You are on page 1of 26

Journal of Leadership & Organizational Studies

http://jlo.sagepub.com/ The Self-Leadership Questionnaire: Preliminary Assessment of Construct Validity


Joe S. Anderson and Gregory E. Prussia Journal of Leadership & Organizational Studies 1997 4: 119 DOI: 10.1177/107179199700400212 The online version of this article can be found at: http://jlo.sagepub.com/content/4/2/119

Published by:
http://www.sagepublications.com

On behalf of:

Midwest Academy of Management

Additional services and information for Journal of Leadership & Organizational Studies can be found at: Email Alerts: http://jlo.sagepub.com/cgi/alerts Subscriptions: http://jlo.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Apr 1, 1997 What is This?

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

The

Self-Leadership Questionnaire: Preliminary Assessment of Construct Validity


Joe S. Anderson Northern Arizona University

Gregory E.

Prussia Seattle University

Executive

Summary

As organizations approach the 21st Century, employee empowerment has become an important concern in coping with current competitive demands. Empowerment demands non-traditional relationships between organizations and their employees. At the heart of empowerment lies employees ability to lead themselves. However, as important as self-leadership skills are to the success of empowerment programs, no valid measure of self-leadership has been available. Three studies are reported which contribute to the refinement and preliminary validation of a measurement of self-leadership skills. Results, limitations, and a suggested research agenda contributing to practical benefits are discussed.

About the Authors: Joe Anderson is Assistant Professor of Management at Northern Arizona University. He received his doctorate from Arizona State Universitys Department of Management in May, 1993. His research interests include stress and coping in organizations, strategic alliance formation in small businesses, and organizational privacy issues. Gregory Prussia is Assistant Professor of Management at Seattle University. He received his Ph.D. from Arizona State University in December, 1993. His current research interests include the causal relationship between organizational structure and strategy, the antecedents and consequences of job loss, and group collective efficacy.

The authors acknowledge valuable suggestions by three anonymous reviewers of the of Management on an earlier version of this paper.

Academy

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

120

Organizations are undergoing considerable change as attempts to respond to domestic and global competitive pressures focus attention on internal organizational processes and efficiency. Current changes in organizations are widespread and expected to continue well into the 21 st Century (Kozlowski, Chao, Smith, & Hedlund, 1993). All levels of organizations are in flux, and recent prevalent downsizing has affected management to a greater extent than ever before (Byrne, 1993). Consequently, an emphasis on employee has as an important feature of many organizational empowerment emerged restructuring efforts (Alvesson & ~Imot, 1992; Conger & Kanungo, 1988; Thomas & Velthouse, 1990). Empowerment of employees is a key foundation of self-managed workteams, participative management, and other attempts to extend quality concepts into organizations. At the heart of the movement toward empowerment of employees is a vision of new relationships between the organization and its employees. Employees are now expected to be intelligent and active participants in their organizations (Aktouf, 1992; Block, 1993). Control in organizations is shifting from external, downward influence by management on employees to a decentralization of power, and an opportunity for workers at all levels to exercise increasing influence over themselves and their tasks (Shipper & Manz, 1992, p. 48).
Recognition is growing that in empowering organizations, managers will need to rely more on internal employee self-leadership, rather than on external leadership as it has been traditionally applied. For example, self-leadership is considered pivotal to employees enthusiasm, commitment, and performance in empowering organizations (Manz, 1986, 1990). In addition, positive selfleadership is argued to contribute to improved performance for both individuals and their organizations (Manz, 1992b). Several authors argue that employee empowerment and the related issue of employee self-leadership are expected to be recurring themes in organizational restructuring, and thus will continue to be of concern to practicing managers well into the 21 st Century (Alvesson & Wilmot, 1992; Manz, 1992a).
on

However, while anecdotal evidence of the effects of employee self-leadership other organizational variables is common and has strong intuitive appeal,

there has been little empirical examination of self-leadership skills and their relation to performance or other important related outcomes in organizations (Cf. A clearer Neck & Manz, 1994; Prussia, Anderson, & Manz, 1997). understanding of the process of self-leadership is needed in order to facilitate empowerment of employees by organizations restructuring for survival in the
new

century.

