the five leader- advisors’ effectiveness. Although together
ship practices significantly more than did the for just a few days, and in an arbitrary
less effective student leaders. This was true relationship in the sense that the members
from both the leaders’ perspectives and from of the groups did not select one another nor
the perspectives of people in their organi- did they select (or elect) their leaders (i.e.,
zations (Posner & Brodsky, 1994). Few the orientation advisors), the effectiveness
gender differences among college students of orientation advisors, consistent with
have been reported, from populations previous studies, was directly related to the
ranging from Greek chapter leaders in the extent to which the leaders engaged in the
Midwest (Adams & Keim, 2000), first-year five key leadership practices. Self-reports by
undergraduates (Mendez-Grant, 2001), or the orientation advisors themselves showed
students enrolled in either hospitality a strong positive relationship between
management or dietetics programs (Arendt, perceptions of effectiveness and the fre-
2004). quency that they reported engaging in these
Sample populations of resident advisors leadership practices.
(RAs) from seven diverse collegiate environ- How leadership practices might be
ments were studied (Posner & Brodsky, affected by various characteristics of the
1993). RAs completed the Student LPI-Self group or setting that students are involved
and distributed Student LPI-Observer forms with has been the focus of additional studies.
to residents and the resident director in their For instance, Posner and Rosenberger (1998)
housing facilities. Effectiveness data across reported that students who were being
several different sources—the RAs, the compensated for being leaders did not
students living in their residential units, and systematically engage in a different pattern
the resident director of each campus—were of leadership practices when compared with
collected, and a remarkably consistent those who were uncompensated for their
pattern was found. RAs who engaged in the leadership responsibilities. They also found
five leadership practices most frequently, as that student leaders working with peers in a
compared to those who engaged in the five non-hierarchical relationship did not engage
practices less frequently, viewed themselves in these leadership practices more or less
as more effective and were also viewed as significantly than those students who were
more effective by their supervisors (resident elected by their peers into official positions

446 Journal of College Student Development

Leadership Development Instrument

of leadership, or hold a hierarchical position Greek affiliation, or race” (p. 58).

such as president of a student organization. Walker (2001) found no significant
Edington (1995) found that the leadership differences in the pre- and post-test admini-
practices were not related to a student’s strations of the Student LPI following a
gender, race, age, work outside the home, leadership development intervention. She
full or part-time student status, or semester cited conversations with researchers at the
in school. Center for Creative Leadership who explain
In addition, it was revealed that students
that leadership development is not
did not vary their leadership practices when linear, rather leadership development
involved in a one-time leadership project will regress and progress. In the process
versus a project or program lasting for an of implementing leadership programs,
entire academic year. However, students who the researchers at The Center found that
returned for a second year in a leadership the immediate post test often showed
position significantly engaged in each of the negative development as opposed to the
five leadership practices more often than pretest. This may be a result of parti-
cipants increased awareness of the
those who were just starting in the same
multiple facets of leadership as they
position (Levy, 1995; Posner & Rosenberger, move through leadership training.
1998). Baxter (2001) found that students (pp. 110-111)
stationed as ROTC unit instructors (typically
in their fifth year of studies) had higher Using a modified version of the Student
leadership practices scores than did other LPI to reflect “self-efficacy” (i.e., I can or
students on the campus. Arendt (2004), in cannot do this), Endress (2000) reported
comparing students, found that those who higher scores at the conclusion of a leader-
had held official leadership positions and/ ship development program than at the
or taken courses in leadership reported program’s start. Completion of the leadership
higher leadership practices scores. class enhanced students’ beliefs in their
Mendez-Grant (2001) investigated the abilities to engage in leadership behaviors.
possible impact of leadership development This finding was independent of the students’
on the retention rates of first-year under- levels of co-curricular involvement. Another
graduates. While she found differences in the study reported that students in leadership
hypothesized direction, they were not positions, without any particular partici-
statistically significant. However, she did pation in leadership development programs,
find that pre- and post-test scores on the had high self-efficacy for leadership prac-
Student LPI were significant for those tices (Bardou, Byrne, Pasternak, Perez, &
students who went through a leadership Rainey, 2003). In these cases, gender did not
education program (treatment) versus those account for any differences in leadership
who did not (control group). Pugh (2000) practices, although females tended to feel
reported that participation in a leadership more supported and encouraged to develop
program resulted in higher leadership as leaders by their advisors than did their
practices scores at time two than at time one. male counterparts. One intriguing finding
These findings, he said, “were not explained was that the type of organization with which
by demographic variables: year in school, a student was involved seemed to influence
family cluster affiliation, gender, GPA, leadership self-efficacy. For example, those

