Having established in the exploratory study that it was possible to deconstruct a post-graduate student’s thinking using Meta-Programmes, and that the Methodology used was appropriate to achieve the principle aim of feasibility, the current study intended to further the stated objectives with additional objectives from which new hypotheses emerged.
With this in mind, it was reasonable to ask the same research question of a larger participant pool to determine if the results were replicable. The question remained: how does the thinking of post-graduate students map against the Identity Compass profile tool?
The objectives from Study 1 were:
- To determine if there are Meta-Programmes common to all post-graduate students.
- To determine if a specific combination of MP’s creates an academic Thinking Style.
- To determine if there are driver Meta-Programmes as suggested in the literature.
The hypotheses from study 1 were also valid in the current study:
- Certain Meta-Programmes have more of an effect on the profile than others and thus might be classed as “driver intentions”.
- Different combinations of Meta-Programmes will be discovered.
The exploratory study demonstrated that the methodological approach was valid. Hence the current study replicated this with the addition of objective 4 and 5 below. The researcher had access to an 8,200-profile database from the Identity Compass profile tool, from which it was possible to extract the profiles of 177 post-graduate students, MBA as well as PhD, obtained over the previous 5 years in order to investigate this.
Therefore, the revised objectives for study 2 were:
- To determine if there are Meta-Programmes common to 177 post-graduate students
- To what greater or lesser extent do 177 post-graduate students have different thinking styles?
- To determine if a similar or different subset of dominant Meta-Programmes arises.
- To discover if any other unique patterns in thinking emerge.
A new hypothesis arose out of the fifth objective:
The objective of deconstructing the thinking of post-graduate students into fifty individual Meta-Programmes using the IC ensured inter-relator reliability as each participant performed the same questionnaire with the same questions. For example, a meta-programme group called ‘Comparison’ contained the Meta-Programmes of ‘Sameness’ and ‘Difference’. As a post-graduate student, an important ability was to be able to notice ‘difference’ in one’s research material (Brown, 2002), and the IC demonstrated to what extent the participants confirmed this Meta-Programme. A student who did not value change, but rather prefers stability and similarity would seek ‘Sameness’ (Weinberg, 2009).
As mentioned in the pilot study, the IC helps to differentiate between intention and attention in context due to the difference in the combinations of the fifty Meta-Programmes. In this regard, it would be anticipated that we saw ‘Difference’ higher in its percentage score than ‘Sameness’ for a post-graduate student.
The research in the current study was a replication of the exploratory study in that the Identity Compass profile tool was used to determine the construction of 177 post-graduate students’ thinking into fifty Meta-Programmes.
The study employed an opportunity sample of 177 Post-Graduate students from Coventry University across two campuses, including 32 PhD students, 11 male and 21 female. There were 145 MBA students from Coventry University London. Participants’ age ranged from 21 to 44 years. As per the pilot study, their countries of origin were, but not limited to: England, India, Pakistan, Korea, China, Japan and Nigeria.
The MBA student data were selected due to the Identity Compass profile tool holding historical data from 5 years’ of High Flyers programmes and it was hypothesised that the initial findings from the pilot study would be replicable on a dataset of similar participants (within-group). The reason for this was that the focus continued to be on how post-graduate students’ thinking mapped against the Identity Compass profile tool.
The PhD students were selected based solely on their capacity to volunteer to be part of the study, their availability to partake in the questionnaire and their philanthropic tendencies as psychology students to assist colleagues. A further reason for the choice of participant data was the potential to differentiate between MBA students and PhD students, as per study 1, to determine if there were differing constructions of an academic Thinking Style at the different academic levels.
As this was an extension of the pilot study, all measures and methods were replicated from study 1, with the main difference being the number of post-graduate student profiles in the data (n=177). This meant that the Identity Compass profile tool used to benchmark student thinking was exactly the same method as in the pilot study.
Objective 5 asked if a benchmark score can be created to normalise the Identity Compass output. It was important at this stage to note what the literature said about such constructs: variables are created by developing the construct into a measurable form (Vigo & Doan, 2015). Any variable must have at least two possible values. Examples of variables include height, age, scores on any scale, and so on. These variables are measured by operational definitions which should identify how the variable is calculated as a numeric value, and specify the range of possible values. It should also establish the variable’s level of measurement (nominal, ordinal, or interval). An example of an operational definition for Awareness would be “the sum of the responses to the Identity Compass profile tool, which can range from 2 to 5.” This is the foundation of the Thinking Quotient scale.
This was achieved by mapping the relevant Meta-Programmes to the behavioural output of Kegan’s Levels of Adult Development (see chapter 1) from a social-emotional perspective, and Laske’s Cognitive Development Framework from a cognitive perspective. See Table 4.10.
Table 4.1: Meta-Programme Mapping to LoAD and CDF
|Caring for Others||Details|
|Caring for Self||Difference|
The principle behind this alignment was those Meta-Programmes that were predominantly gaining their locus of evaluation or control from an external source, such as External, Partner and Caring for Others, were indicative of Kegan’s ‘Socialised Mind’ stage of development, and were thus given the score of ‘3’ to match his stage 3 principle. The level of ‘3-ness’ was then gauged by the difference between ‘Internal’ and ‘External’ such that an individual with these at balance (50/50) was hypothesised to have the choice of either in their response. On the other hand, an individual with 40% difference was only capable of using the higher-valued Meta-Programme in any given situation (see Figure 4.3 as an example).
Figure 4.1: Towards / Away From Visual
From a cognitive perspective, Laske, (2008) stated that a person’s ability to do ‘Abstract’ was a function of higher-level thinking. Thus, an individual at Stage 2 (Laske, 2008) would not be capable of this abstraction, but those at stages 4 and 5 would be, to a greater extent. If we know what a student was not capable of in terms of their thinking capacity, then we have a starting point for their self-awareness score using the principles of balance (choice) above. See Appendix 2 for a full list of Meta-Programmes.
