Graduates with the “Very Good (>2.99)” Degree
It was found that the ratio of graduates with the “very good (>2.99)” degree in Turkey was 17% in 2005 and it increased to 46% in 2020. It was also found that the ratio of those graduated with a “good (2.50-2.99)” degree increased from 35% to 40% in this period whereas the ratio of those graduated with a “moderate (2.00-2.49)” degree decreased from 48% to 14% (Graph 1).
The change in the ratio of those who graduated with a “very good” degree was examined using the stochastic limit coefficient estimations of REE (Table 1). The results showed that the coefficient, which indicated that women are more likely to graduate with a “very good” degree compared to their male peers, is statistically significant. This result is proof that being a woman has a positive and significant effect on performance. Point estimation shows that a 1% increase in the ratio of women among the students increased the average of those graduated with a “very good” degree at the ratio of 0.09% (when all other factors are constant). A significant and positive correlation was found between the percentile of the last student who entered the medical faculty (high entrance score increases as the percentile decreases) and the ratio of those graduated with a “very good” degree. For example, a 10% increase in the entrance percentiles (entrance with a lower score) increases the ratio of those graduated with a “very good” degree at the ratio of 2.5% (when all other factors are constant). This also showed that a university that accepts students from the high percentile (with a low score) will graduate 4% more students with a “very good” degree compared to a university that accepts students from the low percentile (with a high score).
The results obtained in the context of university properties can be summarized as follows. The coefficient, which indicates that it is more likely for students in non-profit private universities to graduate with a “very good” degree than students in public universities, is statistically significant. The point estimation shows that a 1% increase in the students in non-profit private universities increased the average of those graduated with a “very good” degree at the ratio of 0.57% (when all other factors are constant). The coefficient, which indicates that it is more likely for students of a nonaccredited university to graduate with a “very good” degree than students in an accredited university, is statistically significant. The point estimation shows that a 1% increase in the students of a nonaccredited university increased the average of those graduated with a “very good” degree at the ratio of 0.27% (when all other factors are constant). The coefficient, which indicates that it is more likely for students in English medium education programs to graduate with a “very good” degree than students in Turkish medium education programs, is statistically significant. The point estimation shows that a 1% increase in the students in English medium education programs increased the average of those graduated with a “very good” degree at the ratio of 0.11% (when all other factors are constant). A significant and negative correlation was found between the ages of faculties and the ratio of those who graduated with a good degree. For example, the point estimation shows that a 10% increase in the age of the faculty decreases the average of those graduated with a “very good” degree at the ratio of 0.17% (when all other factors are constant). This also showed that young faculties will graduate 6% more students with a “very good” degree compared to relatively older faculties.
A continuous increase was observed in the estimated coefficients in year dummies. Additionally, estimated coefficients for year dummies are statistically significant as of 2009 after checking the features of students and universities (p<.001). The point estimation of 200 dummy (when all other factors are constant) showed that the ratio of those graduated with a “very good” degree, which was 17% in the 2005 academic year, increased to 45% with a 25% increase (Graph 1); and this means that it explains almost all of the 29% increase in the ratio of those graduated with a “very good” degree before checking the variables. These results are proof that there is grade inflation in the pre-graduation medical training in Turkey within the last 15 years.
Table 1. Standard random effects estimates
Variable name
|
Random Effects
|
Students characteristics
|
|
ln (% Female)
|
0.09 (0.079)*
|
ln (percentile of the last student)
|
-0,25 (0.041)*
|
University characteristics
|
|
Non-profit private universities
|
0,57 (0.189)*
|
Accredited university
|
-0,27 (0.125)*
|
English medium education programs
|
0,11 (0.098)*
|
Faculty Age
|
-0,17 (0.081)*
|
σi
|
0.071
|
σe
|
0.063
|
rhoi
|
0.619
|
Within-R2
|
0.432
|
Observations
|
9.618
|
Number of universities
|
25
|
*p<0.001
Note. Robust standard errors corrected for clustering by university are reported in parentheses. σi and σe are the estimated standard deviations for the fixed effects and the error term, respectively, ρi is the fraction of the variation in the dependent variable accounted for by the fixed effects and ρ is the correlation between the fixed effects and the included variables.
Grade Inflation
While the students in the medical faculties in Turkey graduated with an average GPA of 2.41 (SD=0.29) in 2005, this average increased to 3.16 (SD=0.58) in the past 15 years (2020). This difference detected is quite high and statistically significant (t=21.37; p<.001). Accordingly, a marginal increase of 31.12% in the graduation grades between 2005 and 2020 was detected, and this finding is an indicator of high-grade inflation in the pre-graduation medical training in Turkey. Considering the grade averages, there was an increase in every year compared to the previous year (except 2007). The highest-grade inflation on yearly basis was in the grades of students who graduated in 2020 with 4.64% followed by students who graduated in 2017 with 3.62% (Graph 2).
The results of the former analysis showed that the following six factors were associated with the graduation grade: sex, the percentile of the last student who entered the program, types of the universities (public or non-profit private), program accreditation, age of the faculty and medium of the education. The average values corrected according to these six factors were calculated with the ANCOVA for each year (Graph 2). A marginal increase of 29% was detected in the graduation grades in terms of the corrected averages, but there is high-grade inflation in the pre-graduation medical training in the past 15 years even when the factors that affect the grades are taken under control.
