In Support Of An Emergent General Taxonomy Of Perceived Managerial And Leadership Effectiveness: An Empirical Study From The UK Independent Schools Sector
Refereed Paper
Hamlin, Robert G.; Barnett, Clare
Abstract
This paper presents the results of a replication study of perceived managerial and leadership effectiveness withinthe non-profit UK Independent Secondary Schools Sector. Using the Critical Incident Technique (CIT), concrete examples (critical incidents) of observed effective and ineffective managerial behaviour were collected from a purposive sample of participants from three independent Girls High Schools. The data obtained from each school was content analyzed to identify analytical categories for which behavioural statements werethen devised to reflect the constituent critical incidents in each category. These were subjected to a form of realist qualitative analysis to identify sector-related categories (behavioural indicators) of perceived managerial and leadership effectiveness. The paperalso reports a subsequent cross-sector comparative analysis which explored the similarities and differences existing between the deduced behavioural indicators and equivalentfindings resulting frompast studies carried out inother organizational sectors in the UK. The results support Hamlin, Sawyer and Sage (2011) who challenge the widely held view that to be effective managers in third (non-profit) and public sector organizationsneed to adopt different managerial behaviours to those in private (for-profit) organizations because of the inherent differences between sectors. Limitations of the study and implications for future research and practice are discussed.
Keywords: Perceived managerial and leadership effectiveness, behavioural indicators, general taxonomy:
Over the past 80 years or more there has been a plethora of research into the nature of management. Much is known about managerial functions, roles, and activities,but there is a dearth of information regarding managerial behavior. Furthermore, there have beenfew studies that“have empirically examined managerial behavior in the context of performance”(Borman & Brush, 1993, p.4). AsCammock, Nilakant and Dakin (1995, p. 443) claim.drawing on Hales (1986) and Stewart (1989), although past managerial behaviour studieshave provided important insights into the nature of managerial work, little has been done “which relates the activities and behaviors of managers to measures of effectiveness”. Similarly, Noordegraaf and Stewart (2000) claim a major shortcoming ofmanagerial work and behaviour research is the paucity of attention to managerial effectiveness. In addition these writers lament the dearth of studies usingthe ‘managerial behavior approach’, which they characterize as“inductive empirical research that explores managerial work and behaviour as it occurs in organizations with the aim of developing “categories, concepts and theories on the basis of empirical evidence” (Ibid, p. 429).
A contemporary researcher who has addressed,and continues to address the shortcomings outlined above using the ‘managerial behaviour approach’, is Author One (Hamlin). His original study explored the behavioural criteria of managerial effectiveness in UK secondary schools (Hamlin, 1988). And since that time he has conducted with various co-researchers a series of replication studies not only within a diverse range ofUK public, private, and third (non-profit) sectororganizations (see Hamlin, 2009), but also within public sector organizations in Egypt, Mexico and Romania (Hamlin, Nassar and Wahba, 2010; Hamlin, Ruiz and Wang, 2011; Patel, Hamlin and Iurac, 2010), and within private sector organizations in Germany(Patel, Hamlin and Heidgen, 2009). Our study builds upon and extends this past research by identifying the behavioural indicators of perceived managerial and leadership effectiveness withinthe UK Independent Secondary Schools Sector. The reason for choosing this sector is because Author Two (Barnett) had had previous access to such schools through her professional work as an external ‘executive coach’, and hadbeen able to secure the active interest and sponsorship of three Heads of School. As with Hamlin et al’s earlier replication studies thisinquiry has two aims. The first is to identifywhat people working within the respective collaborating schools perceive as effective and ineffective managerial behavior. And the second aim is to explore the extent to which these perceptions are similar or different to those identified by Hamlin et al in other types of organizational settings and sectors within the UK.