JULY/AUGUST 2004 ◆ VOL 45 NO 4 447

 Posner

in activist organizations tended to have scale. Table 1 summarizes a number of these

higher leadership self-efficacies in Model- internal reliability scores across a variety of
ing, Inspiring and Challenging while those student populations. For example, internal
in cultural organizations scored higher on reliabilities (Cronbach’s alpha coefficients)
Encouraging. Additionally, students in on the Student LPI-Self ranged between
professional organizations tended to have α = .63 and α = .75 for fraternity chapter
higher scores for Enabling and Encouraging, presidents and between α = .66 and α = .75
while those in service-related organizations for sorority chapter presidents, while internal
demonstrated higher levels of leadership reliabilities on the Student LPI-Observer
self-efficacy in Modeling, Inspiring, and ranged between α = .75 and α = .84 for
Enabling. fraternity (executive committee) officers and
Only a few studies have explored rela- α = .85 and α = .90 for sorority (executive
tionships between student leadership as committee) officers (Posner & Brodsky,
measured by the Student LPI and various 1994). When these two samples were com-
personality dimensions. For example, several bined internal reliabilities ranged between
researchers have focused on the possible α = .80 and α = .87 for males and between
impact of learning and personality styles on α = .73 and α = .89 for females. Internal
leadership practices. Preferences for andra- reliability coefficients for RAs on the
gogical versus pedagogical learning have not Student LPI-Self were between α = .65 and
generally resulted in any statistically signi- α = .83, while for students in general
ficant leadership differences, although the (Student LPI-Observer) the range was
findings were in the predicted (andragogical) between α = .81 and α = .89 (Posner &
direction (Walker, 2001). Studies involving Brodsky, 1993). The internal reliabilities on
managerial populations have generally the Student LPI-Self for Orientation Advisors
shown a positive relationship between more ranged between α = .56 and α = .66, while
active learning styles and use of the five their constituents’ scores on the Student LPI-
leadership practices (Brown & Posner, Observer ranged between α = .80 and
2001). However, few significant relation- α = .85 (Posner & Rosenberger, 1997). For
ships have been found between the Achiev- students majoring in dietetics (across eight
ing Styles Inventory (Lipman-Blumen, 1996) college campuses) internal reliability scores
and the Student LPI (Snyder, 1992; on the Student LPI-Self ranged from α = .55
Schroggs, 1994). In one study, using a for Challenging to α = .76 for Inspiring
sample of female student leaders, there was (Arendt, 2004). For students majoring in
some correlation between the dimension of hospitality management (across eight college
introversion-extroversion on the Myers- campuses) the range of internal reliability
Briggs Type Indicator and use of the five coefficients on the Student LPI-Self was
practices, but no significant relationships α = .65 for Challenging to α = .78 for
between any of the additional personality Encouraging (Arendt, 2004). Other pub-
dimensions and the leadership practices lished studies not listed in Table 1 have
(Reeves, 2001). reported internal reliabilities for the five
Finally, within these many studies leadership practices between α = .63 for
involving the Student LPI, researchers have Challenging and Enabling and α = .83 for
reported on the internal reliability of the Inspiring for a sample of fraternity and

