Cohen (2009) claimed that ethics and integrity are essential for leadership but although ethics and integrity may not necessarily ensure effective leadership, but their absence may prevent leaders from becoming effective. The study by Abu-Tineh et.al (2008) on Amman school principals in Jordan using Kouzes and Posner's LPI revealed that most of the school teachers in Amman identified their principals as transformational which is attributed to effective leadership. The finding clearly supports the claim by Parry & Proctor-Thomson (2002) that followers' perception of their leaders as transformational are closely related to their perception of leader's integrity. For Kouzes and Posner model of leadership practices, Taylor (2002) acknowledged that it has been recognized by many researchers as truly representative of highly effective leadership practices.
The previous study by Morgan (1993), Posner and Schmidt (1984) indicates that leaders are the primary influence on the ethical behaviour of their subordinates. They claimed that leaders who practice ethical behaviour is likely to impact the ethical behaviours of others in the organization. Simons (1999) by citing Bass (1990); Carlson and Perrewe (1995); Kouzes and Posner (1992) also argued that for subordinates to have trust in their leaders, it requires the "follower to perceive the leader as honest or as possessing integrity". So, this indicates that leadership needs to have integrity to influence the ethical behaviour of subordinates. For Palanski and Yammarino (2007) cited in Hooijberg et al (2010), they claimed that the role of integrity plays in leadership effectiveness has not been examined empirically while Morrison (2001) asserted that integrity is necessary for effective leadership not only in organizations but also in the global context.
Based on the above statements and other previous literatures on leader effectiveness and leader integrity, this study intends to investigate the direction of the relationship between perceived leader effectiveness and perceived leader's integrity in the context of Sabah state's agencies in Kota Kinabalu. The following schematic framework explains perceived leader effectiveness as independent variable and perceived leader's integrity as dependent variable. The relationship of these two variables is moderated by demographic variables such as age, gender and education.
Figure 1 : Theoretical Framework and Hypothesis
Perceived leader's Integrity
(PLIS)
Perceived leader effectiveness
(LPI)
Model the Way
Inspired Vision
Challenge status-quo
Enable others to act
Encourage the heart
Age
Gender
Education
Independent Variables Moderating Variables Dependent Variable
Definition of Variables
Independent Variable
Kouzes and Posner (1995, 2002) asserted that effective leaders are determined not by their position but by their collection of behaviours and practices. Ultimately leader's behaviours and practices will lead them to inspire greater accomplishment and achieve extraordinary results beyond what it is today. Taylor in 2002 and Parry and Proctor-Thomson also in 2002 in studying Kouzes and Posner's leadership behaviours asserted that the five dimensions of leadership behaviours: 'Model the Way', 'Inspire Vision', 'Challenge the process', 'Enable others to act' and 'Encourage the heart' which Kouzes and Posner developed through extensive research of almost 20 years have been recognised by many researchers as truly demonstrating an effective leadership practices.
Using the five dimensions model of effective leader's practices, Kouzes and Posner (1995, 2002) defined them as:
Model the Way,
Model the Way is a set of behaviours that have to do with setting an example or role-model of how to treat others and how to pursue goals in a respectful manner. The leader must lead the way specifically implies that leader must go first by showing an example that they live by the virtue of their words and actions (Kouzes and Posner, 1995, 2002).
Inspire Vision,
Inspire a Shared Vision labels the need for leaders to aspire a shared vision and help others to see that vision which is an idealized collective future. This will create enthusiasm and excitement among employees and followers to work towards positive outlook through sharing of vision and future of the organization (Kouzes and Posner, 1995, 2002).
Challenge the Process,
Challenge the Process is about taking risks, leader's ability to seeing failures as learning opportunities, and finding new and better ways to accomplish goals. It involves creating new ideas and finding new ways of doing things in order to achieve extraordinary results (Kouzes and Posner, 1995, 2002)
Enable Others to Act,
Enable Others to Act is like an empowerment and leader's behaviour that create an environment of trust and mutual respect that will in turn foster collaboration and team work. It also means giving employees the right to make decision especially those of front-line employees. (Kouzes and Posner, 1995, 2002)
(e) Encourage the Heart.
Encourage the Heart includes those behaviours that recognize the contributions of others, celebrate accomplishments, and share rewards as well as motivations to strive harder for the accomplishment of goals and mission of the organization. By giving incentives, rewards and other sorts of motivations and encouragement, this will contribute to the employees and followers to perform better (Kouzes and Posner, 1995, 2002).
