Implementing Personal Construct Theory to explore divergent approaches to substantive knowledge in RE

Divergence is not a new phenomenon in RE in England and Wales. A diverse range of aims and approaches have been promoted since RE provision became compulsory in 1944. Approaches include the phenomenological, interpretative, and dialogic with current debates centring around the benefits and challenges of a ‘worldviews approach’. There have been a range of responses to this approach, and implementation has been diverse. It may appear that at the heart of this divergence is confusion over the definition of the term 'worldview' and what this entails for teachers: adding content knowledge and diverse sources. However, this paper argues that the divergence is much more significantly rooted in the nature of ‘substantive knowledge’ and the power dynamics associated with who defines, selects and assesses the validity of substantive knowledge. In current research, I am testing the hypothesis that teachers’ personal worldviews influence their choice of substantive knowledge in RE. Through 142 questionnaires and 21 interviews, this power dynamic is being investigated. Employing Kelly’s (Kelly, Personal construct psychology, Norton, 1955) Personal Construct Theory illuminated teachers’ personal worldviews relating to substantive knowledge in RE. Through the completion and analysis of RepGrids, the teachers revealed they taught the ‘good’ aspects of religions and ignored the ‘bad’. They prized knowledge from lived experience above knowledge provided by religious institutions and some deemed online sources more useful than those from religious institutions. The power to define ‘substantive’ knowledge, once primarily in the hands of religious institutions, has dissipated to embrace definitions from teachers’ personal worldviews. Whilst enabling a diversity of beliefs to be covered, without examination, hegemonic power may risk the validity of the subject. What is needed, in the midst of this divergence, is not a battle for supremacy, and a rejection of one voice over the other, but an acknowledgement of the power dynamics in defining substantive knowledge in RE.


Divergence: where and by whom?
This article has evolved from a paper that I presented at the AULRE conference, 2023, for which the conference theme was 'Divergence in RE'.Divergence is not a new phenomenon in RE in England and Wales where a diverse range of approaches have been promoted since RE provision became compulsory in 1944, many in response to challenges/opposition to the subject.
I am employing the term divergence to entail a move away in different directions, much as a stream might do finding the path of least resistance around an obstruction as it tumbles towards the sea.As the subject of RE has hit obstructions or oppositions it has sought a path, or paths, to progress forward.The subject has moved through name changes initially from Religious Instruction (RI) to Religious Education (RE) in response to objections about 'indoctrination' or 'Christian instruction' (Freathy & Parker, 2013), in a seemingly linear way.Yet schools have adopted a range of names for the subject, since 1944, signifying different tributaries (Parker et al, 2016) not simply of name but of the nature, role and purpose for the subject: Philosophy and ethics, Religious Studies, Beliefs and Values, Moral and Philosophical studies, Social and cultural studies.This multifurcation includes not only the name and nature but also the pedagogical approach of the subject.A wealth of approaches have been proposed in RE: the world religions approach (Brown, 2000in Grimmitt, 2000), the phenomenological (Smart, 1968), interpretative (Jackson, 1997(Jackson, , 2007)), narrative (Erricker et al., 2011), dialogic approaches (Freathy et al, 2017) and more recently the 'worldviews approach' (Cooling, 2020).

Investigating the root of the divergence
Identifying roots of divergence for RE, throughout history, is important in any attempt to navigate the divergence and understand its impact: a build-up of silt that needs dredging, or rock that needs circumnavigating etc.Current debate/divergence centres on the benefits and challenges of a 'worldviews approach'.There have been a range of responses to this approach both positive (Cooling, 2020;Freathy & John, 2019) and negative (Barnes, 2022), and implementation has been in diverse ways.It may appear that at the heart of this divergence is confusion over the definition of the term 'worldview' (Barnes, 2022, which Cooling acknowledges is 'messy', 2020) and what this entails for teachers: adding extra content knowledge (Barnes, 2022) and diverse sources.However, this paper argues that the divergence is much more significantly rooted in the nature of 'substantive knowledge' and the power dynamics associated with who defines, selects and assesses the validity of that substantive knowledge.Ofsted (2021) calls for RE to deal with 'substantive knowledge', 'disciplinary knowledge' and 'personal knowledge' and defines substantive knowledge to include content and concepts.Ofsted delineates between four aspects: lived experience, knowledge of artefacts and texts, concepts around religious and non-religious traditions and the very concept of 'religion' and 'non-religion' (Ofsted, 2021).Yet Ofsted does not account for, nor question, the power dynamics at play in defining that substantive knowledge.Knowledge, as Freire (1988) and Foucault (1977) would claim, is a reinforcement, or means to reinforce, and maintain power by an elite.