448 Journal of College Student Development

Leadership Development Instrument

sorority chapter presidents and their exec- reliability of the Student LPI over a ten-week
utive council members (Snyder, 1992); period was statistically significant, with
between α = .55 on Challenging and α = .70 correlations exceeding r = .51 (Pugh, 2000).
on Enabling for a study comparing the One observation from these various
campuses’ African-American sorority chap- published reports on internal reliability is
ter presidents with their Caucasian sorority that coefficient scores on the Student LPI-
chapter presidents counterparts (Williams, Self tend to be lower than those from the
2002); and between α = .83 for Challenging Student LPI-Observer. This discrepancy may
and α = .92 for Encouraging in a study be a result of measurement bias (in the form
involving RAs and the students in their of smaller samples sizes for self scores
residence hall (Levy, 1995). versus observer scores) or simply that there
Walker (2001) reports that analyses of is more variation within individuals (self
Student LPI scores with the Social Desir- scores) than across individuals (the per-
ability Index (Crown & Marlowe, 1960) spectives of observers). The sample sizes in
“confirms previous findings that indicated Table 1 suggest that the internal reliabilities
tests of social desirability bias were not showed some sensitivity to sample size
statistically significant” (p. 58). Test-retest because as the sample size increased, the

TABLE 1.
Internal Reliability Coefficients for the Student Leadership Practices Inventory by
Leadership Practice and Sample Characteristic

Sample
Characteristic Modeling Inspiring Challenging Enabling Encouraging

Males (304)1c .80 .86 .80 .87 .86

1c
Females (485) .73 .82 .83 .89 .88
Male (Fraternity) Presidents (65) 1a .63 .75 .60 .70 .73
Female (Sorority) Presidents (96) 1a .66 .75 .74 .67 .69
1b
Fraternity Officers (239) .75 .84 .76 .84 .83
1b
Sorority Officers (389) .85 .90 .85 .90 .90
2a
Resident Assistants (333) .69 .81 .65 .69 .83
RA Observers (1304) 2b .81 .89 .84 .82 .89
Orientation Advisors (78) 3a .61 .61 .56 .61 .66
3b
OA Observers (683) .80 .82 .80 .82 .85
4a
Dietetics Students (283) .67 .76 .55 .62 .73
4a
Hospitality Students (345) .68 .75 .65 .68 .78

Numbers in parenthesis refer to sample sizes.

Sources: 1Posner & Brodsky, 1994; 2Posner & Brodsky, 1993; 3Posner & Rosenberger, 1997; 4Arendt, 2004
a
= Student LPI-Self; b = Student LPI-Observer; c = Combined Student LPI-Self and Student LPI-Observer

JULY/AUGUST 2004 ◆ VOL 45 NO 4 449

 Posner

scores (Cronbach’s alpha coefficients) original statements, were presented to 31

generally increased as well. students in five focus group sessions in order
Overall, the Student LPI shows con- to determine item face validity; that is, the
sistent relationships with various measures extent to which Student LPI statements
of effectiveness, as reported across multiple reflected the actual behaviors of student
constituencies. Moreover, the Student LPI is leaders and the extent to which the state-
robust across different collegiate student ments were expressed in language and
populations (e.g., fraternities, sororities, terminology appropriate for use with college
residence halls, orientation programs, students. A student from each of the focus
academic majors, and the like). It appears groups volunteered to be part of a final focus
to be relatively independent of various group session in which agreement was
demographic variables (e.g., gender, age, reached for each of the statements and from
ethnicity, etc.) and possibly affected by level which the revised version of the Student LPI
of experience and/or education (e.g., pre- emerged.
vious leadership experiences and course In the end, two-thirds of the statements
work). in the revised version were identical with
those from the original version. In terms of
REVISING THE STUDENT LPI revisions, four statements were edited for
clarification or terminology and six state-
With the third edition of The Leadership ments were completely changed, primarily
Challenge, Kouzes and Posner (2002) for conceptual reasons and/or perceived
updated and restructured a modest amount redundancies. Revisions were made by
of their conceptual materials within the five leadership practice as follows: Modeling
practices of exemplary leadership frame- (three new statements), Inspiring (one edited
work. For example, a new chapter was added statement and one new statement), Chal-
(focused on “Finding Your Voice”) in lenging (one edited statement and one new
Modeling the Way. The “small wins” idea statement), Enabling (one new statement),
was moved from the Modeling leadership and Encouraging (two edited statements).
practice to Challenging the Process. The
order in which the five practices were Methodology
discussed was changed. In the third edition The psychometric properties of the revised
of the Leadership Practices Inventory (non- Student LPI-Self were tested using data
student version), the authors subsequently collected from fraternity chapter officers
either revised and/or replaced four items. (N = 604) on over 200 college campuses
These changes, along with regular psycho- across the United States. The same national
metric updating, provided the impetus for fraternity was selected that had been in-
proposing a review of the Student LPI volved in the initial empirical study of
instrument. college student leaders (Posner & Brodsky,
1992). A single college fraternity organi-
Revising the Instrument zation was selected in order to minimize the
A number of new statements were written potential effects of varying national policies
to assess the selected leadership behaviors. and procedures on local organizations.
These new statements, along with the Presumably all of the local chapters were