As identified in the Methodology chapter (2), the Intention, Awareness, Choice and Response factors of how a post-graduate student thinks has been omitted from the profile tools currently available, hence why the Identity Compass profile tool was chosen as it allows a deconstruction of an individual’s thinking into quantitative data for measurement. The literature on Meta-Programme groups predominantly placed the Meta-Programme pairs in a binary role, in that ‘Towards’ was always the opposite of ‘Away From’ on every scale. By virtue of this polarisation, it also stated that a person had a certain propensity for ‘Towards’ thinking in opposition to ‘Away From’ thinking. This tied in with Piaget’s ideas on stage transition as seen in the literature review (chapter 2) where he conceived stage transition to be:
- A, B (or not A)
- A or B
- A with B
This could be extrapolated to Meta-Programme use rather than subtasks (Commons, et al., 1984) if ‘thinking’ were the task, where A = ‘Internal’ and B = ‘External’:
As stated above, this linked Piaget to Commons’ interpretation of task complexity, where it is arguable that he missed the opportunity to look at ‘thinking’ as the task. This will be discussed in chapter 8.
The research question in this study asked if the difference between the two MP scores were indicative of the individual’s self-awareness, akin to Kegan’s (1994) Subject/Object theory. In other words, in Identity Compass terms, if ‘Away From’ was 65% and ‘Towards’ was 90%, (see Figure 4.3) then there was a difference of 25%, which denoted a specific level of unconscious intention, and lack of awareness in the individual’s thinking, which impacted their capacity to respond accordingly. With this score, it suggested the person is goal-oriented to such an extent that they would not see the immediate pitfalls of their decision-making, and thus could not think in a risk-aware manner (Away From).
It was hypothesised that the closer the Meta-Programme pair was in their score, the more capable the individual would be of choosing between the two. For example: should they be 50% and 50% then the individual would be capable of choosing which Meta-Programme was more appropriate in context at that time, and more capable of making a qualitatively better decision. The hypothesis argued that this was a limitation in the individual’s ability to respond in the moment as they were only capable of using ‘Towards’ thinking, and unaware of how to do ‘Away From’, thus restricting their capacity for their Intention, Choice and Response.
A student’s capacity to choose their thinking/behaving aligned with Kegan’s notion of Subject/Object behaviour in that if they hold the two MP’s as Object, they could choose to do either in a given context (Kegan, 1994). This was opposed to the student being ‘Subject to’ one of the Meta-Programmes due to its unconscious nature. If the student were not aware of their Meta-Programme use, and it manifested as an habituated reaction to a situation (without choice), then the student would be ‘Subject to’ that MP (Kegan & Lahey, 1984).
As per the literature review, in Kegan’s framework, development does not occur instantaneously. Instead, he argued, that people move from fully constructing their understanding in a way that is consistent with Stage 2 (for example), to building a bridge to Stage 3 by constructing meaning in two ways simultaneously, and then moving beyond the lower stage by incorporating it into the larger frame of the higher stage.
Kegan’s (1994) work illustrated the difference between stage 3 and stage 4 by describing a couple who were struggling with the issue of interpersonal intimacy in their marriage. He noted that if each spouse had a level of development different to the other (one at stage 3 and the other at stage 4, for example), each would have a different idea (construction) of what it meant to be intimate. The differences being, the spouse at stage 4 would not be subject to their construction of self and would recognise the interdependency of the couple without the emotional binds that come with stage 3 thinking. On the other hand, the stage 3 spouse would not be capable of a dissociative position and would subsume their own values for the values of their partner.
Paraphrasing Kegan’s (1994 PAGE NUMBER) own words from his book, “In Over our heads”, he describes the thinking patterns of a stage 3 person, the relevant Meta-Programmes have been placed next to the description to demonstrate the possible deconstruction.
People at Stage 3 no longer see others as simply a means to an end; they have developed the ability to subordinate their desires to the desires of others [Caring for Others, External, Partner]. Their impulses and desires, which were Subject to them in the Second Order, have become Object [Away From, Own]. They internalize the feelings and emotions of others [Internal to External] and are guided by those people or institutions (like a church or synagogue or a political party) that are most important to them [Trusting, External, Affiliation, Consensus]. They are able to think abstractly [Abstract], be self-reflective about their actions and the actions of others [Team Player] and are devoted to something that’s greater than their own needs. The major limitation of Stage 3 is that, when there is a conflict between important partners, whether a spouse or a business, or even a church or political party, they feel “torn in two” [Consensus, Caring for Others, Partner] and cannot find a way to make a decision [Options]. There is no sense of what they want outside of others’ expectations or societal roles [External].
Kegan (1982) notes, “… The popular literature will talk about [someone] as lacking self-esteem, or as a pushover because [they] want other people to like [them]” (p. 96). He goes on to point out that the very notion of ‘self-esteem’ was inappropriate at this stage because self-esteem implies an internal locus of evaluation for feeling good about oneself. However, those at the third stage do not have an independently-constructed self to feel good about; their esteem is entirely dependent on the perspective of others because they are, essentially, constructed by others’ opinions. Thus, a student at stage 3 would be a model student, following the rules out of loyalty to their peers and university. He would try hard not to break the rules as he wouldn’t want to let his colleagues down. In academia, this student could do anything, as long as he had someone to help make the difficult decisions.
To align this idea with the Thinking Quotient above, this lack of choice in their capacity to respond would manifest as a low score on the TQ scale (see Table 4.11). If we were to extrapolate to all fifty Meta-Programmes comprising twenty groups, we then have the foundations of the benchmark scale based on Kegan’s work for the Thinking Quotient. This would denote what the person was and was not capable of.
Following on from the above rationale for the Thinking Quotient process, for the purposes of the current study, 5% increments between the scores of the paired meta-programmes were used in order to better-differentiate the level of Intention, Awareness and Choice. This ensured the Thinking Quotient output was more accurate. See Table 4.11.
This principle was implemented on each of the 20 groups (13 MP pairs, 5x MP triplets, 1x MP fours and 1x MP fives) taking into account what the percentage difference means in relation to the person’s Intention, Awareness, Choice and Response about their conscious and unconscious thinking style. See Table 4.11, and in full in Appendix 6.
