Promotion to leadership, not just merit, but insider knowledge: What do school principals say?

The interviews in this study provided a balance of both female and male perspectives with the majority (n=8) being experienced practitioners in that they had led schools for seven to fifteen years (see Table 1). The views of comparatively inexperienced principals (n=4) were also evident having served as principals for periods ranging from one to four years. Further, the respondents were drawn from both secondary (n=6) and primary school (n=6) settings. The qualitative data gleaned from the interviews provided details of current merit selection practices from a small, but demographically diverse range of NSWDE principals. Consequently, their personal reflections and insights into not only their own, but also their colleagues’ merit selection practices, greatly assisted the design of the statewide online survey discussed in this paper.

Table 2 presents the respondent profile of the principals who completed the online survey (n=191) and is remarkable in that it is similar to both the age-band and geographical location data of principals published in NSW state government demographic statistics (NSWDE, 2015b; NSWDE, 2017). Further, the weighting of female and male principal voices provided a relatively balanced perspective regarding the in-school merit selection practices utilised by those principals at the time of this study.

The rich data generated from both the interviews and surveys were subjected to a systematic coding process characterised by the identification, categorisation and detailed analysis of overarching textual themes (Bazeley, 2013) by both SPSS and NVivo software in order to address the two RQs shaping this study. Consequently, the raison d’être of the report that follows is twofold, namely to discuss the extent to which NSWDE principals regarded their stewardship of the merit selection paradigm to be objective, and secondly, whether the interplay of non-merit variables influenced their selection practices when recruiting middle-level school leaders. The following section therefore reports the findings applicable to each of the two previously mentioned RQs.

The principals who completed the online survey were asked to indicate via option-button entry, whether they believed merit could be measured objectively when recruiting school leaders (see Table 3). The majority (54.8%) indicated that either it wasn’t possible to do so, or were unsure if merit-based selection could be executed objectively. These findings accord with research reporting that implicit bias, often created by unconscious attitudes and stereotypes, was highly prevalent in merit selection decisions (Beattie & Johnson, 2012; Foley & Williamson, 2019). Principals were also asked to elaborate using their own words, on why they had selected the survey response option in Table 3. Many indicated that the merit selection of school leaders was both a variable and subjective process:

Also, when asked during interview whether they thought merit could be measured objectively, one principal stated:

As well, survey respondents indicated a tendency for the selection of “known” and “trusted” internal candidates – especially those who had relieved in leadership positions over time:

A principal interviewee also suggested that:

A common view also expressed by principals in this study was a desire to select candidates who were the “best fit” for their school. This opinion features widely in scholarly literature with varying definitions (Baltzell & Dentler, 1983; Baron, 1990; Gronn & Lacey, 2006; Walker & Kwan, 2012). However, despite the inherent difficulty in objectively measuring it, the closest thing to a unified definition of best fit has been proffered by Palmer et al. (2016) who consider it a “real” or “perceived” congruence between a candidate’s attributes and aspects of the organization (school culture). This notion was similarly reflected in survey responses:

Further to the idea of best fit, survey respondents completed an open question to explain if and why they considered the merit selection process to be a reliable recruitment tool. Their responses were subsequently coded according to reoccurring or essence-capturing patterns (Bazeley, 2013). As a result, four main themes emerged (see Table 4) with each being assigned a numerical value ranging from one (1) to four (4). The conversion of qualitative data to numerical format enabled SPSS software to generate the resultant frequency profiles. Interestingly, the two most prominent themes to emerge related to the subjective assessment of best fit, and concerns about the integrity of the selection process – especially with regard to the influence of subjectivity and favouritism.

Another area where the objectivity of the merit selection process was called into question was the interview. Principals were asked to complete an open-ended survey question to indicate if and why they considered the recruitment interview paradigm to be an effective way to assess merit. Their responses were coded and analysed by SPSS software generating the frequency data in Table 5 below.

Just over half the survey participants (n=73, 55%) regarded the interview process to be an ineffective way to assess merit. Typical survey commentary regarding principals’ concerns about the integrity of the merit selection process are illustrated below:

Similar concerns were raised during face-to-face interviews with principals:

Survey respondents were also apprehensive about their ability to objectively assess and compare leadership applicants in high stakes selection contexts often characterised by “impression management” (Law et al., 2016) or, to put it more simply, interviewee deception:

Other survey respondents opined that given the subjective nature of the interview regime, it would be better to replace it with a practical component requiring applicants to perform a school-related task for appraisal by the selection panel:

The inability of principals to objectively assess merit may be linked to the amount of NSWDE-initiated merit selection training they had received. Indeed, an online option-button survey question requiring principals to indicate the number of merit selection training-hours they had completed is presented in Table 6. This data demonstrates that the vast majority of principals in this study had received limited professional training regarding the administration of the NSWDE merit selection process (41%, n=71). Indeed, 56% (n=98) of respondents indicated that they had received only two hours or less in-service training prior to leading the merit selection process in their respective schools.