450 Journal of College Student Development

Leadership Development Instrument

structured and organized in roughly similar at a centralized level, seemed fairly typical
fashions, following nearly identical standard of the largest national fraternity organi-
operating procedures and having available zations in both scale and scope.
the same set of support services to the The Student LPI-Self was distributed to
chapters and their officers. all chapter officers attending one of six
The rationale for selecting multiple regional leadership academies held over the
chapters was to minimize the potential academic year. These officers represented
effects of any local campus policies and over 75 percent of the chapter officers in this
procedures and varying quality of student national fraternity organization. Each officer
support services across multiple campuses. attending the academy completed the survey
Both of these sample characteristics in- in a group setting conducted by a fraternity
creased the ability to generalize the findings. staff member or alumni volunteer. Upon
This particular national fraternity is one of completing the survey, each officer trans-
the top five national organizations in terms ferred his responses, but not his name, onto
of number of chapters and membership size. a separate page, provided some demographic
In addition, its chapter services operations, characteristics and placed all of this informa-

TABLE 2.
Demographic Characteristics of Chapter Presidents and
Executive Committee Officers
Percentage of Percentage of
Chapter Executive
Demographic Presidents Committee Officers
Characteristic (n = 113) (n = 491)

School Year
Freshman 0.0 8.8
Sophomore 14.4 43.6
Junior 58.6 37.2
Senior 27.0 10.3
Grade Point Average
Less than or equal to 2.5 4.4 8.2
2.51 – 2.99 22.1 25.2
3.00 – 3.49 43.4 40.5
Greater than or
equal to 3.5 30.1 26.2
Academic Major
Business 41.0 44.0
Engineering 12.4 15.0
Physical Sciences 26.6 29.2
Social Sciences / Humanities 20.0 11.8
Ethnic Background
Caucasian 93.0 91.6
Hispanic/Latino 4.0 4.9
Asian-American 1.0 3.0
African-American 2.0 0.5

JULY/AUGUST 2004 ◆ VOL 45 NO 4 451

 Posner

tion into a sealed envelope that was returned “seldom” and 7 indicating “all the time.”
to the staff member. Following the con- Internal reliabilities (Cronbach’s alpha) for
clusion of the academies, all surveys were this scale in previous studies had ranged
sent to the researcher for tabulation and between α = .77 and α = .94. In this parti-
scoring. In this way, all responses were cular study α = .79.
confidential and anonymous. Respondent
demographics are summarized in Table 2. Findings
A self-report scale that had been used Chapter presidents reported engaging in each
in the early studies of Greek chapter presi- of the five leadership practices more fre-
dent effectiveness was used to assess chapter quently than did the various other officers
effectiveness (Posner & Brodsky, 1992, in their fraternity chapters. Using paired
1994; Adams & Keim, 2000). This scale had t tests this difference was found to be statist-
also been successfully used to assess the ically significant, as reported in Table 3, for
performance of resident advisors (Posner & the leadership practices of Modeling,
Brodsky, 1993) and orientation advisors Enabling and Encouraging but not for
(Posner & Rosenberger, 1997). The scale Inspiring and Challenging. The rank order,
consisted of eight criteria of effectiveness on in terms of frequency, of the five leadership
which each chapter officer rated himself: practices was the same for the group of
1. The brothers view me as effective in presidents and group of all other officers.
meeting the chapter’s objectives. Enabling was most frequently engaged in
followed by Encouraging, Inspiring, Model-
2. I am successful at representing our
ing, and Challenging.
fraternity to faculty and administrators.
In order to determine whether engage-
3. I have developed a strong sense of ment in the various leadership practices was
cohesion and team spirit within the related to effectiveness, respondents were
chapter. categorized into more effective and less
4. I am able to get others in this chapter to effective categories by splitting the sample
volunteer for responsibilities. by average score on the effectiveness scale.
A t test comparing the more effective and
5. When this school year is over, the
less effective chapter officer groups revealed
brothers will be able to talk about the
that those chapter officers who viewed
difference I made.
themselves as more effective also con-
6. I am successful at representing our sistently reported more frequent engagement
fraternity to alumni. in each of the five leadership practices than
7. I am effective at getting the brothers did their counterparts who viewed them-
to care about this fraternity and its selves as less effective.
objectives. Also shown in Table 3 are comparisons
8. I am a positive role model as a chapter on the five leadership practices by ethnicity.
officer. T tests comparing Caucasian students and
students of color (combining the responses
Respondents indicated how descriptive each of the Hispanic/Latino, Asian-American and
statement was about them by using a seven- African-American students) revealed no
point Likert scale, with 1 representing statistically significant differences between