To provide research which can substantiate the theoretical grounding and intuitive appeal of the concept of self-leadership, and further can provide value in practical use, the first step must be validation of a measurement instrument. This study focuses on the measurement of self-leadership skills. Heretofore there has been no valid measure of self-leadership. It is our purpose, therefore, to present the preliminary development of a survey to measure self-leadership behaviors and perceptions in individuals. Conceptual foundations of self-

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

121

leadership theory are now presented, followed by content validation results (Study 1) and further construct validation (Study 2 and Study 3).
Self-Leadership Theory

Self-leadership involves the influence people exert over themselves to achieve the self-motivation and self-direction needed to behave in desirable ways (Manz, 1992b). Particular behaviors and thought processes characterize self-leadership and are hypothesized to influence subsequent outcomes positively. Specifically, three distinct but complementary categories of self-leadership have been hypothesized: behavior focused strategies; natural reward strategies; and constructive thought pattern strategies.
self-leadership strategies refer to specific behaviors that focus self-assessment, self reward, and self-discipline. These behaviors are aimed at managing behavior in difficulty, necessary, but perhaps unpleasant tasks. Five primary constellations of behavioral strategies contribute to enhanced selfleadership: self-observation; self-goal setting; self-reward; self-punishment; and rehearsal. Overall, the foundation for these strategies is a focus on selfawareness, an understanding of why we behave in desirable or undesirable ways, and using strategies to manage this process (Manz, 1992b). Using these behavior focused strategies is argued to promote and encourage successful
on

Behavior focused

&dquo;

;;

behaviors and suppress unsuccessful behaviors.


Natural reward strategies of self-leadership focus on positive perceptions and experiences of tasks to be accomplished. This includes commitment to, belief in, and enjoyment of the work for its own value (Manz, 1992a, 1992b). Natural reward strategies include seeking work activities which are pleasant and enjoyable. Further, perceptions or behaviors relating to existing tasks can be modified in order to increase perceived competence, self-control, or task responsibility. In general, strategies in this category include increasing focus on the pleasant rather than the unpleasant aspects of work.

..

thought pattern strategies focus on establishing and altering thought patterns in desirable ways. Four particular strategies can be used to change thinking patterns and enhance self-leadership: self-analysis and improvement of belief systems; mental imagery focused on positive performance; positive self-talk to motivate and facilitate performance; and use of positive scripts in place of ineffective ones. Manz (1986) asserts that behavioral and psychological scripts are the individual counterparts to organizational rules, policies, and operating procedures.
.

Constructive

..

.&dquo;
..

.....

The construct of self-leadership extends beyond the frequently used term &dquo;selfmanagement.&dquo; Theorists propose that though self-leadership shares some similar components with self-management, self-management tends to focus primarily on behaviors to reduce discrepancies from often externally imposed standards (Manz, 1992b). While self-management is usually used in reference to self-discipline oriented activities related to accomplishing necessary but not usually motivating behaviors (cf. Latham & Frayne, 1989), self-leadership is a
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

122
more

global concept which

also

incorporates additional cognitive and intrinsic

motivational aspects.

Further, self-leadership is distinct from the concept of employee empowerment. Empowerment has been described as &dquo;pushing decision-making down to the lowest level&dquo; in organizations (Roberts, 1994, p. 311). Another view of empowerment describes it in terms of providing employees with the Four As: 1) authority, delegated to the employees to make decisions; 2) accountability, accepted by the employees for their actions and results of their decisions; 3) alignment of employee direction, meaning a shared vision with the organization; and 4) ableness, or development of the employees skills over time (Brower, 1994). A third perspective on empowerment includes management efforts to &dquo;enable; it means to help people develop a sense of self-efficacy... to energize people to take action&dquo; (Whetten & Cameron, 1995, p. 483). In each of these cases, empowerment is the provision of opportunities externally by management, in effect providing an environment in which employees can exercise their self-leadership skills and perceptions. The issue of self-efficacy provides a good example. While empowerment means helping people develop a sense of self-efficacy, previous research demonstrated through sophisticated structural equation modeling techniques, that self-efficacy and self-leadership are separate and distinct constructs (Prussia, Anderson, & Manz, 1997).
.

.
.

In sum, individuals use of behavioral, cognitive, and perceptual self-leadership skills is theorized to build perceptions of control and responsibility which are vital to employees performance in empowering organizations. For this reason, we We examine the measurement of self-leadership skills in this study. the of Kinicki valid measures & Latack, acknowledge (cf. complexity developing 1990), and recognize that our approach in these early stages must be considered preliminary. However, we have built upon multiple sources of existing theory to develop our instrument, and any further pursuit of empirical research in this important topic area depends to a great degree on inchoate steps such as we take here. For example, a valid scale for measuring self-leadership skills is needed as an indicator of employee readiness for empowerment interventions in organizations, and further as a measure of the effectiveness of those interventions after they have been implemented. The content validity of the SelfLeadership Questionnaire is assessed in Study 1. Study 2 examines the scale further for a large sample, evaluating the scales psychometric properties through factor analysis. Study 3 tests the scale with a second sample, further evaluating its psychometric properties by examining reliability estimates and intercorrelations between sub-scales.
.