Dependent Variable
Lawton (2010) defined integrity as referring to principle values and norms in public and private organizations, and this definition is basically similar to definition by Becker (1998) as stated previously. Other researchers defined integrity in different forms and in different manner according to the line of their research. Some researchers related it to matching words with deeds and actions, compliance with rules and laws, commitment to organizational values, but most of them linked integrity to moral values such as honesty, trust, and ethics that are mostly expressed and as similar to integrity and have been applied interchangeably in past literatures.
Peterson (2002) in his study on Perceived leader integrity and ethical intentions of subordinates quoted Sims and Brinkmann (2002) assertion that integrity in organization's leader account for much of the differences in the observed unethical behaviours among various organizations. Peterson also argued that the integrity of the leader would most likely to influence any unethical acts directed towards the organization. For Paine (1997); Sims and Brinkman (2002) also agreed that leaders with integrity will represent as a role-model and can be accounted in establishing the norms and culture as to how subordinates must be treated and taken care of. The statements above underpin how leaders with integrity are important in influencing subordinates' behaviours and establishing the climate of ethics in organizations.
In this study, leader perceived integrity will be assessed using Craig and Gustafson's (1998) instrument of Perceived Leader Integrity Scale or PLIS which has been successful in measuring leader's perceived integrity (Hogan and Kaiser, 2010) which according to Hogan and Kaiser, ratings on the PLIS do not appear to be distorted by impression management and no socially desirable manner will likely to influence the ratings. Craig and Gustafson (1998) stated that the scale was used to assess the respondents' perceptions of the integrity of their leaders. They also asserted that prior research has established that the 32-item PLIS scale has an acceptable level of validity and reliability.
Moderating Variables
Ford and Richardson (1994) hinted that previous studies regarded age, gender, and education level may be related to ethical attitudes of leaders. Hence, this study will examine if age, gender and education have any effect on the relationship between perceived leader effectiveness and perceived leader's integrity. This is also to see if subordinates' perception of leader effectiveness and integrity varies across specific demographics variables (age, gender, education).
Age
Age is an individual chronological age. Previous study indicated that age affects perception of leader effectiveness (Bass, 1990) cited in Hooijberg et. al (2010).
Gender
Leader's gender can be male or female. Previous researchers have demonstrated that gender (Dobbins & Platz, 1986; Eagly & Johnson, 1990) affects perception of leader effectiveness. Thus, this study will predict if gender factor represents one of the attributors to describe leader effectiveness as different people will give different responses.
Education
Different level of education is predicted in this study to affect the level of relationship between perceived leader effectiveness and perceived leader's integrity. Previous study indicated that education affected perception of leader effectiveness (Bass, 1990) cited in Hooijberg et. al (2010). The author predicts that senior ranking personnel with proper educational level would most likely to assess their leaders behaviours and integrity more appropriately and meaningfully.
Research Hypotheses
The research hypotheses are developed from the theoretical framework in figure 1 to examine the relationship between independent, moderating and dependent variables. The research hypotheses to be tested are as follows:
NH 1 : There is no significant relationship between overall LPI and overall PLIS.
AH 1 : There is significant relationship between overall LPI and overall PLIS.
NH 3 : There is no significant relationship between model the way and PLIS.
AH 3 : There is significant relationship between model the way and PLIS.
NH 2 : There is no significant relationship between inspire vision and PLIS.
AH 2 : There is significant relationship between inspire vision and PLIS.
NH 4 : There is no significant relationship between challenge the process and PLIS.
AH 4 : There is significant relationship between challenge the process and PLIS.
NH 5 : There is no significant relationship between Enable others to act and PLIS.
AH 5 : There is significant relationship between enable others to act and PLIS.
NH 6 : There is no significant relationship between Encourage the heart and PLIS.
AH 6 : There is significant relationship between Encourage the heart and PLIS.
NH 7 : The relationship between perceived leader effectiveness and perceived
leader integrity is not moderated by age.
AH 7 : The relationship between perceived leader effectiveness and perceived leader
integrity is moderated by age.
NH 8 : The relationship between perceived leader effectiveness and perceived leader
integrity is not moderated by gender.
AH 8 : The relationship between perceived leader effectiveness and perceived
leader integrity is moderated by gender.
NH 9 : The relationship between perceived leader effectiveness and perceived leader
integrity is not moderated by education.
AH 9 : The relationship between perceived leader effectiveness and perceived leader
integrity is moderated by education.