Table 4.2: TQ Score as per Percentage Difference
Table 4.11 shows the breakdown of the TQ score for four Meta-Programmes: ‘Towards’ / ‘Away From’ & ‘Internal’ / ‘External’. As mentioned, the principle is one of balance. Should each MP be the same score (balance), regardless of where it falls on the percentage scale, the ability of the participant to choose either MP in context was 50/50 and thus was given the highest score (5) for the TQ. The scores then tapered out from this choice position until they reached 25 percentage points difference, at which point there was sufficient difference between them to warrant a subject/object differentiation (Kegan, 1994) and thus a lower awareness score. With the score for ‘External’ being greater than 25% over ‘Internal’, the outcome for the person was such that they could not do ‘Internal’, which meant they were limited in their response in the moment. This reduced their score from a self-awareness perspective. In other words, choice of response diminished with every 5% incremental difference. This principle extended to all Meta-Programme groups and could be tested in a qualitative interview by asking how the participants know they are or are not aware of any particular MP choice, and the difference between the MP’s that then leads to their response and capacity to respond.
Appendix 6 demonstrates the output score for all Meta-Programmes, but particular attention should be paid to the meta-programme groups: ‘Time Orientation’, which contained the individual MP’s of ‘Past’, ‘Present’ & ‘Future’; and ‘Perspective’ which contained ‘Own’, ‘Partner’, ‘Observer’ (see Table 4.12) as these are likely to be modified in the future as a result of continued research due to the difficulty of determining a difference score between three facets within a group. The principle used thus far was to put the highest-scored MP at the top and the behavioural options emanating from the scores of that lead MP.
Finally, Meta-Programmes such as ‘Abstract’ and ‘Future’ were, according to Laske, (2008) only available to those people who reach a mature stage of their thinking (see chapter 1), which implied that a more balanced score for ‘Abstract/Concrete’ was not a level 2 or even level 3 score. It must therefore be level 4 or higher. Hence, where both scores for ‘Abstract’ and ‘Concrete’, were balanced, the TQ score was 5. Then, for each 5% incremental difference, the TQ score went down by 5 points, which meant that at 10% difference, the TQ score would become ‘4’. This was another factor in how the TQ score was determined. Table 4.12 shows the TQ score for each increment of the IC scale.
Table 4.3: TQ Scoring of Three MP Group
|Group||Meta-Programmes||Score % Difference|
Ethical approval was granted from Coventry University ethics committee to use the larger dataset from which the student data were distinguished. Each MBA student had undertaken an Identity Compass profile as part of their High Flyers programme with CU London. The Identity Compass profile forms the basis of the High Flyers programme from a self-awareness perspective for each student. The programme encourages students to go above and beyond their normal academic achievements. Each participant was presented with the same question set relating to their thoughts in the context of post-graduate study.
All fifty Meta-Programmes were included in the initial exploratory factor analysis in order to determine the factor structure, using IBM SPSS Statistics 25 software. Factor analysis regularly yields a five-factor structure that parallels the typical dimensions of the FFM for self, peer, supervisor, and teacher ratings (Goldberg, 1992; Miller, Pilkonis, & Morse, 2004; Ployhart, Lim, & Chan, 2001).
As the data used in this study were not normally distributed, according to Kolmogorov Smirnov test (p < .05 for all the variables included), the principal axis factoring method of factor analysis was used (Fabrigar, Wegener, MacCallum, & Strahan, 1999).
Before performing the analysis, the basic assumptions for exploratory factor analysis were checked to see if they were met. The sample size was adequate in general (n=177), however, the ratio between items and subjects, 50:177 was not, as recommended by (Tabachnick, & Fidell, 2007). According to the correlation matrix, there were many correlations between the items higher than r = .3. Kaiser-Meyer-Olkin Measure of Sampling Adequacy was equal to .90, which was higher than the recommended value of .60 (Kaiser, 1974). Bartlett’s Test of Sphericity (Bartlett, 1954) was statistically significant (p < .001). This all suggested that the data were overall appropriate for factor analysis, with the only limitation being the item/subject ratio.
The principal axis factoring method revealed eleven factors with eigenvalues greater than λ = 1.00. According to the scree plot, only the first five factors offered any potential solution, according to Cattell’s criteria (1966).
The five factors explained 54.64% of the variance. Direct oblimin rotation was used, in order to get more accurate results, as the factors were correlated themselves, as presented in Table 4.13. All items had communalities higher than .40, showing that they were correlating to the other items, except for the MP of ‘Trustful’ (communality = .25). It loaded on the first factor with the factor loading of .32, which was acceptable according to Tabachnick and Fidell, (2007), so it was not excluded. Items ‘Present’ and ‘Looking’ did not load on any of the obtained five factors with the loading of .32 or higher, so they were excluded from the final factor solution.
Table 4.14 presents the factor structure, i.e. the new five dimensions created based on the Pattern Matrix, keeping those with factor loadings higher than .32, and attaching each item to the factor on which it had the highest factor loading.
Table 4.4: Correlations of the extracted factors
Based on the results of the factor analysis, which was effectively saying: these things tend to happen together, it was deduced that there exists a subset of latent dimensions (objective 3) underlying the fifty investigated Meta-Programmes amongst post-graduate students. Each dimension combined between 5 and 14 different Meta-Programmes.
Table 4.5: Five dimensions of Meta-Programmes
|Dimension 1||Dimension 2||Dimension 3||Dimension 4||Dimension 5|
|Achievement||Team Player||Quality Control||Own||Abstract|
|Activity||Caring for Others||Procedures||Internal||Vision|
|Caring for Self||Affiliation||Things||Polar|
From the list in Table 4.14, it is inferred that the Meta-Programmes within the first dimension are the ‘drivers’ in the context of post-graduate thinking in an academic context.
The sample size of 174 was adequate for the linear regression model with five predictors (Tabachnick, & Fidell, 2007, p. 123). The assumption of multicollinearity and singularity was not violated. There were many high intercorrelations of cognitive dimensions, and two low significant correlations of cognitive dimensions with the Thinking Quotient (Cohen, 1988; Table 4.15). Tolerance and Variance Inflation factor (VIF) measures suggested that there were no problems with multicollinearity between any pair of the independent variables.
Table 4.6: Correlations between the variables in the model
|2. Dimension 1||-.14*||1.00|
|3. Dimension 2||-.10||.62***||1.00|
|4. Dimension 3||-.09||.77***||.66***||1.00|
|5. Dimension 4||-.06||.66***||.66***||.70***||1.00|
|6. Dimension 5||-.15*||.76***||.68***||.74***||.69***||1.00|
*. Significance of p < .05, 2-tailed.