:; ,.I

Study 1

Content validity is an important component of construct validity (Schwab, 1980). Content validation assesses whether the items in a measurement instrument adequately represent the performance domain or construct of specific interest (Crocker & Algina, 1986, p. 218). Further, the instrument must not contain items measuring extraneous constructs to qualify as content valid (American Psychological Association, 1985; Cook & Campbell, 1979). Developers of
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

123 measurement instruments are cautionea

against concluding that scales

are

content valid

simply because they

have been constructed to conform to the

theoretical construct.
In content validation, a typical procedure is to have a panel of independent judges determine whether instrument items adequately sample the domain of interest. Judges are either subject matter experts or are supplied with clear descriptions of each item and of the theoretical domains into which the items may be assigned. The judges are instructed to proceed systematically through the instrument, matching each individual item to a list of objectives (Crocker & Algina, 1986, p. 219). They progress through the instrument, just as survey respondents would, but classify the items into their theoretical domains rather than simply answering each individual question. Items assigned into predicted categories indicate that the items assess the specified domain. Items assigned to other than the theoretically specified categories demonstrate that the items
assess

extraneous content

or are

not

independent (American Psychological

Association, 1985).

Study 1 uses the recommended approach of having independent judges determine the content validity of the prototype instrument, the Self-Leadership Questionnaire (SLQ). The prototype was generated by C. Manz and H. Sims from the theoretical and experiential literature on self-leadership skills (e.g., Manz, 1983, 1986, 1992b; Manz & Sims, 1991). The initial SLQ contained 90 items, designed to tap the three theoretical dimensions of Self-Leadership. Each item was intended to represent one of the dimensions of self-leadership: behavioral focused strategies; natural reward strategies; and constructive
thought pattern strategies.
Method

Sample and Procedure


The

prototype SLQ and an introductory manual were distributed to 18 judges at a large southwestern university. The judges included 5 faculty members in the Management Area of the College of Business and 13 graduate students in the College who were enrolled in a graduate Organizational Behavior and Organization Theory course. Judges volunteered to participate in the validation exercise, in response to a memo which briefly introduced the research project and the topic of self-leadership.

respondents were asked to read a six-page &dquo;training manual&dquo; (available by request from the authors) which gave the instructions for the exercise and presented the dimensions of self-leadership theory. Two descriptions of selfleadership theory were included in the training manual: 1) An initial overview of the self-leadership construct adapted from recent work by Prussia, Anderson, and Manz (1994); and 2) a more detailed introduction to self-leadership theory adapted from a recent self-leadership article by Manz (1992b). On completion of their reading, and prior to responding to the questionnaire, respondents were asked whether they had read the introductory material and whether they had
The
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

124
;:

understood it. All respondents answered these questions in the affirmative. Respondents were encouraged to consult the introductory material whenever they felt it would help them in their categorization of the questionnaire items. Respondents then took between 15 and 30 minutes to categorize the SLQ items on the questionnaire
-.
:

The categorization exercise asked the respondents to assign the 90 randomly ordered items in the prototype SLQ into one of the three dimensions of selfleadership or one of two additional dimensions, as follows: 1. Behavioral Focused Strategies 2. Natural Reward Strategies 3. Creating Constructive Thought Pattern Strategies 4. The item fits into more than one of the categories above. 5. The item fits into none of the categories above.

&dquo;

Examples of behavioral focused items were, &dquo;I write specific goals for my own performance; I feel guilt when I perform a task poorly.&dquo; Examples of natural
reward focused items were, &dquo;I look for and try to increase the activities at work that I enjoy; When I have a choice, I try to do my work in ways that I enjoy rather than just trying to get it over with.&dquo; Examples of creating constructive thought pattern strategies were, &dquo;I articulate (vocalize) my images of seeing myself successfully performing a task; Sometimes I talk out loud to myself to work through a difficult situation.&dquo;
~

Analysis

,--

To demonstrate content validity, judges should assign items to the a priori theoretical categories to which they belong (Kinicki & Latack, 1990, p 343). Judges responses for each of the five categories were totaled for the 90 prototype SLQ items. Normal procedure is to drop items with less than 50 to 60% agreement for the judges (cf. Kinicki, Heyl & Callerman, 1986; Kinicki & Latack, 1990). To provide a conservative test of instrument validity, we chose a cutoff of 60% agreement. To ascertain reliability of judges assignments, we also compared the categories of item assignments by the group of faculty judges against item assignments by the group of graduate student judges for each of the 90 SLQ items.
.

