doc toothache
04-14-2007, 09:53 PM
There have been a number of posts on SDN regarding the degree of difficulty of the DAT exam. Most are convinced that the DAT has gotten much harder. A few believe that the degree of difficulty has not changed.
DAT Hard or Easy? Harder or Easier?
The DAT is both hard and easy. It is hard for those who do not do well and it is easy for those who do well. To substantiate the claim that the DAT is harder, there are those who profess to have taken it a second time and promptly declare it to have been harder. In this case, the more appropriate phrase should be: “I just retook the DAT and it was harder” for me. Some have friends who took the DAT for the second time and give a similar verdict. There is no doubt that for those that didn’t do as well the second time around, the exam was indeed harder.
A Case of Misinformation/Misinterpretation
Many posts have been alarmed by the perceived dramatic increase in the DAT scores. One post suggested that the bio score has changed from 17 to 20. Another post suggested that the DAT test has been purposely made harder because of an overabundance of high scores (24-25). Still another (apparently based on insider info) pinpointed the change to a harder DAT to have started in the spring. One agitated OP was ready to call in the National Guard since she believed copies of the DAT exams were out and were made available to a select few of her classmates. Someone else believes that there are only a few versions (4-5) of the DAT exam.
Statistical Evidence
Published statistical data supports the fact that the DAT scores have remained relatively unchanged for both the applicant and enrollee pool. (1) (See Table) For the applicant pool, the AA scores ranged from a low of 17.2 (1999) to a high of 18.0 (2001/02). The PA scores ranged from a low of 16.5 (1998/09) to a high of 17.8 (2003/04). For the enrollee pool, the AA scores ranged from a low of 18.1 (1999) to a high of 19.3 (2006). The PA scores ranged from 17.1 (1999) to 18.4 (2006).(2) The DAT is a standardized test. The purpose of a standardized test is to have results that can be compared between students and from year to year. The results can be documented to have a relative degree of validity and reliability. The results can be generalized and replicated. We may want to think of the DAT using a chemistry analogy. The DAT is a buffer (buffered solution) where the mean scores (pH) remains relatively unchanged by addition of high or low scores (acid or base).
Scores of 24 and Above
The table shows the percentile distribution of scores of 24 and above for various sections of the DAT. (3) For the period from January 2006 to June 2006 (4), the frequencies are, admittedly, higher than expected. It remains to be seen whether or not they will stay at this level when all of the 2006 figures are taken into account. For this period, only 2600 candidates took the test, a figure which is a far cry from the ~12 K expected to have taken the test. Whether or not the percentage of scores of 24 above increases dramatically is not going to be particularly significant. The increase would be of significance only if it had an impact on the mean. This is unlikely to happen.
Number of DAT versions.
In all likelihood, only a limited number of ADA staff knows the number of versions that are available. However, it is inconceivable that an organization such as the ADA, with all the resources at its disposal, cannot come up with virtually an endless number of versions. If, for general chemistry alone, Schaum’s can come up with close to 3000 questions for the Solved Problems Series and over 1100 questions for the Schaum’s Outlines, surely the DAT come up with more than a few versions of the exam.
ADA will make the DAT harder.
It is a mystery why ADA would need to make the DAT harder. While the frequency distribution of some sections of the DAT may fluctuate from year to year, AA scores are less likely to fluctuate. Making an argument for the ADA supporting making the DAT harder is an unenviable task. Keep in mind that for TS each missed questions up to 6 decreases the STD score by a point each. (5) For a score of 24 in TS, a candidate only missed 6 questions. If the ADA wanted to bring the mean down, it could easily do so by changing the range of correct answers needed for a mean of 17 or 18.
Applicants/Enrollees Pool.
There seems to be some confusion regarding which DAT scores are under consideration. As it has been pointed out in a previous post, a distinction needs to be made between the applicant pool and the enrollee pool scores. The applicant pool mean scores are generally lower than that of the enrollee pool. As the number of applicants increases there will be an increase in the number of applicants that have scores above the mean. The net effect of choosing candidates with higher scores will be a shift of the mean to the right, or if you will, an increase in the mean scores for enrollees. In order to reach an AA mean of 20 for enrollees, the applicant pool would have to be ~25K.
