Abstract
Aims and Objectives:
Over 60% of the UK dental schools use LiftUPP. LiftUPP records the performance level of each student at every clinical encounter. Such performances are graded on the scale of 1 to 6, where 4 is described as “competent with minimal verbal prompts from the tutor”.
On average, each dental school collects 2,000,000 data points on their students per annum. Such dataset provides the faculty with the opportunity to track and monitor the clinical progress of their students in a longitudinal manner.
The aim of this study was to use the longitudinal performance data on the LiftUPP to predict the future clinical competence of the students using the mathematical Bayes’ theorem.
Materials and methods:
The study was conducted in collaboration between the three schools: School of Dentistry, The University of Liverpool, Division of Dentistry, The University of Manchester and School of Mathematic, The University of Manchester. The “tooth extraction” data for a single cohort of students over three academic years were used and filtered to include the procedural-based data only. Using the Bayes’ theorem, the data was explored to predict the likelihood of students scoring 4 or higher in future based on their past performances over time.
Results:
The initial analysis confirms that the longitudinal performance data can be used to predict future competence for the cohort; however, due to limited data points for “extraction” only items, the analysis was not possible at student-level.
Conclusions:
The initial data proves the concept of using longitudinal data for prediction of clinical competence using complex mathematical methods. The same analysis can be run on clinical items with higher frequency of occurrence to test this concept at the student-level.
Over 60% of the UK dental schools use LiftUPP. LiftUPP records the performance level of each student at every clinical encounter. Such performances are graded on the scale of 1 to 6, where 4 is described as “competent with minimal verbal prompts from the tutor”.
On average, each dental school collects 2,000,000 data points on their students per annum. Such dataset provides the faculty with the opportunity to track and monitor the clinical progress of their students in a longitudinal manner.
The aim of this study was to use the longitudinal performance data on the LiftUPP to predict the future clinical competence of the students using the mathematical Bayes’ theorem.
Materials and methods:
The study was conducted in collaboration between the three schools: School of Dentistry, The University of Liverpool, Division of Dentistry, The University of Manchester and School of Mathematic, The University of Manchester. The “tooth extraction” data for a single cohort of students over three academic years were used and filtered to include the procedural-based data only. Using the Bayes’ theorem, the data was explored to predict the likelihood of students scoring 4 or higher in future based on their past performances over time.
Results:
The initial analysis confirms that the longitudinal performance data can be used to predict future competence for the cohort; however, due to limited data points for “extraction” only items, the analysis was not possible at student-level.
Conclusions:
The initial data proves the concept of using longitudinal data for prediction of clinical competence using complex mathematical methods. The same analysis can be run on clinical items with higher frequency of occurrence to test this concept at the student-level.
| Original language | English |
|---|---|
| Publication status | Published - 24 Aug 2017 |
| Event | Association of Dental Education in Europe - Vilnius University, Vilnius, Lithuania Duration: 23 Aug 2017 → 25 Aug 2017 http://www.adee.org/meetings/vilnius2017/ |
Conference
| Conference | Association of Dental Education in Europe |
|---|---|
| Abbreviated title | ADEE |
| Country/Territory | Lithuania |
| City | Vilnius |
| Period | 23/08/17 → 25/08/17 |
| Internet address |