Results and Discussion

Results indicated that of the 90 items in the prototype SLQ, 50 items exceeded the criterion for agreement and were correctly assigned by the judges to their proper a priori theoretical categories. Agreement by the judges ranged from 61 % to 94%, indicating moderate to excellent agreement on categorization of those items. For the 50 SLQ items, 27 were assigned to the behavioral focus strategies category, 8 were assigned to the natural rewards strategies category, and 15 were assigned to the creating constructive thought pattern strategies category. Table 1 shows the final categorization of the 50 items remaining in the &dquo;refined&dquo; SLQ, along with the levels of judges agreement for each item.
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

125

TABLE 1 Refined Self-Leadership Questionnaire


with

Assignments and Percentage Agreement by Judges

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

126

The Jou rnal of

Leadership Studies

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

The

Self-Leadership Questionnaire

127

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

128

When item assignments were compared between faculty judges and graduate student judges, a high degree of agreement was found between the two groups. Overall, for the 90 items of the prototype SLQ, the groups agreed on 67% of the categorizations. For the 50 item refined SLQ, faculty and student groups of judges agreed on 94% of item assignments. These results suggest initial support for the content validity of the 50 items representing the behavioral focus, natural reward, and creating constructive thought pattern dimensions of the SLQ. The 40 items of the prototype SLQ which did not meet the 60% criterion for item assignment by the judges were deleted from the SLQ and were not used when we proceeded with Study 2.
.

Study 2

The primary objective of Study 2 was to extend our investigation of SLQ construct validity by collecting data from a sample of respondents to factor analyze the items in the refined SLQ. Further, our intent was to compare the factor analytically derived subscales of the SLQ to the theoretical components of the three main categories of self-leadership (cf. Schwab, 1980). The issue for construct validity is whether the constructs which have been empirically identified through factor analysis correspond to the theoretical constructs intended by the instrument developers. If items representing one theoretically specified dimension or sub-dimension cluster together, for example, this fits the instrument developers expectation in properly representing the construct. If, however, several items that appear to measure the same thing clustered on different factors (or items apparently measuring very different things clustered together on the same factor), this might raise questions about the validity of the construct measured by this collection of items (Crocker & Algina, 1986, p. 232). Study 2 compares the empirically derived factors of self-leadership with the theoretical components of the three main categories of self-leadership presented earlier. Method

Sample and Procedure


sample was composed of 194 students attending a junior-level introductory management course at another large university in the southwest. Mean age for the sample was 22.7 years, 80% were white, and 45% were female. Students volunteered for the research project in exchange for 1 % course credit to be applied to their grades in the course. The student respondents were given a preliminary lecturette on the importance of research in a major university and their obligation to treat the project seriously in exchange for the extra course credit. Further, respondents received instructions for answering the SLQ survey, and were given up to one hour to complete the exercise. All respondents completed the survey well within the allotted time.
The

Respondents were given the refined SLQ containing 50 items, and questions assessing demographic variables such as age, gender, and ethnicity. They were instructed to read each of the SLQ items carefully and &dquo;try to decide how true
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

129

the statement is in describing you.&dquo; Respondents circled numbers on a five point likert scale corresponding to their chosen answers: 1 = Not at all Accurate; 2 = Somewhat Accurate; 3 = A Little Accurate; 4 = Mostly Accurate; and 5 = Completely Accurate.

Analysis respondents who volunteered for the exercise, 189 returned useful, completed surveys used in this analysis. Factor analysis was used to evaluate
Of the 194 further the extent to which the SLQ items assessed the theoretical constructs of behavioral focused, natural reward focused, and constructive thought pattern focused self-leadership skills (Cf. Kinicki & Latack, 1990, p. 350). Exploratory factor analysis with principal components extraction and varimax rotation was applied to the SLQ. A cut off point of .35 was chosen as the critical value for deciding whether an item defined a factor. Further, the eigenvalue greater than one test, and the scree test were also used to define factors (Gorsuch, 1974). Table 2 presents the factor structure of the refined SLQ derived from factor analysis, showing the factors and their component items, the factor loadings, scale reliabilities (coefficient alpha), and eigenvalues. Twelve initial factors were derived in this analysis, however examination of the substantive meanings of the components dictated elimination of two factors (&dquo;Task Satisfaction&dquo; and &dquo;Not Coherent&dquo;) due to the incoherence of the factor or items measuring inconsistent dimensions (Cavusgil & Zou, 1994). In addition, both these factors showed poor internal consistency estimates, with coefficient alphas (.56 and .54, respectively) far below the acceptable level for basic research (.70) suggested by Nunnally (1978). Items composing these two factors were deleted from further analyses.
)
.