Research Design
This study constitutes a field study with hypothesis testing method articulated to establish a correlational relationship between independent variable (perceived leader effectiveness) and dependent variable (perceived leader integrity) in wanting to understand to what extend leadership must be seen and perceived to be effective and possess integrity and the nature of relationship between perceive leader effectiveness and perceived leader integrity. Hence, correlation analysis was used to test the existence and strength of relationship of each dimension of effective leadership and PLIS score. The author's interference in this study was minimal as the author has not interfered with normal activities in all the identified ministries and departments in this study. In so far as subordinates' rating was concerned, they were assured of anonymity and confidentiality to avoid future backlash if the respective leaders felt less comfortable with the rating scores and no information whatsoever was presented to the respective heads of departments and ministries about the subordinates' rating to preserve present goodwill and climate of understanding and to ensure subordinates to give their ratings fairly, appropriately and justifiably. After all, this process needed to be implemented with care.
This study was intended to examine subordinates' perception of leader effectiveness in relation to leader perceived integrity which this study hopes could be used later on to address the issues of integrity in Sabah's agencies in Kota Kinabalu. The rating of perceived effectiveness and integrity of leaders who comprise of heads of ministries and departments in categories Jusa B, C, Grade 54 and 48 was carried out by mean of bottom-up approach and not at organisational level as a whole or below that of the rater. Therefore, the target population in this study was those categories of Sabah civil service personnel from Grade 27 to Grade 44 and how they assessed their respective heads of departments as leaders in their organizations pertaining to perceived effectiveness and integrity using the LPI and PLIS measurement score. Thus, the methodology used in this study was by means of quantitative technique using structured questionnaire to conduct the research.
The perceived leader effectiveness was measured using Leadership Practices Inventory-subordinates' assessment by Kouzes and Posner (1995, 2002) which is a 30-item 10-Likert-scale questionnaire comprising all the six dimensions of effective leadership practices, that is six items for one dimension : Model the Way (items 1,6,11,16,21 & 26), Inspire vision (items 2,7,12,17,22 & 27), Challenging the Process (items 3,8,13,18,23 & 28), Enable Others to Act (items 4,9,14,19,24 & 29) and Encourage the Heart (items 5,10,15,20,25 & 30).
The perceived leader integrity was measured using the Perceived Leader Integrity Scale (PLIS) which was a reviewed one and it contained 30-items questions which was different from the original PLIS of 32-items developed by Craig and Gustafson (1998). The author felt that modification needed to be done on the original PLIS involving two items in the questionnaire: item number 7 (Is evil) and 18 (Is a hypocrite) were omitted as it reflected high bad impression on leadership in public organizations and could be ill-perceived by leadership. The PLIS was reverse-scored to demonstrate that higher point will indicate leader's high perceived integrity and likewise through response alternatives of 5-point likert scale from 1 - strongly agree to 5 - strongly disagree and respondents have to select the specific point indicated to represent a definite assessment of leader's integrity. Rating was done by subordinates to measure the perceived integrity of leaders as suggested by Hogan and Kaiser (2010) to be the best source of information about leader's integrity although Parry and Proctor-Thomson (2002) in their study indicated that there was evidence to suggest that ratings of leaders by subordinates might be different from ratings of the same leader by peers or super-ordinate managers. However, in this study, subordinates' rating has been used to assess leaders' integrity within the context of Sabah civil service. It was hoped that the use of both LPI and PLIS in this study would be able to provide empirical evidence relating to some theoretical considerations concerning the integrity of effective leadership.
A questionnaire of five pages was designed for this study and to avoid misinterpretation of the questions by the respondents, the author has it written bilingually, both in English and Bahasa Melayu. To confirm that the author was indeed currently this study and as a support for approval in getting required information from relevant parties, a letter of confirmation from the Dean of Business and Economics School has been obtained (in Appendix A). In addition to that, a letter of support has also been requested from the State Secretary's Office to ensure high response and due cooperation from respondents in answering the questionnaire (Appendix B). After the required data has been collected, the statistical package for social sciences (SPSS) was used to analyse the data.
Instrument Design
A combined structured questionnaire developed by Kouzes and Posner (2002) and Craig and Gustafson (1998) has been employed as one set of questionnaire for data collection in supporting the quantitative method used in this study. The general questions part (Section A) of the questionnaire was used to gather personal demographic characteristics of respondents of the sample population. The second Part (Section B) consists of the Leadership Practices Inventory (LPI) with 30 descriptive items questions (6 items for each leader's practice) for five different leadership practices to appraise leader behaviours by subordinates. The responses were based on a 10-point Likert scale ranging from 1 (almost never) to 10 (almost always). In the third part of the questionnaire (Section C), a 30-items questionnaire (reviewed) measuring the leader's perceived integrity was adopted from Craig and Gustafson (1998) using five-point likert scale ranging from 1 - strongly agree to 5 - strongly disagree. Most of items asked the respondents to assess the likelihood that leader would engage in unethical behaviours.