452 Journal of College Student Development

Leadership Development Instrument

how frequently students of color and Cauca- self-reports of effectiveness by respondents.

sian students reported engaging in these Likewise, individual respondent character-
leadership practices. ANOVAs for the three istics did not account for significant differ-
other demographic variables (year in school, ences in leadership practices in either study.
GPA, and academic major) were not statis- Furthermore, the internal reliabilities of the
tically significant (results not shown). Student LPI for these two samples were
Internal reliability coefficients (Cronbach’s relatively consistent.
alpha) are shown in Table 3 for each leader- While numerous studies have used the
ship practice. Modeling had the lowest Student LPI, this is the first study to make
reliability for the chapter presidents group use of the newly revised Student LPI. As
and the chapter officers group, and Encour- such, there are several limitations that should
aging had the highest reliability for both be acknowledged at this stage in the devel-
groups. opment and implementation of the revised
These results from the revised Student Student LPI. First, this particular sample is
LPI are compatible with those found in limited to male respondents, and the findings
previous studies involving the original may not generalize to female respondents.
Student LPI. This is particularly the case Second, only one type of student leader was
when comparing these results with Posner studied, which too may limit the gener-
and Brodsky’s (1992) prior study with the alizability of these findings. Third, other
same target population (i.e., fraternity demographic, institutional, and organiza-
chapter presidents). In both studies, scores tional characteristics beyond those addressed
on the Student LPI differentiated between in this study might account for differences

TABLE 3.
Comparisons for Leadership Practices by Position, Effectiveness, and Ethnicity;
and Internal Reliability Coefficients (Cronbach’s alpha)

Sample
Characteristic Modeling Inspiring Challenging Enabling Encouraging

Chapter Presidents (113) 22.75* 22.89 22.09 23.99** 23.37**

Other Officers (491) 22.09 22.33 21.47 23.03 22.50
More Effective (287) 23.53*** 23.87*** 22.82*** 23.79*** 24.09***
Less Effective (317) 21.02 21.13 20.46 22.68 21.37
Caucasian (485) 22.18 22.31 21.46 23.11 22.49
Non-Caucasian (43) 21.91 22.74 21.42 22.96 23.02
Chapter Presidents (113) 0.55 0.75 0.61 0.63 0.77
Other Officers (491) 0.59 0.67 0.64 0.61 0.77

Numbers in parentheses refer to sample sizes.

*p < .05. **p < .01. ***p < .001.

