We have previously shown that the CI model correlates with teaching staff and peer estimates of professionalism in undergraduate students in a UK medical school [10
] and have previously reported that in a single cohort of students there was a small increase in CI % [11
] which could reflect an improvement in conscientious behaviour which may be attributable to teaching and clinical exposure. However, we have now expanded this study to include 3 consecutive cohorts thus considerably increasing the pool of students in which conscientiousness points were observed. There was some minor variation between cohorts but when the data is pooled there is no significant change in CI % scores between years 1 and 2 of studies. Our findings in this paper demonstrate that conscientious behaviour in medical students, as measured by using a conscientiousness index, does not significantly change from year 1 to year 2 of studies suggesting that the CI is measuring the stable trait of conscientiousness and is not modified by teaching.
The trait of conscientiousness is listed as one of the ‘Big Five’ domains of personality [14
] and has been shown to positively correlate to exam performance in UK university students when measured qualitatively by personal inventory [15
]. Similar approaches have found that conscientiousness significantly predicts final scores in examinations in pre-clinical medical students with those students who scored low in conscientiousness more likely to gain an unsuccessful outcome [16
]. Conscientiousness is also a good predictor of job performance [17
] and this may be of particular importance in the medical profession where as previously mentioned a lack of conscientious behaviour has been associated with unprofessional behaviour in clinical practice [13
]. Other ‘Big Five’ personality traits may also have a role to play as predictors of professionalism such as agreeableness, as could some facets of traits such as warmth and positive emotions (extroversion) and ideas (openness to experience). However, the problem arises here of how to easily and reliably measure other traits and facets as this is normally attempted by qualitative methods by use of questionnaires. The collection of CI points producing a scalar measure of conscientiousness provides a situated measure of this behaviour in the context of the early stages of study in undergraduate medicine. It is possible that the CI may also be measuring a different aspect of this trait compared to the more generalised measures of the ‘Big Five’ personal inventory. However, even if the CI is measuring something ‘different’ the findings presented in this study, together with previously publish data, suggest the CI model may have an important role in identifying early on a lack of conscientious behaviour in medical students [10
]. Further validation of the reliability of the CI may be achieved by adoption of similar approaches in other professions where identifying problematic behaviour at an early stage would be of benefit. Advantages of employing the CI model are its ease of collection, production of a quantitative outcome allowing direct comparison between students and the fact that it records actual student behaviour. Conscientiousness may not be the only personality trait that contributes towards professional behaviour but it is one that is perhaps the most important and easy to measure [10
]. Although the CI may not directly measure professional behaviour it may play an important role in identifying individuals more likely to exhibit negative behaviour in the workplace.
CI % scores were generally high, with mean values of ~90% for both years 1 and 2 for the combined data from all three cohorts (Figures and ; Table ) which suggests that medical students are on the whole a highly conscientious grouping. Overall, in addition to CI being reported to be stable within a single academic year [10
] this study shows similar negatively skewed distributions for both years 1 and 2 of study (Figure ). This suggests the CI % is stable and although there are different components contributing to the final CI points totals, a similar trait of conscientiousness is being measured in both years 1 and 2 for each cohort. This is further reinforced by the significant positive correlation of year 1 and year 2 CI % scores observed in Figure and Table . The observed correlation of 0.54 for the combined cohorts (Table ) would represent a significant effect size in educational terms [19
]. The remaining variance may reflect the degree of change between individuals and cohorts, as in one cohort the mean CI % scores show a small increase and in the other two cohorts it decreases slightly with a net overall effect of virtually no change for the combined cohorts (Figure ; Table ). This provides further evidence for the reliability of the CI as a measure of conscientious behaviour and raises the interesting question of why the mean CI % scores did not indeed increase in year 2 overall.
It may be argued that an increase in mean CI % scores would be expected to occur in year 2, as previously reported [11
], as students are exposed to further teaching, are further exposed to the concept of professional behaviour in doctors and their peers and experience more patient contact. However, as personality is thought to be stable during adulthood [20
] conscientiousness may be difficult to modify in medical students without specific intervention. This could potentially raise the issue of preventing future problems of negative behaviour in practice by screening out likely individuals with low conscientiousness at early stages of the undergraduate curriculum or even at the admissions stage to medical school. Using conscientiousness as a screening mechanism during the admissions process would require a CI model to have been implemented at the students previous educational institution or workplace. This may be difficult to achieve in secondary education in the UK (normally up to age 18), with the reliability of the data also questionable as personality is reported as being stable post 18 years [20
], but would be less problematic for graduate students where CI data could more easily be collected. It is also possible that applicants to medical schools could fake the desired personality traits at interview but use of a CI model would measure actual conscientiousness longitudinally and not just that displayed on a single occasion. This could be used as an argument for offering places at medical school only to graduate students with a satisfactory conscientiousness profile. However, we have previously reported that there was no difference in the CI % between medical students who were graduates and those that had commenced their medical studies directly after completing secondary education [11
]. One further possibility why the CI did not significantly change from year 1 to year 2 of studies is that patient contact occurs early (week 2) in the first year of the medical programme at our institution so this may act as a switching-on for the requirement of professional behaviour from an early stage.
In the 2007–8 and 2008–9 cohorts virtually all students who performed in the top half of the class for CI in year 1 show a decrease in CI % score for year 2 (data not shown) with the opposite trend seen in the 2006–7 cohort. This effect may be due to regression to the mean [21
] as the combined data shows no significant overall change in CI % scores (Table ). As conscientiousness is also linked to exam performance [16
] it is also possible that students with low a CI in year 1 also performed poorly in exams and put extra effort into year 2 studies, for example, in the form of increased attendance which would improve CI scores. The components comprising the possible total CI points available for year 2 also differs from year 1 so it may be more difficult to achieve the highest scores and a levelling out of CI points occurred. The high CI scores generally achieved by students may have been due to the relative ‘ease’ at which students could accumulate some points and may have contributed towards a ceiling effect on the data obtainable. However, the range of scores achieved (Table ) and the number of different categories of data collected (up to 19 for year 1 and 12 for year 2) suggest this is more likely to reflect that medical students are a highly conscientious group and thus achieve high CI scores. Further analysis of the components of the CI index are required to further explore this, which lies out with the scope of this study.
Although we have reported no change in CI score and hence conscientious behaviour between year 1 and year 2 of studies it has been reported that in adulthood, up to the age of ~40 years, conscientiousness, aspects of extroversion (social dominance and emotional stability), open mindedness and agreeableness all improve [22
]. This suggests that conscientiousness may indeed improve in the later years of undergraduate study and/or post-graduation. The relative components of personality that contribute towards professional behaviour may thus also vary with age. For example, does conscientiousness remain stable while other traits and facets change altering the relative contribution of conscientiousness towards professional behaviour. To provide further evidence of the validity and reliability of the CI as a measure of conscientiousness an extension of the project into the later stages of undergraduate medical training and beyond is therefore required. This would map out a students’ conscientiousness throughout their entire medical training programme. It may then be possible to determine more definitively the predictive validity of the CI in highlighting individuals at risk of experiencing postgraduate disciplinary procedures and importantly determine more precisely whether conscientiousness remains a stable characteristic throughout medical school unaffected by teaching. The use of a CI may also be of use in other professions such as teaching where it would be an advantage to flag negative behaviour in individuals at an early stage in training.