***. Significance of p < .001, 2-tailed.
Normal P-P Plot of Regression Standardised Residual suggested that the data were very close to normally distributed. Scatterplot visualisation suggested that there were not many atypical points, and no violation of some of the assumptions of linearity, normality, and independence of residuals, and homogeneity of variances. Maximum Cook’s distances lower than 1 also indicated that there were no atypical points, and hence no cases to exclude from the analyses.
Finally, the dependent variable TQ was close to normally distributed, based on skewness and kurtosis absolute values lower than 1.96, as well as based on the analysis of the histogram. However, the overall regression model was nonsignificant, F (5, 168) = 1.07, p = .38. Accordingly, none of the five cognitive dimensions predicted the outcome variable TQ, as presented in Table 4.16.
To determine if objective 4 was achievable, the effects of the five cognitive dimensions on the TQ score were tested using multiple linear regression, with five cognitive dimensions as independent variables, and the TQ score as dependent variable. As with the first regression model, three cases were identified as outliers, since they had Mahalanobis distances higher than the recommended threshold of 20.52 (Tabachnick, & Fidell, 1996).
Table 4.7: Regression coefficients of cognitive dimensions with TQ as the outcome
|Unstandardised Coefficients||Standardised Coefficients||t-value||Sig.|
The final multiple regression was conducted without these three cases, on the sample of 174 participants. Before that, the assumptions of conducting multiple regression analysis were evaluated.
The average TQ score for the 177 post-graduate students was 3.19 and can be seen in Figure 4.4. Although there are a number at level 3.0, there are a significant number above this score, which suggests that if this were correlated with Kegan’s levels of adult development, the average post-graduate student has a measure of thinking construction marginally higher than the average population.
Figure 4.2: Average TQ Score for PG Students
As the pilot study had demonstrated the methods used were appropriate, the main aim of this study was to extend the findings of the pilot study, to determine how the thinking of 177 post-graduate students mapped against the Identity Compass profile tool. The aim was to discover any commonalities in the clustering of Meta-Programmes that might form ‘Thinking Styles’, as per objective 2, and to uncover how this thinking influenced their construction of self as a post-graduate student. This research started from the premise that Meta-Programmes exist as high-level abstracted patterns of thinking and responding (Hall, 2005) and can be reliably identified and measured by the Identity Compass profile tool.
Table 4.8: Meta-Programmes ranked by Median score
|Pilot Study (n=32)||Study 2 (n=177)|
|Realisation||80||Caring for Self||80|
|Caring for Self||75||Consensus||80|
As per the pilot study, a hierarchy of Meta-Programmes emerged from the data. However, this was a different hierarchy compared to the pilot study findings. A comparison of hierarchical MP’s can be seen in Table 4.17. What was interesting was that the same seven Meta-Programmes appear in both studies, with the obvious exception of ranked order. ‘Information’ is about learning and appears second in both studies. The MP of wanting to get on with their work (Activity) became the highest unconscious habituated MP with a greater number of student profiles. Goal orientation (Towards) also remained important, as did the unconscious MP of receiving the degree (Achievement). These similarities were understandable as the pilot study was a subset of study 2. However, this ranking is different to the question raised in objective 1, which will be discussed next.
The results of the 177 profiles demonstrated that there were Meta-Programmes common to all post-graduate students, as per Objective 1, and these Meta-Programmes had a specific meaning in the context of academia. Table 4.18 shows the first dimension listing the highest factor Meta-Programmes. As mentioned, the factor analysis was effectively saying that if you answer a certain way with ‘Internal’, you are likely to answer in a similar way with ‘External’. It was possible that dimension 1 lists those MP’s most accessible to the students, and dimension 5 the least accessible. As students are not encouraged to think in an ‘Abstract’ or ‘Global’ way due to the nature of the academic system, it made sense that those two Meta-Programmes were unavailable. Taken in context, the combination of those twelve Meta-Programmes in dimension 1 demonstrated a potential thinking style as per objective 2, and highlighted the intention and direction of thinking behind the profiles for the students. It was possible that the combination was an academic Thinking Style: however, it was more likely to be a Thinking Style common to post-graduate students, suggesting it was also available to non-students alike. See Table 4.18. These component MP’s of dimension 1 are discussed next.
Table 4.9: Five dimensions of Meta-Programmes
|Dimension 1||Dimension 2||Dimension 3||Dimension 4||Dimension 5|
|Achievement||Team Player||Quality Control||Own||Abstract|
|Activity||Caring for Others||Procedures||Internal||Vision|
|Caring for Self||Affiliation||Things||Polar|
The first Meta-Programme was ‘Achievement’ which was about attaining the reward, and in context, this was about the student getting their post-graduate degree. They were actively motivated (consciously or unconsciously) to get the certificate. McClelland (1985) determined there are three basic motives, with Achievement being a stable preference over time. The factor analysis for post-graduate data would support his perspective in this context.
The fact that ‘Information’ was second demonstrated that ‘learning’ was a key (conscious) Meta-Programme to all post-graduate students profiled. Also, a low score in this MP might indicate a tendency towards arrogance.
The third MP was ‘Activity’ which showed an intention to get things done. This could be considered an appropriate MP for post-graduate students in terms of workload and outcome.
‘Long Term’ demonstrated that the student was capable of planning beyond the confines of their studies (usually 12 or 36 months).
‘Future’ also showed that the future was at the forefront of the student’s thinking, and was indicative of higher level thinking according to Laske (2015) and Jaques (1989) in his ideas on ‘span of discretion’.
‘Caring for Self’ was a common MP for foreign students studying in the UK. It emphasised their prerequisite to take care of their own needs first.
‘Pre-Active’ demonstrated that a student was conscientious and willing to work for their degree.
‘Realisation’ was the MP that determined if the students followed through on their ideas, and as it was the very next MP on the list, this suggested that post-graduate students, at least in this context, did indeed follow their ideas through to fruition. This implied from the moment they were assigned a group activity, up to the moment they hand in the finished work.
‘Task’ suggested the student was more concerned about getting the job done than the people involved in the actual task.
‘Difference’ was tenth, and this was an important Meta-Programme for post-graduate students as it essentially showed how a researcher maintained their search filter for what was new and different to what had been written previously. It was indicative of the student’s intention to grow through their research and aligned with the findings of Brown, (2004).