In conducting their research Hamlin et al followed Yukl (1989) by not separating the terms ‘leadership’, ‘leading’ and ‘leader’ from ‘management’, ‘managing’ and ‘manager’. Instead, as explained and justified in Hamlin (2009), they used the terms ‘managerial leader’ and‘managerial leadership’ to give recognition to the everyday task of ‘leading’ people as performed by most managers at all levels of management in most organisations. Thus, for their replication studies, and necessarily for ours, the term ‘managerial and leadership effectiveness’ embraces both ‘manager effectiveness’ and ‘managerial leader effectiveness’, but not ‘organizational leader effectiveness’
Conceptual Background
Past managerial and leadership behaviour studies
Various attempts were made in the 1950s to link the observed behaviours of managers with measures of their effectiveness (Flanagan, 1952; Guest, 1956; Jasinski, 1956; Kay, 1959; Ponder, 1957). Although differences were found between the managerial behaviour of effective and less effective managers and supervisors, the studies only revealed the degree of brevity, variety, fragmentation and interpersonal interaction that characterize managerial work. The issue of managerial effectiveness received further attention in the 1970s/1980s (See for example, Boyatzis, 1982; Burgoyne and Stuart, 1976; Latham and Wexley, 1977; Luthans, Rosencrantz and Hennessey, 1985; Luthans, Welsh and Taylor, 1988; Morse and Wagner, 1978). But most of these studies were similarly focused on the time devoted to, or the relative frequency of particular activities (O’Driscoll, Humphries, and Larson, 1991; Shipper, 1991), and failedto explore what differentiates highly effective from less effective managers (Borman and Brush, 1993; Hales, 1986; Martinko and Gardner, 1985, 1990). Stewart (1989) highlights the lack of comparability between these paststudies due to the ‘haphazard’, ‘arbitrary’ and confusing ‘mix’ of coding categories used, and argues researchers should free the mind from existing category codes and instead demonstrate the existence of broad categories of managerial behaviour as deduced from empirical evidence. Furthermore,she arguesthey should identify the ways individuals tend to be ineffective as well as effective, which is a viewechoed in the more recent writings of Amabile et al (2004) and Burke (2006).
In the field of leadership behaviour research most studies have been survey-based and focusedon a single level of analysis using pre-determined behavioural dimensions and questionnaire items. But asConger (1998) claims, these tend to be reductionist and ‘sterile’ in character inn order fro the instruments to be applied in all types of organizational settings. Consequently, researchers end up measuring the presence and frequency of ‘static terms’ which fail “to help us understand the deeper structures of leadership phenomena” (Ibid., p.109). Alvesson (2002) complains about the positivist bias in leadership research which favours laboratory experiments or questionnaire studies. And Yukl, Gordon and Taber (2002) express concern that despite the large amount of empirical research on effective leadership behaviours, there is still a “lack of agreement about which behavior categories are relevant and meaningful for leaders” (Ibid.p.15). And as these writers also claim, it is very difficult to compare and integrate one taxonomy with another when different terms have been used by various researchers to refer to the same type of behaviour, and when the same term has been defined differently by various theorists. In the absence of clear and unequivocal dimensions of managerial performance, including behavioral effectiveness criteria against whichperformance can be measured, we suggest managers/managerial leaders will operate and behave on the basis of their own individual personal preferences. This situation and the continuing lack of agreement on what differentiates behaviourally effective from ineffective managers/managerial leaders suggests further studies of perceived managerial and leadership effectiveness are warranted.
Another compelling reason to explorefurther the issue of what constitutes managerial and leadership effectiveness is the widely held belief that to be effective managers and leaders in public sector and third (non-profit) sector organizations need, or may need, to adopt different managerial behaviours to those in private sector companies because of the inherent differences between the sectors (See Baldwin, 1987; Fottler, 1981; Hooijberg and Choi, 1998; Peterson and Van Fleet, 2008). But as Hooijberg and Choi (1998) point out, few researchers have explored whether management/leadership styles do vary across the sectors. In sharp contrastto this ‘contingency’ view of management,various other writers believe in the ‘universality’ of management across organizations, organizational sectors and nations, and argue that logic suggesting the existence of universal/near universal leader behaviours and generic behavioural managerial competencies is compelling (See Bennis, 1999; House and Aditya, 1997; Woodruffe, 1992). These conflicting and controversial viewssuggestmore research is required to find out the extent to which the contextualized perceptions of managerial and leadership effectiveness ‘within’ and ‘across’ organizations and organizational sectors are the same or different.