Principals completed an open-ended survey question to specify the extent to which non-merit variables (NMVs) such as personality, age, gender and ethnicity influenced their in-school leadership recruitment decisions. As can be seen in Table 7, the generation of frequency data revealed that non-merit factors played a significant role in the merit-based decision making of NSWDE principals whilst appointing middle-level school leaders to their leadership teams.

Clearly the majority of principals in this research held the view that in-school deliberations for the selection of middle-level school leaders were heavily influenced by factors other than merit. Principals suggested that:

Further to the above, an option-button entry question in the online survey also revealed that 68.6% of NSWDE principals (n=120) in this study also found it very difficult to exclude non-merit variables from their selection panel deliberations (see Table 8).

When asked to expand upon their reasoning for the option each had selected, principals’ responses were coded and analysed by SPSS software to reveal three basic response-types. Firstly, the majority of respondents (n=103) held that it was human nature to be subjective and hence, it was not possible to exclude non-merit variables from the recruitment process. Alternatively, a smaller group (n=48) maintained that a strict adherence to selection criteria would facilitate greater objectivity in the assessment of potential leadership candidates. Whereas only one principal was unsure whether non-merit variables played any role in their previous merit selection experience (see Table 9).

The role of human nature was highly prevalent in the principals’ survey responses. Indeed, disclosure of the subjective nature of their merit selection practice was evident in responses such as:

Additionally, one principal who participated in this study’s interviews suggested that:

Clearly, this subjective approach to the selection of school leaders is not in keeping with the meritocratic ideal.

This paper has explored some key inconsistencies associated with the merit selection of school leaders in NSW government schools. Advocates of merit selection maintain that every candidate has an equal chance of attaining a leadership position in what is purported to be the fairest way of building school management hierarchies (Hausser, 2013). Despite that claim, this research suggests that the notion of a “presumed objectivity of merit” (Foley & Williamson, 2019) is problematic. Further, it revealed that merit-based leadership recruitment in NSWDE schools is not always an objective, fair and equitable process.

Principals in this study indicated that non-merit variables were difficult to ignore in leadership recruitment decisions (Table 3). This was surprising given that systems justification theory (Jost et al., 2004) intimates that beneficiaries of merit selection are more likely to support it as a fair and legitimate procedure (Lambert et al., 2016). Nevertheless, their concerns accord with scholarship that has found implicit bias to be common in leadership recruitment regimes (Foley & Williamson, 2019). Indeed, research has revealed that managers often demonstrate an affinity for job applicants who are like themselves (Blackmore et al., 2006; Eagleson et al., 2000), are physically attractive (Desrumaux et al., 2009) or are deemed the best fit for their organisation (Palmer et al., 2016). Clearly, recruitment outcomes based upon such variables run counter to the meritocratic ideal of objective selection on a level playing field.

Another inconsistency regarding the merit selection of school leaders is an assumption that principal-led recruitment panels were either unaffected by the interplay of non-merit variables or were objective enough to quarantine their subjectivities (Foley & Williamson, 2019)— suppositions shown to be baseless (Lenton et al., 2009; Pronin et al., 2004). A more likely influence on the school-based selection paradigm, however, is the quality and duration of principals’ merit selection training. To that end, data in Table 6 highlight the need for more extensive and on-going professional training to enhance the ability of NSWDE principals to objectively identify merit despite the influence of non-merit variables.

This can be achieved by the adoption of a hybrid-training model (Todd et al., 2017) utilizing annual real-time, face-to-face workshops (as opposed to solitary online platforms). Principals would thereby have an opportunity to critically reflect upon and refresh their merit selection skills in an inclusive, collaborative, collegial environment facilitating the sharing of merit selection experience and expertise (Johnson et al., 2000).