Results and Discussion

On the basis of the analysis above, the data were represented by ten underlying factors. Behavioral focused strategies of self-leadership were represented by six factors; natural reward strategies were represented by one factor; and Constructive thought pattern strategies were represented by three factors. Refer to table 3 for the factors representing the main categories of self-leadership. Behavioral focused strategies of self-leadership were represented by six factors. Self Goal-Setting (factor 2) was composed of seven items (alpha = .85)

specifically addressing issues of setting goals: mentally setting goals, writing down goals, and thinking about personal goals for performance. Self-Reward (factor 4) was represented by three items (alpha .91) directly focusing on self reward behaviors for good performance.
=

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

130

0
1&

1
E
a
.

c 0 #

I a
a

2: It:

12
~q N

4)
m

Q H

(a

8
.
=1

o
n

15

~
t-

LL

1 as

~ -e
$ i
a
Z
0

CO

.%

~ 1
ua
<

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

131

0)

i co
M
E
S

a
cm

m 0
0

I
11

L?

1.9
I

CC

li

CO

I
I

9
Urr

<

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

132

Ii
1.

F-

f
0
H

I 1
I
I

I
11

.t

I
4 d

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

133

rn

cm

~
co

0)

E
0 m

a
cm
s

0 0

ts
2
u
c

U
cm
c

16
11

U
U i6
0

.cm
0

m
w
ca

i
~

~o .m
Z
11 m

i6 .9? tm

0 2
cn
a

a~
rn

0 0

y 0
.:;
c~

s m
11

<

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

134

(I)

1
E
a

i
cm

.. E

0 ~ 9
>

-09
c

C.)

~
a
11

I
j
G)

L)

I
1!
at

I
11

S
Z
m

if

1k
I

12
fit

<
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

135

CO)

N
0)

co

Cl)
0

cv n-

ij>

H
N

9 7
To

0 0 U
0)

c
41)
U
11

U
0 i6

F s

~
72
t0

W-

Ta

To
Z
li m

d
cm

a>

0
s 0
.>

0
a~ c~
u

(D ~ m
<

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

136

0) l~ II

o
0)

fl
M
C

&oelig;

.2 T5
N cr %

cm

0 N
0) $$

~
:2 i
o

:&scaron; j o u

a) m a
,

m
cm
L

e
O

0
0

= E
O
M

H
0
16
2 H
0 c

s 0. s
In

U
cm C
cv

&oelig; c:

v as jo

2
0
11

~ g
% Q) 2
&dquo;

t0 0 cm

o
flh
C

0 ca

m
&dquo;D
t0

a~
L
~5
z
11 m S ;6

#
as

.~

8 c:
as

3
c 0

im

E E

8
0

Ei
a

~
W
o
Z

0
.

Q. N

Q)

S
0 m J5 a)

g~
co

m
11

<

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

137

TABLE 3 Sub-scales of the

Self-Leadership Questionnaire.

The Self Punishment scale (factor 6) was composed of four items (alpha = .75) primarily addressing self displeasure and responses to poor personal performance. Self-Observation (factor 7) was represented by five items (alpha = .73) dealing with paying attention to and keeping track of personal performance. Self-Cuing (factor 9) comprised two items (alpha = .82) focusing on respondents use of notes and lists for keeping track of what they need to accomplish. Finally, factor 10 was interpreted as Self Withholding, represented by two items (alpha = .76) specifically dealing with withholding positive rewards when performance was poor. Each of these scales corresponded to components representing behavioral focused strategies for self-leadership as presented in theoretical presentations of the construct (Manz, 1992b). Natural reward strategies of self-leadership have been theorized to fall into two main dimensions: natural rewards for tasks, and rewards for thinking style. The former focus on building enjoyable features into necessary tasks to make the work more enjoyable. The latter focus on identifying and concentrating on pleasant rather than unpleasant aspects of tasks (Manz, 1992a). This category of self-leadership was represented by only one factor in our analysis, which was interpreted as Focusing Thoughts on Natural Rewards (factor 8). This factor was defined by four SLQ items (alpha = .69) addressing both focusing thoughts on positive aspects of work and building enjoyable features into the work. While items composing this factor clearly represent natural reward strategies of selfleadership, the factor composition suggests that this category may be represented by one scale rather than two as previously theorized.
Constructive thought pattern strategies of self-leadership were represented three factors. Visualizing Successful Performance (factor 1) was represented

eight

(scale alpha .85) primarily addressing visualizing or successful performance on challenging tasks. Self-Talk (factor 3) was defined
items
=

by by imagining

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

138
,

by five SLQ items (scale alpha .84) focused on talking to oneself to get through difficulties. Finally, Evaluating Beliefs and Assumptions (factor 5) was represented by five items (alpha .79) specifically dealing with examining beliefs and assumptions in problem situations. Each of these factors has been theorized to represent components of this self-leadership domain (cf. Manz, 1992a, Neck & Manz, 1994).
= =