Theoretical Framework
As previously mentioned, most past studiesof managerial effectiveness have used quantitative methods to explore the predictive power of managerial abilities, personality and behaviour to determine managerial performance. But the resultsshow less than 10 percent of variance as measured against external objectiveperformance standardscan be accounted for by these factors (See Tsui, 1984a; 1984b). Tsui (1990)argues that a different conceptualization of managerial effectiveness is required, andadvocates instead the use of the multiple-constituencymodel (MC)of organizational effectiveness. The central tenet across all variants of MC is that “an organization is effective to the extent it satisfies the interests of one or more constituencies associated with the organization” (Ibid., p. 458). She also argues that the social test of effectiveness at the unit level is a conceptual analogue to the MC approach at the organizational level. Similarly, we argue the social test at the behavioural level is a conceptual analogue to the MC approach at the subunit level. By using the MC approach, managers are perceived as operating within a social structure consisting of multiple constituencies or stakeholders (i.e. superiors, peers, subordinates) who have their own expectations of and reactions to the manager. Managerial behaviour is seen to be largely influenced by the desired behaviour expectations of the respective stakeholders. And theseexpectations determine the criteria by which a manager’s performance and reputation are perceived and judged. Tsui (1984a) refers to such perceptual consensus as managerial reputational effectiveness, and suggests with Ashford that the type of managerial behaviours exhibited by managers have reputational consequences (Tsui and Ashford, 1994).
How managers are perceived and judged can cause peers, superiors and other key people either to give or withhold important resources such as information and co-operation. Furthermore, perceptions can affect positively or negatively the attitude of employees, which in turn can determine how subordinates perform and affectively engage with the organization (Hall, Blass, Ferris and Massengale, 2004). As Shipper and Davy (2002) claim, numerous studies have shown that managerial success is defined in terms of either employees’ attitudes, managerial performance or both. Indeed, as Eagly, Makhijani and Klonsky (1992) found from a review of 147 studies, ‘satisfaction with the manager’ was the dependent variable in 74% and ‘managerial performance’ in 58% of the studies. Clearly, least effective or ineffective managerial behaviours which cause dissatisfaction on the part of key stakeholders, and lack of engagement on the part of employees, can severely damage a manager’s reputational effectiveness. And thiscan lead to managerial failure. But, as discussed already, few researchers have attempted to identify the specific types of managerial behaviours that differentiate‘good’ (effective) from ‘bad’ (ineffective) management/managerial leadership. Building upon the past work of Hamlin et al who have attempted to do so, our ‘replication and extension’ study in the UK Independent Secondary Schools Sector addresses two questions as follows:
- How are the behavioural indications and contra-indications of perceived managerial and leadership effectiveness manifested within UK independent secondary schools?
- How and to what extent are the identified behavioural indicators resulting from Question 1 similar or different to those identified within other organizations and organizational sectors in the UK?