The eleventh MP was ‘Concrete’ which was about knowing the facts, such as who does what and how, and when it will be delivered. This could have been an environmentally-driven MP as research is key in post-graduate study.
The final Meta-Programme in dimension 1 was ‘Observer’ which was a student’s ability to dispassionately step back from their studies and consider the neutral perspective. It does not associate with feelings, and thus highlights the point raised in the literature review that stated in order to be a high level thinker, one must move through emotion in to cognition.
Vermunt (1998) suggested that ‘constructive friction’ was a useful tool for disrupting a student’s thinking by way of a mis-match of teaching/learning styles, which would be made easier to achieve with an appropriate tool such as the Identity Compass. This perspective aligned with Piaget’s (1985) concept of disequilibrium and was supported by the third dimension in Table 4.18. The findings of the factor analysis showed that those Meta-Programmes in dimension 1 were opposed by their MP pair in dimension 3. Thus, in order to cause friction as per Vermunt, (1998) one only needed to offer the post-graduate student a ‘Sameness’ MP where they habitually used ‘Difference’ and their disequilibrium would be in effect. This also supported the findings from Dunn, Griggs, Olson, Beasley & Gorman (2010) who found that the variation in Meta-Programmes required students to focus on those MP’s that were less habituated in order to grow their thinking in context. It is possible that what Dunn, et al (2010) meant was that a student would gain an awareness of their intention and a choice of response would result. This idea could be tested further on a greater number of IC profiles.
Objective 5 asked if it were possible to create a benchmarking tool to normalise the Identity Compass output. As mentioned in the Measures section, this was achieved by mapping the relevant Meta-Programmes to the behavioural output of Kegan’s Levels of Adult Development (see Appendix 6) from a social-emotional perspective, and Laske’s Cognitive Development Framework from a cognitive perspective. This mapping allowed an inference to a student’s underlying Intention, Awareness, Choice and Response as a function of their MP combinations. This provided a score against which the four attributes were measured.
Objective 3 asked if a similar or different subset of dominant Meta-Programmes emerged from the larger student data. When compared with Table 3.7 from the exploratory study, the combination of CI’s was clearly different. The obvious difference being the number of student profiles in the data (pilot n=32; current n=177). Although each study found a subset of Meta-Programmes, the current study had five dimensions and the pilot study had 4. What was interesting was the principle found in the combination of MP’s in each dimension remained the same. In the pilot study, dimension 1 contained the majority of Meta-Programmes, suggesting that all profiles were influenced by more MP’s than in study 2. However, those MP’s in dimension 3 in each study were the polar pairs to those contained in dimension 1. This suggested that in each study, the principles of disequilibrium as put forward by Piaget (1964) were supported by the fact that should a student use (unconsciously) an MP in dimension 1, then a focus on the opposing MP would be necessary in order to balance their thinking. The two factor analyses supported this principle and thus supported the findings of Dunn et al. (2010).
The current study only included profiles of post-graduate students at one university in the UK, and thus any conclusion on their thinking or their Thinking Style would be premature without the existence of a control group for comparison. With a number of psychological concepts emerging from the current study, a larger dataset investigation of the same principles is recommended to determine if the same concepts/hypotheses can be found, taking into account the need for a control group.
Should the same objectives and hypotheses be applied to a larger, more diverse dataset, the results might differ. This is something to be considered as the basis for a third quantitative study.
Objective 4 asked if any other unique patterns of Meta-Programmes emerged from the data. From 174 profiles, the results have demonstrated that there were too few records to discern unique combinations of Meta-Programmes sufficient to be considered Thinking Styles. However, what was interesting in this second study was the combination offered by the factor analysis in relation to Piagetian schemata (1965), and Kegan’s (1994) Levels of Adult Development. The data for the TQ score was non-significant in the current study, however, there was suggestion that the TQ score was different because the combination of Meta-Programmes within each was different. To test this further would require a much larger dataset. For example: it might be that a post-graduate student is more likely to be higher on the TQ scale than someone who does not attain the same level of education for a reason other than CI combinations, such as General Intelligence. This would also be investigated in a larger dataset.
The combination of the Meta-Programmes allowed a student to construct different meaning-making in their thinking process as their awareness of their intention increased. Inferred from Kegan (1994), this suggested that their experience of being a student would vary depending on where they found themselves on the TQ scale, because the degree of choice of behaviour differs at each TQ level. Thus, the hypothesis was that lower TQ-level thinkers had less awareness of their Meta-Programme use in context (academia). Further study on a greater number of participants would offer a more profound understanding of this hypothesis and allow the TQ score to develop into a more robust system of measuring awareness in the moment.
The standard definition of metacognition is one of learning strategies and a learning process requiring awareness of each to be successful in the classroom (Flavell, 1979). Meta-Programmes are thus not metacognition as they offer no such educational process or strategy. However, they do offer a way of thinking about one’s thinking from an intention and awareness perspective once fed back to the participant, as discussed. From a post-graduate and thus self-directed learning perspective, students must be able to assess their own knowledge relative to the research, which involves metacognitive knowledge (Hmelo et al, 1997). What metacognition does not offer, and was thus a differentiator for this study, was a definition of the intention behind a student’s thinking that determined where they placed their attention. For example: if a student unconsciously used ‘Sameness’ as a heuristic, this impacted how they approached assignments in comparison to a student who used ‘Difference’. This direction of intention (comparison) is not accounted for by developmental psychologists in the manner it is described here as a thinking shortcut.
The results of the current study demonstrated the importance of a student’s thinking when comparing the output of two individual meta-programmes, such as ‘Internal’ and ‘External’. According to the literature, it was apparent that the Identity Compass, and other similar tools (Daniels, 2010) measured what they purported to measure: Meta-Programmes. However, from the current study, it was evident that the Identity Compass, when understood differently, was capable of measuring more than the standard definition Meta-Programmes. Arguably, it was also measuring a participant’s Intention and Awareness. Consequently, it could be argued that the label of “Meta-Programme” is a misnomer, and a more functional label for the fifty Meta-Programmes uncovered by the Identity Compass is: ‘Cognitive Intentions’.