In sum, each of the ten meaningful, interpretable factors derived by factor analysis from the refined SLQ correspond to theoretical components of the major categories of self-leadership. Thus, empirical testing in this study, coupled with results from Study 1, further argues for construct validity of the refined SLQ instrument in measuring self-leadership skills, cognitions, and perceptions in our sample (cf. Crocker & Algina, 1986). To continue our investigation into the construct validity of the SLQ, we proceeded to Study 3.

Study 3 Study 3 had two objectives in extending our examination of construct validity of the SLQ. The first was to use the factor analytically derived self-leadership scales from Study 2 in a second sample to test scale reliabilities across samples. The second was to further investigate the psychometric properties of the SLQ. We examine the relationships among the SLQ scales to begin placing the SLQ in a nomological network of self-leadership (Cronbach & Meehl, 1955; Whitely, 1983). To begin to relate self-leadership to other theoretical variables, we examined the literature on self-leadership for important psychological and contextual factors connected with self-leadership. The existing literature is rich in theoretical, anecdotal, and experiential associations between use of selfleadership skills and positive performance. In addition, self-leadership has been related frequently in this literature to positive affect and self-confidence in those who use self-leadership behaviors, cognitions, and perceptions (e.g., Manz, 1986, 1990, 1992a, 1992b). However, there has been a paucity of empirical research upon which to base hypotheses about how the overall self-leadership construct is related to other psychological or contextual variables of interest (Cf. Neck & Manz, 1994; Prussia, Anderson & Manz, 1997), nor for that matter have we found any investigation into how self-leadership dimensions relate to each other. In an initial attempt to remedy this situation, we examine scale intercorrelations of the SLQ in this study.
.

Method

Sample and Procedure


~

obtained for this study from 155 students enrolled in three separate senior-level, undergraduate entrepreneurship courses at the same southwestern university as in Study 2. Respondents ages ranged from 20 to 49 years, and the average was 27. Sixty-six percent of the sample was male and 88% were seniors or graduates. Respondents averaged 9.3 years of work experience. The refined SLQ surveys as used in Study 2 and questions regarding demographic information were administered to students at the beginning of the semester in
Data
were
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

139

order to obtain information about respondents self-leadership strategies prior to in-class lectures and discussions of the concepts of self-leadership. Two of the classes were taught by the first author and the third was led by the second author. In order to ensure valid combination of the data sets, following data collection, t-tests indicated no significant differences in demographic variables existed between respondents in the two authors classes.

Analysis
Of the 155 entrepreneurship students surveyed, 153 returned completed and useful responses. Scale means, standard deviations, and reliability estimates were calculated for each of the valid factors derived through the factor analysis
in

Study 2. properties

Correlations among the SLQ scales were also calculated. The scale and intercorrelations are shown in table 4.

Results and Discussion One aspect

validity of an instrument is replicability. is Replicablity represented by a stable factor structure across different samples (Crocker & Algina, 1986). We examined the scale reliabilities between the samples in Study 2 and Study 3 to determine whether the factor structure was stable between the samples. In only one case did scale reliability drop significantly below Nunnallys (1978) recommendation of .70 for useful scale reliability. This was for the Focusing Thoughts on Natural Rewards scale, the only factor representing natural reward strategies of self-leadership. We concluded from this evidence that there may be an area of instability across samples for this dimension of the factor structure of the SLQ. contributing
to construct

in this

case

A further aspect of construct validation addresses the relationships among the subdimensions of the theoretical categories of the construct of interest. Subdimensions of each of the theoretical categories should be more closely related to other subdimensions of the same major category than to those of other categories. This finding suggests discriminant validity for the instrument in question (Crocker & Algina, 1986; Schmitt & Stults, 1986). To determine discriminant validity for the SLQ scales, correlations within each of the major categories of self-leadership should be larger than those across categories. For behavioral focused strategies, for example, we expect to see larger correlation coefficients among the scales representing that category than between those scales and scales representing other categories. Examination of table 4 for each of the major categories of self-leadership leads to the conclusion that the scales in many cases do not discriminate between behavioral focused, natural reward, or constructive thought pattern dimensions of self-leadership. Often correlation coefficients are larger across categories than within categories. While this result could be seen as an argument against the internal coherence of the SLQ, it may also be that the &dquo;three distinct, yet complementary dimensions of self-leadership&dquo; (Manz, 1992b) are more distinct in theory and more closely associated and complementary in the empirical. This study argues
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

140

Cl)
.