3
Knowledge linked to power, not only assumes the authority of 'the truth' but has the power to make itself true.All knowledge, once applied in the real world, has effects, and in that sense at least, 'becomes true.' (Foucault, 1977, p. 27).
Who defines the 'substantive knowledge' for RE that is taught in schools?For whoever does holds immense power: religious or non-religious institutions, exam boards, syllabus writers, SACREs, subject leads etc. (Flanagan, 2021).These powerbrokers make significant decisions on subject content in the curriculum which may well be impacted by their own personal worldviews or the institutional worldviews that they represent.Figure 1 denotes the potential impact of personal worldviews on choices of subject content knowledge or 'substantive knowledge'.The figure contains lists of possible areas where personal worldviews may formally, and informally, impact choices of subject content, or substantive, knowledge are suggested.These are coded as formal and informal due to the nature of these decisions.Formal choices are those codified in curriculums or syllabus whereas informal choices are made are more flexible, such as who to invite to speak etc.These are in an exemplary, but not exhaustive, list.

Proposal for research
In current research, I wanted to investigate and test the hypothesis that teachers' personal worldviews influence their choice of substantive knowledge in RE.This research examines one potential obstacle that teachers move around in potentially divergent ways -their understanding of what constitutes knowledge in RE.

A theoretical basis
A critical realist approach (Bhaskar, 1975) underpins this research.In its simplest form, Critical realism (CR) is a philosophical approach concerned with ontology, the nature or study of being, which delineates between the 'observable' and the 'real' world (Bhaskar, 1975).This research was concerned with enabling teachers to examine their 'observable' world, how they teach RE, and then investigate their 'real' world, the way that their personal worldviews potentially, and perhaps subconsciously, impact their teaching.Employing a CR stance in RE has precedence (Wright, 2007(Wright, , 2013) ) but is not without critics (Jackson, 2008).Wright is concerned with applying CR to investigate RE as a field of research rather than applying CR to determine how RE is taught, which this research focuses on.The epistemology/methodology may be the same, but the applications are different.Critiques of Wright's work appear to centre around any assumptions of objective truth such as 'some final and undisputed account of religions that can be uncovered by scholarship' (Jackson, 2008:22).However, within this research, examination of self does not necessitate similar conflict as what is searched for is not fallible knowledge of the 'world' or an account of religion(s) but rather self-knowledge -examining or excavating that which is consciously or subconsciously held -identifying action, reaction and impact.A CR philosophical stance was employed to inform the questionnaire design in asking teachers to identify their observable world -the nature of their RE teaching.In tandem with this observable world questions are posed in order to attempt to unearth the 'real' world in a dialectic between personal worldviews and the religious or non-religious worldviews to be taught.