It was understood that they could also be considered and named cognitive biases. However, it was preferred here to use ‘Intention’ in order to differentiate from other meaning-making systems. There is an argument for meaning-making that needs to be considered when describing any phenomenon, and that is ‘essentialism’. This had an impact on the reasons behind the reframing and renaming of Meta-Programmes. One needs to be careful when naming categories of thinking, elements of intellect or facets of behaviour, to ensure the perceiver is not convinced of some deeper reality that simply is not true (Barsalou, Wilson, & Hasenkamp, 2010). A concept contains the essential information necessary to be able to distinguish its instances from its non-instances. They are information reducers and augmenters. The process of conceptualising, or forming concepts, involves reducing complex information to simpler, more essential, and manageable forms for the purpose of facilitating other cognitive processes, such as classification, identification, storage, and retrieval. Thus, to explain and predict human conceptual behaviour, the key is to decipher the nature of the essential information that concepts encode and to discover how it is encoded. Stated in different but equivalent terms, we wish to understand how multiple MP’s are represented as one (Vigo, 2015)
Humans tend to go beyond the simple naming of an essence of something and perform an obtuse meaning-making on that ‘thing’ in order to find evidence for it. James (1890) observed this when he said: “Whenever we have made a word … to denote a certain group of phenomena, we are prone to suppose a substantive entity existing beyond the phenomena, of which the word shall be the name. (PAGE NUMBER)” He knew that naming something created an essentialism which Bloom, (2004) argues is how individuals think about events and objects in their world view. This allows the individual to give more deference to the word, rather than the object, which creates an ‘essence placeholder’, which in turn promotes essentialism. This meaning can convince the individual that there is a deeper reality to the ‘object’ in the material world (Barsalou, Wilson, & Hasenkamp, 2010; Medin & Ortony, 1989). Consequently, the context is ignored (Mesquita, Barrett & Smith, 2010).
This problem occurs in psychology so dramatically that entire fields are considered immovable and unchallengeable due to the nature of the belief behind the essentialism. This is particularly true in Trait theory, where it is sacrosanct that ‘traits do not change’ (Woods and Anderson, 2016). Behaviours, mental states, and humans are, it is believed, all determined by ‘deep, unchanging internal forces’ (Mesquita, et al., 2010).
The issue is a circular paradox in that names create meaning, yet we must name our observable phenomena. However, we must do it intelligently to avoid essentialism as a matter of course. The pervasiveness of essentialism has shaped Western psychology, where models of the mind have become fragmented due to the assumption by psychologists that emotion, memory, the self, attitudes, personality traits and more, are different entities with distinct organising properties and causes (Bruner, 1990). By concentrating on a singular mental state or behaviour, it is easy to miss its embeddedness in the larger system. From a Constructed Development perspective, this would be to focus on a single Cognitive Intention and miss the awareness process by which all fifty are measured: Dynamic Intelligence.
According to Gendron & Barrett, (2009) psychological states, behaviours and traits are not entities but events constructed out of a more basic set of processes, which are shaped by context. Feldman-Barrett, et al., (2010) goes further and states that mental events as well as behaviours are states that emerge from our immediate interactions with our environment. She calls this moment-by-moment interaction the context principle. By extension, one Cognitive Intention can serve as the context for another, hence the objective of discovering ‘driver programmes’ in the factor analysis throughout these studies. It is suggested in the current study that Gendron and Barrett’s “more basic set of processes” are those Cognitive Intentions outlined in the current study, and it is one’s level of awareness of their relationships that guide our response in the moment. Feldman-Barrett seems to be suggesting that these processes are out of awareness, and thus unchangeable. With the hypotheses outlined in this thesis, a level of awareness would allow for a change in one’s response in the moment, offering a reprieve from the context principle.
Feldman-Barrett, Mesquita & Smith, (2010) state that those observed functions of psychology, such as thoughts, feelings and actions, are not necessarily driven by single causes, but are the emergent results of multiple transactive processes. It is argued here that those transactive process are the fifty Cognitive Intentions: they precede the state that leads to the response in the moment, hence the necessary reframing from ‘Meta-Programmes’.
In support of this renaming, it should be considered whether each individual Cognitive Intention is a heuristic; a shortcut for thinking and behaving in support of Piaget’s (1971) schemata. An example would be the use of stereotypes. A stereotype is a cognitive function that serves as a predictor for an individual’s future behaviour (Snyder, Tanke and Berscheid, 1977) based on prior thoughts and experience about a particular thing or group, albeit, an overly-extended schema. From a dual process perspective (Evans, 1975), System 1 appears to be the self-regulator of reactive thought, forming quick conclusions and allowing for one to act upon impulses within the context of reality on a simpler level. Whereas System 2 is in control of deeper, less compulsive ways of filtering information, which might describe an unconscious heuristic. Many people do not monitor the output of their System 1 reasoning, and they might lack the competence to switch to more effective methods if they did (Stanovich & West, 2000). It might also be that System 1 thinking is a shortcut of the System 2 heuristics. Kahneman, (2011) states that passing judgement too quickly reinforces what the self believes as true: ‘jumping to conclusions’ bias, assumptions, and biases about what is being assessed. This points to, albeit negatively in this example, ‘Internal’ being a potential bias in one’s thinking, and a shortcut to an internal locus of evaluation.
An alternate perspective on Cognitive Intentions is their potential for conforming to the definition of a perceptual set. A perceptual set refers to a predisposition to perceive things in a certain way (Bruner, 1990). In other words, we often tend to notice only certain aspects of an object or situation whilst ignoring other details. This noticing directly aligns to the Cognitive Intentions of ‘Sameness’ and ‘Difference’, whereby it is explained by Maus (2011) and Hall (2005) that an individual initially filters for difference in general. What Hall and Maus do not explain is whether an individual is capable of performing this ‘noticing’ with a conscious intention, which further supports the need for a redefinition with a deeper understanding of what a Cognitive Intention pair achieves from an awareness perspective.
From the current study, it was shown that the combination of Cognitive Intentions, when understood in the context of heuristics and schemata, offer a deeper understanding of an individual’s deconstruction of their Thinking Style. However, this is not metacognition as the post-graduate students were not thinking about their thinking in the traditional metacognitive context, affecting learning and tasks at hand, and the strategies employed (Baker and Brown, 1980; Palincsar and Brown, 1981). As well as it being almost impossible to determine if or when a student is engaging in metacognitive behaviours (Markman, 1977), the current study also suggests that participants are not aware enough of their cognition to be able to control and monitor them in the traditional sense, but more from a complexity perspective (Markman, 1977).