Factors that Affect Course Grade
Class Size Differences
The total number of students in the class is at the top of factors that affect grades. The instructors had a greater opportunity to get to know students in classes with fewer students and they tend to be more flexible when they have more information about students’ efforts. Even having a closer relationship can make it less likely for the instructor to give a “bad” grade with the fear of upsetting the student 48. Thus, the correlation coefficient of the relationship between class sizes and average grades was examined. The results showed a significant and negative correlation between average grade and class size (r=-.36, p<.001). Accordingly, the fewer students in the class, the higher their average grade is.
Differences Regarding the Academic Degree of the Instructor[2]
Another one of the potential factors that affect the course grade is the academic degree of the instructor. All of the lessons in medical education in Turkey are given by instructors with a doctoral degree (Ph.D. & M.D.) And these instructors have a hierarchical structure among themselves as an assistant professor, associate professor, and professor. In the data set of the study, 71% of the lessons were given by instructors with the highest two degrees (associate professor and professor). According to the results of ANOVA, the average grade (M=2.93, , SD=0.98) in lessons given by instructors titled professors, the highest academic degree, was higher than the average grades in lessons given by associate professors (M=2.87, , SD=0.95) and assistant professors (M=2.61, , SD=0.91) (F=47.29; p<.001). In conclusion, the academic degree of the instructor affected the course grade.
Grade Differences
A potential deviation factor is the grade. The number of students is quite high in the first or second-grade courses because of the students who repeat the class and the upper-grade students who were unable to continue their classes. According to ANOVA results, there is a significant difference between the grades and average course grades (F=59.78, p<.001). The senior year (6th Grade) classes have the highest average grade (M=3.47, SD=0.54) followed by the fifth (M=3.21, SD=0.66) and fourth (M=3.01, SD=0.69) grades. The first grade (M=2.13, SD=0.89) has the lowest average grades followed by the second (M=2.41, SD=0.88) and third (M=2.74, SD=0.91) grades. In conclusion, the grade affects the course grade.
Content (Field) Differences
It is known that both the competencies expected from the students and the lessons they take are more complex and difficult according to the fields of the courses. For example, the courses in fields like surgical medicine where courses like medical pathology are predominant, are relatively more difficult than courses in basic medical sciences and require students to make more efforts. In this regard, another factor that potentially affects the grades is the field of the course. The differentiation of average grades of the courses based on their fields (Basic Medicine, Internal Medicine and Surgical Medicine) was examined with the ANOVA. The results showed that there was a difference between average course grades in terms of the field of the course (F=39.27, p<.001). The lowest grades were given in basic medicine (M=2.31, SD=0.93) while the highest grades were given in surgical medicine (M=3.07, SD=0.59). The average grade in internal medicine was 2.87 (SD=0.63). Accordingly, the field of the course affects the course grade.
University Differences
The universities in Turkey are divided into two as public and non-profit private universities. Education is free of charge in all programs of public universities (including the pre-graduation medical training). Education programs of private universities are paid. The education fee of the medical faculty in private universities changes between 58.000 TL and 151.000 TL (M=93.80, SS=23,93). However, it is mandatory for private universities to spare at least 15% of the total placement to beneficiary students for each program. Thus, another factor that potentially affects the grades is the type of university (public or private). In this regard, the differentiation of average course grades of the program based on the type of the university was examined with the t-test. The results showed that there was a difference between the grades given in public universities and grades given in private universities (t=29.33, p<.001). The course grades in private universities (M=3.11, SD=0.86) are quite higher than the course grades in public universities (M=2.54, SD=0.82). In conclusion, the type of university affected the course grade.
English medium education programs
The accreditation of medical faculties in Turkey is made by the Association for Evaluation and Accreditation of Medical Education Programs. The association is a quality agency recognized by the Higher Education Quality Committee and World Federation for Medical Education and carries out national and international medical education accreditation operations. The accreditation process of medical faculties in Turkey started in 2011, and the accredited education program is carried out in 41 of the existing 111 medical faculties (TEPDAD, 2020). Accreditation expresses the assessment and external quality assurance process that measures whether the pre-determined academic and field-specific standards are met by a higher education program. In this regard, it is an expected process for accredited higher education institutions to have a standard for both education and assessment-evaluation processes. Thus, another factor that potentially affects the grades is the accreditation of the program. In this regard, the differentiation of average course grades of the program based on whether the university is accredited was examined with the t-test. The results showed that there was a difference between the grades given in accredited programs and the grades given in non-accredited programs (t=21.93, p<.001). The course grades in non-accredited programs (M=3.11, SD=0.86) are quite higher than the course grades in accredited programs (M=2.76, SD=0.79). In conclusion, the accreditation of the program affected the course grades.
Age Differences of Faculties
The age of the faculty is closely related to various variables from the recruitment of lecturers and physical opportunities to the quality of various education programs. The ages of the faculties are at the top of factors that affect grades. The sufficiency of lecturers and physical opportunities are more limited in relatively young faculties; thus, expectations from students tend to be lower in such faculties. Thus, the correlation coefficient of the relationship between the ages of faculties and average course grades was examined. The results showed a significant and negative correlation between the age of the faculty and average grade (r=-.62, p<.001). Accordingly, the younger the faculties are, the higher the average course/lesson grade is.
Footnote:
[2]This analysis was carried out though separate courses the outside of the structure of “committee” carried out by many faculty members.