How to elicit personal worldviews on knowledge? (a methodology)
Eliciting deeply held personal worldviews is a challenge and one which previous research has attempted to address (Flanagan, 2020).Attempts to elicit views by reflective writing exist in relevant literature -through the use of questionnaires (Schraw & Olafson, 2002), written reflection (Kanning, 2008;Chen & Huang, 2017), vignettes (Joram, 2007) and autobiographical accounts (Kyles & Olafson, 2008).Yet these have been found wanting.Korthagen and Wubbels (1995, p. 53) point out that reflection can often be overwhelming and vague: 'too big, too vague and too general for everyday application'.Lastrapes and Negishi (2012) in their own research with teacher training students attempted to ward against another danger of reflective writing: reinforcing bias and prejudice.Joram (2007) noted the danger of oversimplification in the self-reflections of trainee teachers.Therefore, I rejected mere written reflection and opted for an approach underpinned by Personal Construct Theory (PCT).
The fundamental theoretical basis of PCT is that each person develops their own rules, constructs, or worldviews (my addition) in order to interpret who and what they experience in life: people, social interactions and life events (Kelly, 1955).Personal construct theory (PCT) is a natural partner for this research.Within PCT, individuals work as 'the scientist' (Kelly, 1955) devising constructs, maps, models or worldviews of the world, to enable them to make sense of the world.These are formed by experience and employed to understand experience: A person lives his life by reaching out for what comes next and the only channels he has for reaching are the personal constructions he is able to place upon what may actually be happening (Kelly, 1977, p. 359).
This accords with CR in that this research is attempting to enable teachers to identify their 'observable' world -what they teach as well as their 'real' world -what underpins their decision, choices and actions in their RE teaching.There are debates around whether PCT can be best described as 'naïve realism, critical realism, radical realism or yet something else' (Epting, 2007).However, Epting rejects these as irrelevant as they miss the distinction in what he deems Kelly's pragmatic work, which aimed to cast off traditional terms and notions.
The open question for man is not whether reality exists or not, but what he can make of it.If he does make something of it, he can stop worrying about whether it exists or not.(Kelly, 1969, p. 25).
Therefore, Kelly is concerned, not so much with a CR focus on ontological questions of reality but with socio-constructivist ideas of meaning making.Some may see a tension between the two as PCT asserts that events may be interpreted by individuals in a myriad of different ways yet CR posits the existence of a 'real' world.Kelly acknowledges that and deems this 'constructive alternativism' (1955 cited by Burr et al, 2014, p. 2).PCT focuses on the teachers' constructed meanings: from their thoughts, emotions and actions in response to and as part of life events.This 'construct system' or worldview, is the lens by which the teachers see the world.PCT lends itself to investigate the 'real' and 'observable' nature of teachers' worldviews.
Employing PCT proved effective for the teacher participants.Indeed, one participant commented on the revelatory nature of the exercise: just occurred to me that we've been encouraged and entrusted to come up with some core concepts, none of which I've mentioned.This teacher had helped his RE department decide on three main concepts in school: 'ultimate questions', 'living a good life', and 'identity and community'.Through engaging with the process of this research, however, the two main constructs which were identified for his RE teaching were individual as opposed to dogmatic and set versus free.
I think the concepts that I've identified are essential to the RE curriculum at my school, do fit in with that, but there's a few more but I guess what I've done today without realising, I guess, is sort of saying, well, yes, we've got these concepts at school and they're all these concepts that kids need to know.But ultimately, what does it amount to?What does it boil down to?
The teacher found that their underlying key constructs for RE actually had more to do with autonomy and individualism.This may well reveal elements of his own personal worldview in terms of what he values, and further analysis is needed to interrogate this.

Sampling
The sample for this research included both trainee teachers and practicing teachers in order to provide a broad sample of experience, knowledge and understanding of how to teach RE.This research relied on volunteers who self-selected to complete the questionnaires and to participate in the Repertory Grid Method (RGM) sessions.Links to the online Questionnaires were sent out from the end of January 2021 to coincide with the National Association of Teachers of RE (NATRE) Strictly RE conference, in England, attended by approximately 350 RE teachers.Additionally, the surveys were sent out by St Culham Gabriel in their spring newsletter, through Learn Teach Lead RE networks, by Diocesan Advisors, by NATRE RE Advisors, by RE Initial Teacher Education lecturers at various institutions, through Facebook and Twitter.This may skew the sample as teachers who attend these events are passionate about teaching RE and are engaging with training.They may not be representative of the general teaching population in the UK therefore making generalisations inadvisable.However, I recommended to these participants that they seek out teachers who do not engage with these networks to fill in the survey in order to gain data that may reveal explanations beyond the participating group.