Perceptual experience is about our beliefs about our environment, in that it helps to justify them, and represent our world around us in what we see and hear (Schellenberg, 2014). There are three questions that have motivated the study of perceptual experience that help us to deconstruct this: the Epistemology-question asks how our perceptual experience justifies our beliefs and yields knowledge of our environment given that perceptual experience can be misleading? The Mind-question asks whether perceptual experience brings about conscious mental states in which our environment appears or seems a certain way to us. This is what Korzybski (1951) meant when he said: ‘the map is not the territory’. Finally, the Information question asks how does a sensory system convert a myriad of informational input into mental representations that we then attribute to the world?
With these three questions in mind, and the arguments in the literature review, it is apparent that the answer could be the use of a cognitive heuristic based on our past experience of our environment. The third question hints towards a Constructivist perspective, which is the thread throughout this thesis, and which points to the potential use of Cognitive Intentions as one method of thinking construction, employed as heuristics.
It could then be further hypothesised from the Identity Compass output via the Thinking Quotient scale that the score attained by each participant (e.g. 3.1), is a scaled measure of their self-awareness, as it also pertains to the level of awareness of the participant’s use of the fifty Cognitive Intentions. This awareness is brought into the individual’s consciousness typically in the feedback process. It has already been suggested that an individual’s level of self-awareness from a Cognitive Intentions combination perspective could be tested qualitatively in a fifth study.
Finally, as the data suggests there is scope for a change in one’s habituated thinking patterns, with the use of Cognitive Intentions as the guide for this deconstruction, there is a suggestion that should the Thinking Quotient uncover a scale of self-awareness, then with the aid of the Cognitive Intentions, a guided reconstruction is possible that would affect an individual’s thinking style, which is a more constructed development approach.
Taking into account the findings in the studies so far, it would be interesting to discover if the same or similar combination of Cognitive Intentions creates the same counter-pattern (Dimension 1 versus Dimension 3) within a factor analysis that supports the principles present in the current data. Were they to, the implications on the objectives and hypotheses were such that students did not have a discernible Thinking Style different to the general population, and other styles do exist.
Table 4.14 demonstrates that the five dimensions within Table 4.16 do not predict the outcome variable TQ score, and although this addressed objective 5, it did suggest the results of the pilot study were artefactual. As per objective 3, a different combination of Cognitive Intentions was correlated with different scores on the Thinking Quotient scale, however one cannot assume causation. Again, it would require an investigation on a much larger dataset to determine a more robust relationship between CI combinations and TQ scores.
What was inferred from the results of the factor analysis, however, was an intention in the thinking of the post-graduate students to maximise their experience within academia based on the prominence of Cognitive Intentions such as ‘Information’, ‘Activity’, ‘Achievement’, ‘Towards’ and so on. Given the description of each (in Appendix 2), the intention behind their use was mirrored in both the pilot study and the current study. What is missing, and was pointed out in the literature review, is the awareness of the students that they are actually utilising a shortcut in their thinking in the first place, the lack of choice one has from this unaware state of mind, and finally the individual’s inability or capacity to respond in the moment according to the choices this awareness would offer. As it stands, there was a suggestion that post-graduate students lack the capacity to respond in the moment based on their habituated Cognitive Intention combinations. This warrants further investigation on a larger dataset.
For those students who utilised the combination of CI’s in the manner in which they are defined in the factor analysis, the data suggested that this is the standard Thinking Style for post-graduate students. What was not determined, and thus requires further study, was those students whose intention was to either stand out, or blend in, and what their specific TQ score was in relation to their CI combination. In other words, for the bespoke ‘High Flyers’ programme within Coventry University London, was there a difference in how those High Flying students construct their thinking when compared to the average student on campus?
As per the literature review, theories of learning and learning styles have been well researched in the past (Sims & Sims, 2006). It was discussed in the literature review that understanding ‘meta-programmes’ (now Cognitive Intentions) contributes to the creation of an optimal teaching and learning environment (Wium, Pitout, Human & du Toit, 2017) and conversely, ignoring those students whose Cognitive Intentions (CI) are dissimilar to those habituated CI’s of the lecturer will be less conducive to an optimal learning environment (Hawk & Shah, 2007). Where the original studies focused on an individual Cognitive Intention, the current study offers a deeper understanding of that principle by the use of combinations of CI’s for students on an MBA course at Coventry University London. It could be argued that not only do lecturers and students have differing Cognitive Intention combinations, so too does each taught subject, such as engineering, accounts, art, design and computer programming. Brown (2003) determined those CI’s that combine to produce the thinking behind accountancy modules by profiling accountancy students. By extrapolation, a computer sciences student might need to display a different combination of Cognitive Intentions to succeed at that course than if they undertook a nursing degree. With this hypothesis in mind, there is evidence that once a lecturer becomes aware of their own CI combination (Thinking Style), they could be more affective at facilitating learning (Wium, Pitout, Human & du Toit, 2017). There is thus potential to deconstruct the constituent Cognitive Intentions required in a specific module at university and align a student profile to it before their education commences, matching the post-graduate student to the course content, as Jaques achieved with organisational complexity (1989).
This would lend support to research by Chen & Zhang (2007) who discovered that the cognitive complexity of underachievers was lower than that of higher achievers. From a CDT perspective, it would be possible to predict which students would do well at university using the Thinking Quotient tool. This is a major contribution to complexity theory and justifies the use of student data for the first two studies within this research.
The results for study 2 have given five dimensions of Cognitive Intentions, with the most-available CI’s by post-graduate students being Dimension 1. The question that requires further and longitudinal research is: if ‘Achievement’ is about gaining the reward at the end of the degree, does this CI motivation diminish over time the closer the student gets to graduation? There are a vast number of combination questions such as this that could be asked and answered with a variety of longitudinal studies on CI combinations in context.