JULY/AUGUST 2004 ◆ VOL 45 NO 4 453

 Posner

in leadership practices. For example, cultural coefficients for some of the individual
diversity, leadership experience levels, practices are desirable.
socioeconomic status, chapter size, insti- Fourth, future studies of more increas-
tutional size and type, and other variables ingly diverse college student populations
may impact leadership practices. Fourth and would help to further our understanding of
finally, while Cronbach’s alpha coefficients both leadership and student development. It
across leadership practices are adequate, would be useful, for example, to look at more
stronger internal reliability coefficients diverse populations, such as student body
would help bolster this psychometric prop- officers, officers in professional clubs, sports
erty of the instrument. teams, peer educators, and even graduate
students. Equally revealing would be studies
THE STUDENT LPI AND STUDENT investigating various socio-cultural differ-
LEADERSHIP DEVELOPMENT ences, such as socioeconomic status, sexual
identity, disability, and the like. Expanding
Reviewing studies of college student leaders the investigation of collegiate leadership
involving the Student Leadership Practices education outside the United States would
Inventory, including the revised version, sug- also be of interest (some studies are currently
gest several working conclusions, thoughts underway with college students in the West
about future research, and areas for further Indies and Japan). Further instructive would
student leadership development efforts. First, be studies investigating the relationship
leadership matters. Student leaders repre- between leadership and such factors as
senting a variety of campus leadership cognitive complexity, personality typology,
positions who practice leadership behaviors thinking and learning styles.
measured in the Student LPI regard them- Fifth, studies examining the impact of
selves as more effective and are regarded by various leadership development programs
observers as more effective than those who and classes, especially over time, would
do not engage as frequently in leadership assist greatly in understanding just how
behaviors. Limitations of the most recent leadership is developed. For example, in one
investigation notwithstanding, analysis of study the use of a conceptual leadership
revised Student LPI data also revealed that framework and feedback was shown to
those who viewed themselves as more enhance the effectiveness of the fraternity’s
effective leaders than their peers consistently pledge education program and to signi-
reported engaging more in each leadership ficantly increase leadership practice scores
practice. Second, this finding from previous in a pre- and post-test condition (Matsos,
studies, and confirmed by the most recent 1997). Leadership development, as a visible
investigation, is robust and relatively un- component of new member development,
affected by a range of demographic variables was postulated to serve as a path for better
(e.g., gender, ethnicity, year in school, age, aligning the fraternity experience with the
GPA, academic major). Third, the revised aspirations of the academic community.
Student LPI meets acceptable psychometric Studies investigating just how leadership
standards of reliability and validity (Hair, development occurs would be invaluable not
Anderson, Tatham, & Black, 1998). How- just for those involved and responsible for
ever, somewhat higher internal reliability student leadership development, but also for

454 Journal of College Student Development

Leadership Development Instrument

people who provide leadership education engagement in leadership behaviors seem

for corporate, civic and community related to opportunities that students have to
organizations. reflect on their leadership experience, and
Finally, leadership educators and other themselves, as leaders. These opportunities
student affairs professionals can take comfort can be further facilitated through case
and even pride in knowing that leadership studies, leadership shadowing programs,
education programs and leadership classes journaling, guest speakers (role models), and
are apparently influencing the leadership advanced or follow-up experiences.
behaviors of students (at least in comparison
with these students’ undergraduate non-
Correspondence concerning this article should be
participating peers). It would be enlightening addressed to Barry Z. Posner, Dean, Leavey School
to know how long after graduation this of Business, Santa Clara University, Santa Clara,
influence might continue. More frequent California 95053; bposner@scu.edu