Examining and analysing potential relationships between personal worldviews and knowledge in RE
Through questionnaires (N = 134) and interviews (N = 21), the power dynamic and interrelationships between teachers' personal worldviews and substantive knowledge were examined.
Francis's attitudinal studies (1978) and Francis & Greer (1990), Valk's (2009) framework tool for worldview identification and Larkin, Freathy, Doney and Freathy 's (2020) implementation of that tool provided a guide on which to devise appropriate questions.
The semi-structured interviews (N = 21) were conducted through Microsoft Teams or Zoom, employing the repertory grid method, RGM (Kelly, 1955).The RGM relates to Personal Construct Theory (PCT) yet rather than merely conceived as an additional element the RGM embodies the theory: RGM 'is personal construct theory in action' (Bell et al., 2004).Employing Kelly's (1955) PCT illuminated teachers' personal worldviews relating to substantive knowledge in RE.The teachers were asked to identify 8 elements of knowledge for teaching RE.Through the process of triadic elicitation constructs were identified.These were placed onto the grids and assigned numerical values to excavate aspects of their personal worldviews.Principal component analysis (Pearson, 1901) was undertaken to analyse these grids.The interviews employed the RGM to facilitate an archaeology of the self (Ricoeur, 1970): to make conscious that which is subconsciously held.
The range of data was analysed in appropriate formats for the data produced: Pearson Coefficient Correlations (Norman, 2010) and Rstudio for the quantitative data from the questionnaires, Principal Component analysis for the RepGrids (Pearson, 1901), and Nvivo and thematic coding (Braun & Clarke, 2006) for the qualitative data from the interviews.

Findings on potential relationships identified between personal worldviews and knowledge
The results of the questionnaires and the RGM interviews provided a complex and contrasting picture.The questionnaires (N = 134) were completed by teachers who were teaching throughout England and Wales and were primary (77) and secondary (57) teachers at a range of stages of their careers (Fig. 2).
The questionnaire was designed in two sections: the first concerned with personal worldviews and the second with the teaching of RE.The aim for this research was to identify any relationships between teachers' personal worldviews and their choice of substantive knowledge for RE, so the focus for analysis was on connections between the first and second sections of the questionnaire (Tables 1 and 2).
The results of the questionnaires were analysed for significant correlations by calculating Pearson coefficient correlations which is an appropriate form of analysis for Likert questions (Norman, 2010).The most significant correlations between the personal worldviews section (Q5 to Q14) and the teaching RE section (Q15 to Q34) of the questionnaire are displayed on the table below.As a standard rule a result of 0 to 0.3 is deemed a weak positive correlation, between 0.3 and 0.5 is a moderate positive correlation, and above 0.5 is a strong positive correlation (Schober et al., 2018).
From the table it is evident that the key questions to reveal moderate positive correlations concern the spiritual realm, prayer, and ethical rules and morals based on a religious faith/tradition.The highest of these is the connection between Q14 and Q20: if a teacher

Trainee
Newly Qualified Teacher 2 to 5 years 6 to 10 years over 10 Fig. 2 Questionnaire respondents by years of teaching themselves follows a set of ethical or moral rules based on a religious faith, they are more likely to teach about ritual and sacraments as a method for accessing/communicating with the divine.This is highly significant in terms of the metaphysical or spiritual nature of religions.If rituals are taught in a procedural way devoid of any spiritual realm this may present a deficit model of any religion.Therefore, from this initial analysis correlations are revealed which need further analysis, particularly on the metaphysical nature of the subject.