The 177 post-graduate students were recruited from one UK university, albeit across two campuses. For the purposes of identifying the students’ Cognitive Intentions, the study relied on the Identity Compass profile tool, which is predominantly an organisational tool, rather than being aimed at academia. A further possible limitation of the data was that a proportion of the participants were not native English speakers, however, they did conform to the minimum standard of English for post-graduate study in the UK. This limitation could be mitigated in future studies by offering foreign students a profile in their native language. Further to this, a comparison could be made between two Identity Compass profiles by one student in both English and their native tongue to discover not only meaning-making differences, but their self-construction differences as a national, and a foreign student. Studies using students-only were potentially less generalisable to the wider population, with opinion varying. Kardes (1996) and Lucas (2003) argued that using students was acceptable, especially in a psychology environment. However, others such as Sears (1986) and Wintre, North, and Sugar (2001), articulated apprehension about the use of university students in behavioural research. Sears went further and suggested that university students bias the output of studies as they tended not to be representative of the wider population due to having stronger cognitive skills, more flexible attitudes, more accommodating behaviour, and more transient peer group relationships than older adults. As this study was specifically designed to test for the Thinking Styles of post-graduate students, the limitations mentioned represent a potential for bias, not an actual bias. It was for this reason that a control group of non-students would be investigated next.
The main result coming from study 2 was the creation of a tool that normalised the output of the Identity Compass tool that allowed for the alignment of the combinations of Cognitive Intentions with Kegan’s (1994) Levels of Adult Development in order to demonstrate that there are different levels of self-awareness at the different levels of the TQ scale. However, the caveat here was that the data were non-significant and there was no control group, which suggests the need for a third study on the larger dataset mentioned in the introduction to either validate or refute the principles investigated in the current study.
A second important outcome was the renaming and repurposing of what were known as Meta-Programmes to Cognitive Intentions. This has the potential to impact a number of disciplines, not only the field of Neuro-Linguistic Programming, as mentioned in the literature review, but also metacognition and stage development. The consequences of this will be discussed in chapter 8.
The findings of study 2 regarding Cognitive Intention combinations were different to the pilot study, which implied that the pilot study findings were artefactual and potentially misleading. Although a positive outcome of study 2 was a specific combination of what could be termed ‘Academic Cognitive Intentions’, the first dimension in Table 4.18 cannot be considered an academic Thinking Style, rather a style of thinking found in post-graduate students. This leaves the combination open to non-academics to utilise in a different context.
What was interesting in the results of the pilot study and study 2 was that the use of Cognitive Intentions transcends learning styles (and thus academia) and offers an insight into an individual’s (domain-general) thinking strategy. With this in mind, an emphasis on general thinking and awareness would be beneficial in future studies. Another important finding was the first dimension in Table 4.18 demonstrated that the post-graduate students did not know they used those Cognitive Intentions (in context) and did not know the impact on their thinking and behaving those CI’s had individually and collectively. This implied a further study on CI awareness is necessary in order to discover the extent to which post-graduate students and the general public are unaware of their construction of self based on CI use.
Grasha, (1984) contended that in order to learn, students need to be pushed, and to avoid the potential for ‘bore-out’ (Csikszentmihalyi, 1990). Therefore, it is possible that a negative side to not knowing the student’s CI combinations (or Thinking Style) might lead to apathetic, argumentative behaviour, often mis-identified by the lecturer as laziness (Coffield et al., 2004). However, using the lens of Dynamic Intelligence, it is simply be a mismatch of lecturer/student Thinking Styles. This was an important finding as it supported previous work by Brown and Graff (2003) and Daniels (2010) as well as Pashler (2009) where the alignment of lecturer and student Thinking Styles afforded a more productive academic outcome.
From the findings demonstrated here, the relationship between Cognitive Intentions such as ‘Internal’ and ‘External’ was responsible for and accurately predicted an habituated Intention and Awareness that led to a choice of response in the moment. As demonstrated in the current study by the Thinking Quotient, once this intention and awareness was defined, it was then measured by interrogating the numerical relationship between them.
Although it was disappointing that the results of this study were not significant, due to the nature of the findings in the factor analysis, it was hypothesised that on a much larger sample of profiles, it might be possible to validate the Thinking Quotient, in order to offer a robust measure of a person’s level of awareness and positively address the research question that Dynamic Intelligence exists as a conceptual measure of self-awareness in the moment. Also, future research on each Cognitive Intention with regards to validating an intervention would be a worthwhile study, in line with Vermunt’s (1998) ideas.
Considering the supervisor plays the role of both coach and mentor to the research student (Wisker, et al., 2008), and the efficacy of each approach is different, Luecke (2004) suggests mentoring allows the transfer of tacit knowledge, whereas coaching helps the process and methodology of research (Blow, 2005), a further study on the efficacy of the supervisor/student relationship would be an interesting outcome of the current study whereby the results of good or bad supervision were investigated due to the dependency on the supervisor for the process (Yew et al., 2011).
Brown and Graff (2003) demonstrated a link between students’ Cognitive Intentions and their performance in summative assessments. However, this current study is important as they did not go so far as to measure the Cognitive Intentions from the perspective of Intention, Awareness, Choice and Response, which leaves room for a future study on post-graduate students’ awareness of their potential Thinking Style. This would expand the research by Brown and Graff. Further to this, a study on whether the different levels of the TQ score produce different degree outcomes (2:2, 2:1, First) would be interesting to the academic system. For example: in line with Chen and Zhang’s (2007) findings, is there a similar correlation between those who scored 2.5 on the TQ scale and their final degree classification of 2:2, and those who scored 3.8 on the TQ scale and their final degree classification of a 1st? This would provide further evidence that CDT was capable of standing alone as a theory of cognitive development, and thus contributing to psychology overall.
There is evidence that cultural self-awareness has an impact on a student’s well-being as well as self-identity (Lu and Wan, 2018). A future study could investigate, initially any combination of Cognitive Intentions that hints at cultural division in Thinking Styles. This would reinforce such studies as Markus & Kitayama (1991) and Phinney (1990) whose work focused on how culture contributes to and helps shape the development of self.
With further research on a larger dataset, as mentioned in the introduction, a deeper definition of the Thinking Quotient could emerge out of the various combinations of the Cognitive Intentions, offering all levels of thinker a more robust explanation for their habituated thinking and behaving in context, further validating the Theory of Constructed Development and adding to the genre of stage development and personality.
Finally, it is important to demonstrate the levels of awareness each individual has in their conscious or unconscious choice of Cognitive Intentions. These extrapolations will be investigated next in study 3.