>1
, .

65
n

a
N a

-.0
0

ii m
~

6Hfl
~

ca

9 4)
i a c
K5
0

.2

2
a

0
M

00 q (t) Lt) VV ~Q,Q,


11 II 11

tf)-

c.

iii
t-

O
z

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

141

for that conclusion. Moderate to strong scale intercorrelations for several scales within and across dimensions suggest that people tend to use more than one self-leadership strategy in confronting their tasks. Discussion

separate studies provided preliminary support for the construct validity of the SLQ instrument to measure self-leadership behaviors, cognitions, and perceptions. Analysis by judges in Study 1 provided 50 valid items from the prototype instrument assessing behavior focused, natural reward, and constructive thought pattern strategies of self-leadership. Study 2 supported this evidence of content validity, providing factor analysis of data from a large sample to examine responses to the 50 refined SLQ items.
Ten

Three

meaningful, interpretable factors consistent with self-leadership theory merged from the factor analysis of the sample data. Estimates of internal consistency were acceptably high for the factors retained in Study 2. Behavioral focused strategies were represented by six factors, natural reward strategies were represented by one factor, and constructive thought pattern strategies were represented by three factors.
Study 3 used a separate large sample to investigate further the psychometric properties and intercorrelations of the SLQ. Reliability estimates for the scales, with only one exception, were acceptable and this finding argued for a moderate to good level of replicability for the instrument. While examination of the scale intercorrelation matrix disclosed poor discrimination among the categories of self-leadership, this was interpreted as an argument that the major theoretical categories of self-leadership may be less distinct empirically than theoretically.
Scale intercorrelations within and across dimensions demonstrated that respondents used multiple self-leadership strategies when dealing with challenges in their school and work lives.

Despite contributions of the present study to developing a valid measure of selfleadership, three limitations must be noted. The first limitation pertains to the generalizability of the study results. Since both of our samples were undergraduate students, there may be characteristics of the samples that differentiate them from other samples, and these differences may limit generalizability. Nevertheless, to the extent that our student samples represent other samples, results should be generalizable to other groups. The second limitation of the present study is the size of the sample used in the factor analysis in Study 2. While the sample size is acceptably high for most analyses, it has been argued by some methodologists that exploratory factor analysis of the type used here should attempt wherever possible to conform to the ratio of 10 respondents per survey item being factor analyzed. While many researchers question the importance of conforming to the 10-to-1 ratio, this may limit the stability of the factor structure proposed in this research. The third limitation regards the measure of Focusing Thoughts on Natural Rewards, which fell below the recommended level of scale reliability (.70). The low level of reliability may possibly weaken results, since the scales reliability represents its maximum
Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

142

possible correlation with other measured variables (Nunnally, 1978). Thus, reported correlations between this scale and others should be regarded with
caution.
In conclusion, several authors argue that employee empowerment and the related issue of employee self-leadership are expected to be recurring themes in organizational restructuring for the foreseeable future (Alvesson & Wilmot, 1992; Manz, 1992a). Further research efforts should be aimed in two directions. First, we suggest continuing refinement and development of scales to measure self-leadership, including the SLQ reported here, particularly with samples of 500 or more to respond to concerns regarding a stable factor structure for the instrument and the construct it represents. Only with valid instruments can we proceed to understand the relationships between self-leadership behaviors, cognitions and perceptions and other pivotal variables of interest.

Second, the studies reported here are a preliminary step toward validation and use of measurements of self-leadership. Further research is needed to place self-leadership in a nomothetic network of psychological and contextual variables (Cronbach & Meehl, 1955; Whitely, 1983). Empirical research should be directed toward the relationship between self-leadership and other psychological variables such as self-esteem, mastery, self-monitoring, and other traits. In addition, self-leadership should be examined further as it relates to self-efficacy and confidence, for example, psychological variables which are task or domain specific (Cf. Prussia, Anderson, & Manz, 1997). Further, determinants of self-leadership should be examined, better to develop training methods to enhance individuals self-leadership skills. Finally, the relationship between selfleadership and performance measures should be investigated further. With the results of this suggested research agenda, the process and results of selfleadership will become better understood, and will surely contribute to improved implementation of employee empowerment programs in organizations.
With

generation of more practical knowledge of the relationships between selfleadership and empowerment, we assist in the survival of organizations in the 21 st Century.

References
Aktouf, O. (1992). Management and theories
of organizations in the 1990s: Toward a critical radical humanism? The Academy of

(1993). From patemalism 45-48. stewardship. Training, 30,


Block, K.

to

3, 407-431. Management Review, 17,


Alvesson, M., & Wilmot, H. (1992). On the

Brower, M. (1994). Implementing TQM with self-directed teams. In Costin, H. (ed).