The nature of substantive knowledge
One primary difference on the questionnaires between teachers was in answer to the questions on the metaphysical nature of the subject.To investigate this further data was split into discrete groups and placed on scattergraphs: those who declared following a religious faith and those who did not.Q23 was analysed further.At an initial glance Fig. 3 appears to display the importance of the spiritual realm for teachers who declared that they followed a religious faith.Whereas Fig. 4 appears more disparate or divergent.
Scattergraphs demonstrate some significance, but the lack of numerical value may hide more significant data.Therefore, a process of 'jittering' was undertaken through Rstudio to make explicit all data.'Jittering' is a process which addresses these types of challenge of overplotting and reveals the hidden numerical values.Q23 On a scale from 0-10, how much would you agree/disagree with the statement 'The spiritual realm is no longer important to teach in RE'?
Fig. 3 Scatter graph for answers from those who declared that they followed a religious faith Figures 5 and 6 display the jittered version where the numerical values are apparent.This clarifies the sense that the spiritual realm is important for those who declared a religious allegiance (Fig. 5).Whereas the spiritual realm is seen as less important for those teachers who declared no religious affiliation, see Fig.   Q23 On a scale from 0-10, how much would you agree/disagree with the statement 'The spiritual realm is no longer important to teach in RE'?Further evidence of divergence in the teachers' views can be observed in the source of knowledge for their RE teaching.For example, the results of the ranking order question, Q22, provide insight.Figure 7 reveals the initial results for Q22, Place these in order of importance to you in teaching RE: Figure 8 reveals the combined totals for each of the ranking order questions.After calculation the overall order for importance were: 1. Lived experience, 2. Traditions, 3. Texts, 4. Doctrine.
These results raise questions about authoritative knowledge.However, the ability to draw conclusions is somewhat hampered by the lack of clear definition of 'lived experience' within this question.The term 'lived experience' is employed frequently in the RE community (Ofsted, 2021, e.g.RE Online refers to 'lived experience' 488 times) and often this refers to the actual life experience of believers in local communities near the school or in the UK.I had chosen the term to reflect that definition but had not explicitly defined it for this question.However, further questions make this more explicit.Are the teachers employing 'lived experience' to refer to the experience of one believer in the community of the school or of the religious community in the UK or the world?Dependent on definition, this might reveal a shift from 'expert' to 'practitioner' or from community to individual?Fig. 6 The jittered results for the teachers who declared following no religious faith Further questions in the questionnaire did pertain to the source of knowledge: where they choose to use knowledge from-religious institutions (Q28), academic/scholars (Q29) and media outlets (30).In these Likert scale questions percentages were calculated for scores of 9 or 10 given by teachers, indicating they highly valued these.The results revealed that only 13% of teachers highly valued religious institutions as a source of substantive knowledge, whereas 16% highly valued knowledge from academics or scholars and 16% highly valued knowledge from media outlets.Whilst these are not statistically wide variations, they do signify which sources of knowledge are most highly valued.However, when the data is split into those who declare a religious faith and those who do not the figures are more revealing.For those who declared that they did not follow a religious faith only 6.2% highly value religious institutions as a source of substantive knowledge (those giving Q28 a score of 9 or 10), 17% highly valued academics or scholars as a source of substantive knowledge (Q29) and 18% highly valued social media as a source.For those who declared following a religious faith 20% highly valued religious institutions, 16% highly valued academics, and 14% highly valued media content.A comparison might be seeking knowledge about cancer from a patient as opposed to an oncologist: both have valid knowledge, but experiential and academic knowledge differ.These figures reveal a significant divergence in the teachers' choices of the source of substantive knowledge and seem to be related to their own personal worldviews.
Overall, in the questionnaires, the teachers prized knowledge from lived experience above knowledge provided by religious institutions and some deemed online sources more useful than those from religious institutions.The power to define substantive knowledge, once primarily in the hands of religious institutions, has dissipated to embrace definitions from media outlets and lived experience.Whilst this may be welcomed by some as a more representative picture of religious communities current lived experience, others may caution media outlets as a source of authoritative knowledge.Media outlets have a range of practices in terms of requirements for proof or evidence and often have no peer review process.Therefore, materials can be published online as factual/substantive knowledge about a religion or worldview which may be far removed from the general practices of many who follow that religion or worldview personally.Indeed, Mamodaly (2016) exhorts teachers to develop media literacy and criticality in their lessons in order to portray the nuanced pictures of religion(s).