Further, the survey data also indicated that principals in this study considered that human nature injected selection bias (see Table 9). This accords with research findings indicating that even those individuals who overtly endorsed egalitarian values, inadvertently made stereotypical associations based upon implicit (or unconscious) bias in the recruitment context (Devine et al., 2012; Foley & Williamson, 2019; Van Nunspeet et al., 2015; Washington & Kelly, 2016).

In order to eliminate implicit bias, all government school principals should be required to undertake bias awareness training, the underlying premise of which holds that implicit selection bias is a habit that can be broken with practice (Devine et al., 2012; Foley & Williamson, 2019). Accordingly, a proven professional learning regime that could be utilized is one designed and implemented by Devine et al. (2012) who demonstrated the possibility of achieving dramatic, long-term reductions in levels of implicit bias amongst recruiters. Initially developed to reduce race bias during recruitment, this program up-skilled trainees in the use of multi-faceted habit-breaking intervention techniques. Since it equips participants with a wide range of bias reduction strategies, it could readily translate to any school-based merit selection setting to enhance the objectivity of both principals and their panels.

Another concern expressed by many principals in this study was their perception that the panel interview process was fraught with bias and subjectivity (see Table 5). Research has demonstrated that such bias can be markedly reduced by conducting several separate interviews rather than by an overall reliance on a panel (Bohnet, 2016). Given that interview panels often comprise influential individuals who dominate group discussion, there is always a risk that their views may propagate bias and “group-think” (Fontana & Frey, 1998) within the entire panel. Thus, this strategy therefore affords individual panel members with opportunities to participate in a series of one-on-one interviews, develop rapport with each candidate and draw their own selection conclusions (Nankervis et al., 2014) prior to reconvening as a panel. Panelists would then table and defend their selection decisions as required to determine a consensus candidate.

Whilst it hasn’t been externally evaluated, another possible way to further minimize interview subjectivity and bias could be provided by the emerging field of artificial intelligence (AI). Research conducted by Stockholm’s KTH Royal Institute of Technology in association with AI company, “Furhat Robotics”, has resulted in the trial of the world’s first robot capable of conducting bias-free job interviews (Savage, 2019). Tengai the robot has conducted multiple test interviews during its trial period engaging with a diverse range of recruiter-volunteers to “learn” as many selection behaviours as possible in order to counteract the narrow influence of its programmers (Savage, 2019).

Tengai, unlike traditional robots, has been designed to mimic speech patterns and subtle facial expressions as well as to deliver interview questions in an identical manner, tone and order. The obvious advantage of utilizing this is that unlike human recruiters, robots don’t succumb to unconscious bias regarding non-merit variables such as age, gender or ethnicity. Hence, Tengai interviews and records applicants thereby allowing interview transcripts to be analyzed by human recruiters using only the words on the page to ascertain whether interviewees should move on to the next phase of recruitment. Interestingly, Tengai has conducted real job interviews throughout 2019 (Savage, 2019).

Another selection strategy that could be incorporated into the NSWDE merit selection paradigm is the elimination of interviews from the merit selection process entirely. A principal-recruitment trial conducted by Wildy, Pepper and Guanzhong (2011) in conjunction with the Western Australia Education Department has demonstrated that recruiters are able to dispense with traditional interviews, relying instead upon an applicant’s curriculum vitae, written referee reports and a school-based performance-based task. Wildy et al. (2011) were able to quantitatively rate each of these components using predetermined standards-based assessment criteria to determine if applicants should be placed in a “pool” awaiting an offer of a principal’s position during the following twelve-month period. This empirically tested recruitment regime should be further explored by the NSWDE given that it markedly reduced the impact of subjective selection practice and provided selection panels with an opportunity to observe how prospective leadership candidates operated in a given school context.

The principals in this study revealed that their merit selection practice is, more often than not, a subjective process fraught with implicit bias. This perspective is problematic given that the culture of the NSWDE has long embraced an unshakable belief in the meritocratic ideal and its associated checks and balances. The preceding data has demonstrated that this narrative doesn’t reflect the reality of middle-level school leadership recruitment practice in NSWDE schools. Additionally, it also highlights the challenge for the NSWDE to modify its merit selection regime to produce more objective and equitable recruitment outcomes for all school leadership applicants. Finally, whilst the inherent nature of implicit bias in the merit selection practices of NSWDE principals may be difficult to eliminate entirely, there is much evidence in scholarly literature revealing that its merit selection protocols (policies and procedures) can be positively modified to produce a fairer and more objective recruitment regime – an outcome than can only be achieved if the NSWDE commits to a long term evaluation of its middle level school leadership merit selection practice.