Research Methodology
The subject focuswas managers in ‘academic’ and ‘pastoral’ managerial roles withinthreeindependent Girls High Schools. These included Heads of School, Deputy Heads, Heads of Departmentand House Masters/Mistresses. We adopted a realist (post-positivist) paradigmatic stance (Guba and Lincoln, 1994; Madill, Jordon and Shirley, 2000; Ponterotto, 2005), and we used Tsang and Kwan’s (1999) notion of ‘empirical generalization replication’ plus Berry’s (1989) ‘derived etic’ approach to applied research. The study was conducted in two stages as follows:
Stage 1:This stage addressed Question 1. Concrete examples of effective and ineffective managerial behaviour (critical incidents-CIs) were obtained from a convenience yet purposive sample of managerial staff and teachers within each of the three schools using the well established Critical Incident Technique (CIT) of Flanagan (1954). This was applied in accordance with the CIT protocols used for Hamlin’s original andsubsequent replicationstudies. Prior to the CIT interviews taking place each interviewee was briefed on the purpose of the research and the anticipated benefit to their school; what was hoped to be achieved at the CIT interview; what was meant by certain key terms that would be used- such as ‘critical’, ‘incident’ and ‘critical incident’; what the interviewee would be asked at the interview; how to prepare for it; and the academic code of ethics that would be applied. They were also briefed on the following two definitions: (i) Effective managerial performance is ‘behaviour which you wish all managers would adopt if and when faced with a similar circumstance’, and (ii) Ineffectivemanagerial performance is ‘behaviour which, if it occurred repeatedly, or was seen once in certain circumstances, might cause you to begin to question or doubt the ability of that particular manager in that instance’. Typically the CIT interviews lasted for 75-90 minutes. During this time the interviewee was asked to describe up to five effective and five ineffective CIs that they had personally observed within the past 6 to 9 months. Depending upon their own role/position in the school structure, the CIs described could relate to behaviours exhibited by managers above or below them, or at the same level. Those CIT interviewees who were managers were not allowed to offer CIs relating to their own managerial practice. When capturing a CI the researcher recorded it as far as possible in the actual words and phraseology used by the respective CIT interviewee. On each occasion the recorded data was reflected back to check the accuracy of interpretation of the described CI. Because of the strict code of anonymityapplied the CIT interviewees were required not to reveal the identity of the particular managers whose behaviours they had described.
The ‘unit of analysis’used for the data analysis phase of Stage 1was the ‘critical incident’. Because of the limited amount of CIT data obtained from each school the critical incidents (CIs) were combined and treated as one sector-related data set. Through a process of open and axial coding (Strauss and Corbin, 1998) and a form of ‘summarising’, ‘inductive category formation’ and ‘explicative’ content analysis (Flick, 2002; Mayring, 2002)applied at the semantic level (Braun and Clarke, 2006), theCIs were initially scrutinized for evidence of ‘sameness’, ‘similarity’ or ‘congruence’ of meaning and clustered accordingly. Sameness existed when the sentences or phrases used to describe two or more CIs were identical or near identical. Similarity existed when the CI sentences and/or phrases were different but the kind of meaning was the same. Congruence existed where there was an element of sameness or similarity in the meaning of certain phrases and/or key words. These behaviouralclusters of effective and ineffective CIs weresubjected to a form of realistqualitativeanalysis (Madill, Jordon and Shirley, 2000) using a variant of ‘selective’ content analysis (Miles and Huberman, 1994; Flick, 2002) to identify and develop a smaller number of positive (effective) and negative (ineffective) core categories. Behavioural statements (BSs) were then devised to reflect in essence the meaning held in common to the respective constituent CIs of each category. We refer to these core categories and behavioural statements as the behavioural indications and contra-indications ofperceived managerial and leadership effectiveness within the UK Independent Secondary Schools Sector.
Stage 2:To address Question 2 the BSs resulting from the Stage 1 researchwere deductively classified and categorized using a form of structuring content analysis (Flick, 2002, Mayring, 2002). For this purpose the ‘behavioural statement’ was the unit of analysis, and the 17 behavioural criteria constituting the latest version of Hamlin’s (2009) ‘generic framework’- now referred to as an emergent ‘general taxonomy’ of perceived managerial and leadership effectiveness- were used as the coding categories. The extent of overlap between the BSs and those underpinning the generic behavioural criteria was then deduced.
Ensuring internal consistency and external validity: To ensure consistency in the application of the CIT data collection and analysis protocols Author One trained and coached Author Two prior to her conducting the CIT interviews, and subsequentlyprovided further guidance and support as required. To ensure and enhance the validity of the research findings in terms of their plausibility, trustworthiness and credibility, a form of ‘investigator triangulation’ was applied for the data analyses (Easterby-Smith, Thorpe & Lowe, 1991). Initially this involved the authors working independently of each other in analysing the respective sets of collected CIs and deduced BSs. They thencompared and contrasted their individual analyses and interpretations in order to arrive at a mutual confirmation of where there was convergence and divergence (Knafl & Breitmayer, 1991). Overall there was general agreement regarding their respective judgments and perceptions of the commonalities existing between the respective sets of data. Minor discrepancies and inconsistencies that arose were resolved through critical examination and discussion to reach a consensus.