The range of knowledge
Insight from the Repertory Grid Method (RGM) reveals a divergence in teachers views on significant knowledge in RE.The teachers (N = 21) identified 94 different elements of knowledge in their RE teaching.Such a diversity might be surprising in any other subject.For example, subject knowledge in maths is clearly set within the National Curriculum (2014).But the variety in maths may appear in terms of procepts.This intertwines procedural knowledge and conceptual understanding.For some researchers, teachers gaining strong conceptual knowledge alongside the subject content knowledge is vital (Gray and Tall n.d. retrived 2023).Through PCA, the two main constructs for each teacher were identified -again divergence emerged and homogeneity was absent.21 teachers identified a wealth of different constructs for teaching RE.With the two main constructs, of these 21 teachers, no two were identical.For example, the following two figures show the crossplot PCA analysis for two teachers, Teacher A and B. The constructs on the x and y axes are very different and the 8 elements are also different.These were chosen by the individual teachers and demonstrate what are important to them in teaching RE.
The main constructs for teacher A, in Fig. 9, are open deep attitude vs surface superficial, and written down recorded vs everyday life.The plotting of the elements of knowledge on their grid demonstrate where these interconnect with the constructs.Thus, in their perspective, RE is primarily concerned with enabling pupils to develop an open deep attitude to religion(s) and non-religious worldviews as well as learning about codified bodies of knowledge.
However, for Teacher B, the two main constructs are everyday life vs religious issues and correcting misunderstandings vs entrenching views, Fig. 10.The location of the elements on the grid reveals the teacher's views on what is important in their RE teaching.Thus, from their perspective, RE is primarily concerned with correcting misunderstandings and the impact of religions on everyday life.These two teachers' constructs are not mutually exclusive but do reveal a different focus.Both are concerned with the pupils' knowledge of religion(s), but one is a corrective and the other a developing of an 'open' attitude.How this may impact their actual practice would be an area for further examination.
For Teacher A, everyday life is held in contrast to written down or codified knowledge whereas for teacher B everyday life is held in contrast to religious issues.Teacher B therefore seems to see religious issues as unrelated to everyday life.
For Teacher A, knowledge of history and culture are important but the need to delve deeper into the lived experience is highly significant but from a cultural, rather than a religious perspective.Religious traditions, text and institutions are important but lived experience and teachers' attitudes is vital for pupils' understanding.
For Teacher B, entrenching views plays no part in their RE lessons, rather challenging misunderstandings, examining texts, practice and modern interpretations are most important.Everyday life plays a significant role in terms of religious teachings, ethical theories and social justice.Application of religion(s) seems the lens through which RE is studied.Further analysis of their questionnaire answers and interviews may reveal links between personal worldviews and their RE teaching.For example, teacher A declared a strong allegiance to a faith whereas teacher B declared a nominal faith.Their different personal worldviews on religion are reflected in their constructs.For teacher B, religion is examined in how it impacts on social justice and ethics: which might be termed the application of religion(s).Whereas for teacher A, religion is examined in terms of sacred texts, interpretations, and practice: which might be termed the study of religion(s).

Concluding thoughts
Preliminary findings from this research indicate that teachers' personal worldviews create divergent approaches to defining 'substantive knowledge' in RE and to their choice of the sources of that knowledge.These findings demonstrate that teachers may be unaware of their personal constructs in relation to RE.These may impact curriculum design, substantive knowledge choices etc., consciously, and subconsciously.The RepGrid method (Kelly, 1955) proved an effective tool to elicit unconsciously held views that enabled the teachers to examine their own knowledge choices, and curriculum design.However, further analysis is needed to ascertain the impact of teachers' personal worldviews on their RE teaching.
Divergence is evident in RE in the UK, in defining and sourcing substantive knowledge, and questions remain as to whether the RE community can stem the divergence and unify.However, a subsequent question is whether unification is what is necessary.If unification entails acknowledging one stakeholder (religious institutions, or academics, or media outlets, or exam boards, etc.) as the primary source of authoritative substantive knowledge without examination this hegemonic power may risk the validity of the subject.What is needed, in the midst of this divergence, is not a battle for supremacy, and a rejection of one voice over the other, but an acknowledgement of the power dynamics, including teachers' diverse personal worldviews, in defining substantive knowledge in RE.An acknowledgement and consideration of which voices are promoted or demoted in the process of defining and choosing substantive knowledge in RE may help to ensure that all voices are heard.licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Fig. 1
Fig.1Potential influence of personal worldviews on SCK(Flanagan, 2021) Figures 5 and 6 display the jittered version where the numerical values are apparent.This clarifies the sense that the spiritual realm is important for those who declared a religious allegiance (Fig.5).Whereas the spiritual realm is seen as less important for those teachers who declared no religious affiliation, see Fig.6.
Figures 5 and 6 display the jittered version where the numerical values are apparent.This clarifies the sense that the spiritual realm is important for those who declared a religious allegiance (Fig.5).Whereas the spiritual realm is seen as less important for those teachers who declared no religious affiliation, see Fig.6.

Fig. 4
Fig. 4 Scatter graph for those who declared following no religious faith

Fig. 7 Fig. 8
Fig.7The results for order of importance to teachers in their RE teaching

Fig. 9
Fig.9Crossplot analysis for Teacher A

Table 1
The most significant Pearson Coefficient Correlations

Table 2
Percentages for the most important source of substantive knowledge scoring 9 or 10 on the Likert scale