The questionnaires have been despatched to the respective heads of departments and ministries accompanied by letter from the State Secretary's Office. Upon completion in answering the questionnaires by respective ministries and departments, the author himself collected them all to ensure that all questionnaires were duly answered and collected for statistical analysis.
3.7 Unit of Analysis
The unit of analysis in this study was the career civil servants from Grade 27 up to Grade 44 in all the state's 10 ministries, the Chief Minister Department, and the other 21 departments in Kota Kinabalu as well. In this study, data was collected and gathered from each individual's civil servants and was aggregated according to each grade so as to see the pattern and differences among the grades of civil servants in perceiving their leaders effectiveness and integrity.
3.8 Sampling Technique
This study used probability sampling technique as the identity of respondents was known based on their grades of position as contained in the Sabah's Estimate of Emolument 2010 which was the identified sampling frame and they have a known chance of being chosen as subjects in the sample which could represent generalizability of the population parameters. This research gave focus on career civil servants in Grade 27 up to Grade 44 only totalling 820 in number distributed according to the ministries, the Chief Minister Department, and other departments in Kota Kinabalu which constituted the population parameters. The specific reason for the selection of these ategories of civil servants was because they were the ones that were directly affected and impacted by leader's behaviours and integrity which according to Parry and Proctor-Thomson (2002) followers were the direct recipients of leader's behaviours and thus they were in a position to observe leader behaviour firsthand. They were also actively involved in any organization's management meeting compared with those of lower categories and hence, their rating was of particular importance pertaining to leader's perceived effectiveness and integrity as they were in direct position to observe leaders' behaviours. In other words, they were relevant, appropriate and meaningful in the context of this study.
These categories of civil servants were considered to be of middle and second-upper management according to present understanding of ranking in the state administration. Any category of civil servants from Grade 48 and above was regarded to represent upper management which directly involves in policy planning and decision-making while those categories of lower than Grade 27 is what was normally referred to as the Chief Clerk and personnel below that in the opinion of the author did not represent high impact subjects in this study as their involvement and interaction with heads of ministries and departments was lesser than those of Grade 27 and above. In such a scenario, their knowledge of leader's behaviour is somehow limited and depended upon descriptions and inferences by their next superior. The Grade 27 was the cut-off point or previously referred to as Scale B11 under the Cabinet Committee Report (CCR) 1970 before the introduction of the present new remuneration system. In a more specific term previously used, it is a division B personnel and any grade below it is the division C which comprised of clerical and technical assistant staff and subsequently the clerical assistant and all the lower level of personnel. Hence, their perception in this study is not relevant and meaningful as they are responsible to their Grade 27 personnel and above. In other words, this study only focuses on perceived leader effectiveness and integrity at the higher management level, that is, Grade Jusa B,C, Grade 54 and 48 in respective ministries and departments by the next level of direct subordinates.
This study also did not intend to involve heads of departments and permanent secretaries of all ministries to participate in the survey questionnaire by way of leader's self-assessment as previous research has indicated that self-assessments might tend to inflate abilities by significant amounts (Atwater, Rousch, & Fischthal, 1995; Morgan, 1993; Randall, Ferguson, & Patterson, 2000). This statement was in tandem with suggestion by Bem, 1967; Nisbett et al. (1982) that self-report was usually not an effective method for identifying the factors that influence person's behaviours. The same thing might occur when superiors assess the leadership, the scores also might tend to be inflated compared to those of peers (Parry & Proctor-Thomson, 2002). Of course, in this study, the participation of the State Secretary as the highest administrator of the Sabah civil service was not duly required as this study only intended to assess subordinates' assessment of leaders' effectiveness and integrity and not superordinate.
The table 3.1. below describes the distribution of categories of civil servants according to their grades of position based on Sabah's Estimate of Emolument 2010 (which does not include statutory bodies):-
Table: 3.1. Distribution of Grades of Position in Sabah state's ministries and departments in Kota Kinabalu as contained in Sabah's Estimate of Emolument 2010
No.
Grade of positions
Div.
Chief Minister Dept.
All
Ministries
Other Depts.
Total
Proportionate
Sampling
(25% of the elements)
Using
Dispro-
portionate Sampling
1.
Grade 27/29
B
164
60
81
305
76
50
2.
Grade 32
52
31
19
102
26
30
3.
Grade 36/38
20
2
25
47
12
4.
Grade 41
A
137
69
80
286
72
86
5.