Readings

in Total

Quality Management.

emancipation in management and f organization studies. The Academy o , 3, 407-431. Management Review, 17
idea of
American

Harcourt, Brace: Orlando, FL. pp. 403-420.

Byme, J. (1993, February 15). Requiem for yesterdays CEO. Business Week, p. 32-33.

Psychological Association. (1985). Standards for educational and psychological , Washington DC. testing

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

143

Cavusgil, S. & Zou, S. (1994). Marketing An strategy-performance relationship: investigation of the empirical link in export market ventures. Journal of Marketing, 58, 121.

Manz, C. (1992a). Self-leadership ... the heart of empowerment. Journal for Quality and

Participation, July/August,

80-85.

Manz, C. (1992b). Mastering Self-Leadership:

Empowering yourself
Kanungo, R. (1988). The empowerment process: Integrating theory and practice. The Academy of Management

for

personal
NJ. Prentice

Conger, J.,

&

excellence. Hall.

Englewood Cliffs,

Review, 13 , 639-652.

Manz, C., & Sims, H. (1991). Superleadership:

Campbell, D. (1979). Quasiexperimental design: Design and analvsis issues for field settings, Boston, MA: Houghton
Cook, T., &
Mifflin Co.

Beyond the myth of heroic leadership. Organizational Dynamics, 19 , 18-35.


Neck, C., & Manz, C. (1994). Thought self-

Crocker, L., & Algina, J. (1986). Introduction to


classical and modern test theory, New York: Holt, Rinehart & Winston.

leadership : The impact of mental strategies training on employee cognition, behavior and affect. (currently under review). Nunnally, J. (1978). York: McGraw-Hill Psychometric theory. New Book Company.

Gorsuch,

Philadelphia,

R. (1974). Factor PA: Saunders.

Analysis.
Prussia, G., Anderson, J., & Manz, C. (1997). Self-leadership and performance outcomes: The mediating effect of self-efficacy. Journal of Organizational Behavior (in press).

Kinicki, A., Heyl, J., & Callerman, T. (1986). Assessing the validity of the Cox, Zmud, and Clark material-requirements-planning audit instrument. Academy of Management Journal,

, 3, 633-641. 29
Kinicki, A., & Latack, J. (1990). Explication of the construct of coping with job loss. Journal of Vocational Behavior, 36, 339-360. Kozlowski, S., Chao, G., Smith, E., & Hedlund,
J.

Roberts, M. (1994). Becoming customer oriented. In Costin, H. (ed). Readings in Total Quality Management. Harcourt Brace: Orlando FL. pp.305-319.

Schmitt, N., & Stults, D. (1986). Methodology


review: Analysis of Multitrait-multimethod matrices. Applied Pgychological Measurement, , 1-22. 10

(1993). Organizational downsizing: Strategies, interventions, and research implications. International Review of Industrial and Organizational Psychology, New York: John Wiley & Sons. Frayne, C. (1989). Selfmanagement training for increasing job attendance: A follow-up and replication. Journal of Applied Psycholoqy, 74 , 411-416.
Latham, G., &

Schwab, D. (1980). Construct validity in organizational behavior. In Staw, B., & in Research Cummings, L. (Eds.), CT: behavior Greenwich, (3-43). organizational
JAI Press.

Improving performance through self-leadership. National Productivity


Review,
summer, 297.

Manz, C. (1983).

Shipper, F., & Manz, C. (1992). Employee selfmanagement without formally designated teams: An alternative road to empowerment. Organizational Dynamics, 20, 3, 48-61.
Thomas, K., & Velthouse, B. (1990). Cognitive
elements of empowerment: An interpretive The model of intrinsic task motivation. , 666-681. 15 Academy of Management Review,

(1986). Self-leadership: Toward an expanded theory of self-influence processes in organizations. The Academy of Management Review, 11 , 3, 585-600.
Manz, C. (1990). Beyond self-managing work teams: Toward self-leading teams in the workplace. In R. Woodman & W. Pasmore (Eds.), Research in Oraanizational Change and

Manz, C.

Whetten, D., &


Collins : New York.

Cameron,

K.

(1995).

DevelopingManagement Skills.

Harper-

Development, Greenwich, CT: JAI Press.

Whitely, S. (1983). Construct validity: Construct representation versus nomothetic span. 3, 179-197. Psychological Bulletin, 9

Downloaded from jlo.sagepub.com at Universiti Tun Hussein Onn Malaysia on November 28, 2013

You might also like