JULY/AUGUST 2004 ◆ VOL 45 NO 4 455

 Posner

REFERENCES
Adams, T., & Keim, M. (2002). Leadership practices and Levy, M. (1995). Followers’ perceptions of leaders:
effectiveness among Greek student leaders. College Prototypes and perceptions of resident assistants.
Student Journal, 34, 259-270. Unpublished master’s thesis, University of Maryland.
Arendt, S. (2004). Leadership behaviors in undergraduate Lipman-Blumen, J. (1996). The connective edge: Leading
hospitality management and dietetics students. Un- in an interdependent world. San Francisco: Jossey-Bass.
published doctoral dissertation, Iowa State University. Matsos, C. (1997). Student leadership development as a
Astin, A. (1993). What matters in college? San Francisco: supplement to college fraternity pledge programs.
Jossey-Bass. Unpublished doctoral dissertation, University of Alabama.
Bardou, K., Byrne, S., Pasternak, V., Perez, N., & Rainey, Mendez-Grant, M. (2001). A study of freshman interest
A. (2003). Self-efficacy and student leaders: The effects groups and leadership practices at Texas Women’s
of gender, previous leadership experiences, and insti- University. Unpublished doctoral dissertation, University
tutional environment. Journal of the Indiana University of North Texas.
Student Personnel Association, 33-48. Posner, B., & Brodsky, B. (1992). A leadership development
Bratton, J., Grint, K., & Nelson, D.L. (2004). Organizational instrument for college students. Journal of College Student
leadership. Mason, OH: South-Western. Development, 33(4), 231-237.
Baxter, W. (2001). Engineering leadership. Unpublished Posner, B., & Brodsky, B. (1993). The leadership practices
master’s thesis, Texas A&M University. of effective RAs. Journal of College Student Development,
Brodsky, B. (1988). Development of a modified version of 34(4), 300-304.
the Leadership Practices Inventory for use with college Posner, B., & Brodsky, B. (1994). Leadership practices of
students. Unpublished master’s thesis, San Jose State effective student leaders: Gender makes no difference.
University, California. NASPA Journal, 31(2), 113-120.
Brown, L., & Posner, B. (2001). Exploring the relationship Posner, B., & Rosenberger, J. (1997). Effective orientation
between learning and leadership. The Leadership & advisors are leaders too. NASPA Journal, 35(1), 46-56.
Organization Development Journal, 22(6), 274-280. Posner, B., & Rosenberger, J. (1998). The impact of
Crown, D., & Marlowe, D. (1960). A new scale of social situational factors on students’ leadership behaviors.
desirability independent of psychopathology. Journal of Working paper, Leavey School of Business, Santa Clara
Consulting Psychology, 24, 349-354. University.
Daft, R. (2005). The leadership experience (3rd ed). Mason, Pugh, D. (2000). College student leadership development:
OH: South-Western. Program impact on student participation. Unpublished
Edington, L. M. (1995). College classroom leadership doctoral dissertation, University of Georgia.
practices: What gender has to do with it. Unpublished Reeves, T. (2001). Psychological type and leadership
doctoral dissertation, Ball State University, Indiana. practices of women student leaders. Unpublished doctoral
Endress, W. (2000). An exploratory study of college students’ dissertation, University of Florida.
self-efficacy for relational leadership: The influence of Schroggs, C. (1994). Transformational leadership charac-
leadership education, cocurricular involvement, and on- teristics and achieving styles of selected student leaders.
campus employment. Unpublished doctoral dissertation, Unpublished doctoral dissertation, University of Missouri,
University of Maryland. Columbia.
Freeman, F., Knott, K., & Schwartz, M. (1994). Leadership Schwartz, M., & Gimbel, K. (2000). Leadership resources:
education 1994-1995: A sourcebook. Greensboro, NC: A guide to training and development tools. Greensboro,
Center for Creative Leadership. NC: Center for Creative Leadership.
Hair, J., Anderson, R., Tatham, R, & Black, W. (1998). Snyder, N. (1992). Empowering leadership and achieving
Multivariate data analysis (5th ed.). Englewood Cliffs, style: A study of gender differences between fraternity and
NJ: Prentice Hall. sorority presidents. Unpublished master’s thesis,
Komives, S., Lucas, N., & McMahon, T. (1998). Exploring University of Maryland.
leadership: For college students who want to make a Walker, J. (2001). Leadership development of students
difference. San Francisco: Jossey-Bass. engaged in experiential learning: Implications for
Kouzes, J., & Posner, B. (1987). The leadership challenge: internship programs in textiles and apparel. Unpublished
How to keep getting extraordinary things done in doctoral dissertation, University of North Carolina at
organizations. San Francisco: Jossey-Bass. Greensboro.
Kouzes J., & Posner, B. (1998). Student leadership practices Williams, E. (2002). Differences in Greek leadership styles.
inventory. San Francisco: Jossey-Bass. Unpublished research paper, Central Michigan University.
Kouzes, J., & Posner, B. (2002). The leadership challenge Wren, J. T. (Ed.). (1995). The leader’s companion: Insights
(3rd ed). San Francisco: Jossey-Bass. on leadership through the ages. New York: Free Press.

456 Journal of College Student Development