Grade 44
30
14
36
80
20
40
Total =
403
176
241
820
206
206
Based on their total number of 820 which is the defined population parameters indicated through the Sabah's Estimate of Emolument 2010 as the sampling frame, this study will apply a sample size of 206 which is also consistent with Table 10.3 for Sample size for a given population size (Appendix C) designed by Krejcie and Morgan (1970) quoted in Sekaran and Bougie (2009). This sample size of 206 respondents will asses all the 21 heads of departments and 11 ministries including the Chief Minister Department in Kota Kinabalu only. Looking at the table above, there appears to be uneven distribution of elements in the sample and the situation becomes less favourable if using proportionate sampling of 25% of the elements where for Grade 44 which is the highest personnel in this study, only 20 samples will be collected compared to Grade 27 which is the lowest for 76 samples. This is to the author's opinion is not reflective of justified subordinates' rating of leader's perceived effectiveness and integrity as superior personnel is in a better position to rate their leader's perceived effectiveness and integrity due to their direct interactions and experiences coupled with level of education and responsibilities.
Consistent with requirement of homogeneity in stratified sampling of each grade, the author wished to apply the disproportionate sampling method the distribution of which is reflected in table 3.1. to reflect generalizability of all the identified grades of civil servants and is therefore representative of the sample. Of course, the more the sample size, the better.
The sampling technique relies on the fact that the selected sample is the convenience and appropriate contributory factors to enable investigation into leader's perceived effectiveness and integrity as direct connection really exists between leadership and the targeted respondents in the execution of daily routine activities of the day. These routine activities represent the type of quality service delivery that hinges on the interconnection of the respondents and the leadership. The leadership establishes all the necessary policies, guidance and direction and the respondents executive the activities that deliver the service to the clients and the people. As far as public perception is concerned, the ministries and departments selected for this study truly represents the main agencies where perceived leader effectiveness and integrity can be carried out.
Data collection procedure
The collection of information was drawn from all the 10 ministries plus the Chief Minister's Department, other 21 departments located in Kota Kinabalu, Sabah. The participants or respondents in this study were mostly career civil servants who were not politically-appointed. A structured questionnaire was employed for the respondents to respond. First, a pilot study using the questionnaire was also conducted with a sample of 50 UMS MBA students taking BZ8006/8007 who were fully employed. Then, a subsequent data collection was initiated where a total of 290 sets of questionnaires made up 35% of the population were sent out by the author himself to all the ministries and departments in Kota Kinabalu so as to generate a sample of 206 in which the author predicted a response rate of 25% of useable samples to be gathered and analyzed in this study. A Chief Clerk attached to each ministries and department was authorized to administer personally the questionnaires and to oversee the distribution and completion of the questionnaires within the 5 days time frame given to each sample for the Chief Clerk to deliver the questionnaires back to the author. The author believes this would save time and energy as well as money used to travel to all the ministries and departments to collect them.
3.10 Data Analysis
The statistical package for social sciences (SPSS) was applied in this study for data analysis. Validated items were used to measure the variables in the construct. Descriptive statistics was also used to obtain frequency distribution of respondents according to age, gender and education; the min, max, mean and variance of independent and dependent variables. This was followed by a reliability analysis to evaluate the reliability of all the dimensions by using Cronbach's Alpha. This is to ensure that all the variables used in this study are reliable for analysis under the conditions of a local survey.
Factor analysis was also conducted to determine the factors that respondents identify as comprising leader's effectiveness. Then, the result of the factor analysis was used as a proxy in hierarchical regression analysis probing the relationship between perceived leader effectiveness and perceived leader integrity. Correlation analysis was also applied to determine the strength of relationship between perceived leader effectiveness and integrity. Then, hierarchical multiple regression model (HMRM) was also applied to predict the effect of moderating variables on the relationship between perceived leader effectiveness and perceived leader integrity.
Table 3.2. Statistical Methods for hypothesis Testing
No.
Analysis
Statistical Method
1.
To test the frequency of age, gender and education of respondents and the max, min, means and variance of independent and dependent variables.
Descriptive statistics
2.
To test the reliability of measure
Cronbach's Alpha
3.
To determine the factors (inspired vision, model the way, enable others to act, challenge the process and encourage the heart) that respondents identify as comprising leader's effectiveness.
Factor analysis
4.
To test the existence and strength of relationship between each dimension of effective leadership with overall PLIS score.
Correlation analysis &Hierarchical Regression analysis
5.
To investigate the effect of leader's demographic factors on the relationship between effective leadership (independent variable) and perceived integrity (dependent variable).
Hierarchical Multiple Regression Analysis