Applicant Pool Enrolee Pool
Year AA PA AA PA
1998 17.3 16.5 18.2 17.2
1999 17.2 16.5 18.1 17.1
2000 17.8 17.1 18.5 17.7
2001 18 17.6 18.8 18.1
2002 18 17.7 18.8 18.1
2003 17.8 17.3 18.5 17.5
2004 17.8 16.7 18.7 17.3
2005 17.7 16.7 18.9 17.4
2006 17.97 17.46 19.29 18.45
Scores at 24 and above (Percentage)
Year Count QR RC Biology G. Chem O Chem TS AA PA
1988 2631 0.3 2.5 0.4 0.9 1.2 0.4 0.1 0.8
1994 9323 2.11 2.7 0.7 3.1 3.4 0.5 0.4 0.4
1999 6645 3 8.1 2.3 4.8 5.8 1.2 1.3 0.8
2004 11363 2.5 7.1 1.2 4.3 4.6 1.5 1.1 0.2
06 pre 2600 4.4 8.5 2.4 5.5 8 2.5 0.9 1.4
Dental Admission Test
Standard Score Analysis
2004
# of Items Mean S.D.
Quantitative Reasoning 40 16.31 3.22
Reading Comprehension 50 18.89 3.05
Biology 40 17.24 2.64
General Chemistry 30 17.15 3.47
Organic Chemistry 30 16.99 3.71
Survey of Natural Sciences 100 17.12 2.68
Perceptional Ability 75 16.35 2.44
(1) R. Weaver, U.S. Dental School Applicants and Enrollees: A Ten Year Perspective, J. Dent Educ. 2000 Sep; 64(12): 867-874.
(2) R. Weaver, U.S. Dental School Applicants and Enrollees: 2003-2004, J. Dent Educ. Sept 2005:69(9):1064-72.
(3) www.ada.org/prof/ed/testing/dat/dat_users_manual.pdf
(4) www.ada.org/prof/ed/testing/dat/dat_score_frequency.pdf
(5) www.ada.org/prof/ed/testing/dat/dat_estimate_score.pdf
DAT Hard or Easy? Harder or Easier?
The DAT is both hard and easy. It is hard for those who do not do well and it is easy for those who do well. To substantiate the claim that the DAT is harder, there are those who profess to have taken it a second time and promptly declare it to have been harder. In this case, the more appropriate phrase should be: “I just retook the DAT and it was harder” for me. Some have friends who took the DAT for the second time and give a similar verdict. There is no doubt that for those that didn’t do as well the second time around, the exam was indeed harder.
A Case of Misinformation/Misinterpretation
Many posts have been alarmed by the perceived dramatic increase in the DAT scores. One post suggested that the bio score has changed from 17 to 20. Another post suggested that the DAT test has been purposely made harder because of an overabundance of high scores (24-25). Still another (apparently based on insider info) pinpointed the change to a harder DAT to have started in the spring. One agitated OP was ready to call in the National Guard since she believed copies of the DAT exams were out and were made available to a select few of her classmates. Someone else believes that there are only a few versions (4-5) of the DAT exam.
Statistical Evidence
Published statistical data supports the fact that the DAT scores have remained relatively unchanged for both the applicant and enrollee pool. (1) (See Table) For the applicant pool, the AA scores ranged from a low of 17.2 (1999) to a high of 18.0 (2001/02). The PA scores ranged from a low of 16.5 (1998/09) to a high of 17.8 (2003/04). For the enrollee pool, the AA scores ranged from a low of 18.1 (1999) to a high of 19.3 (2006). The PA scores ranged from 17.1 (1999) to 18.4 (2006).(2) The DAT is a standardized test. The purpose of a standardized test is to have results that can be compared between students and from year to year. The results can be documented to have a relative degree of validity and reliability. The results can be generalized and replicated. We may want to think of the DAT using a chemistry analogy. The DAT is a buffer (buffered solution) where the mean scores (pH) remains relatively unchanged by addition of high or low scores (acid or base).
Scores of 24 and Above
The table shows the percentile distribution of scores of 24 and above for various sections of the DAT. (3) For the period from January 2006 to June 2006 (4), the frequencies are, admittedly, higher than expected. It remains to be seen whether or not they will stay at this level when all of the 2006 figures are taken into account. For this period, only 2600 candidates took the test, a figure which is a far cry from the ~12 K expected to have taken the test. Whether or not the percentage of scores of 24 above increases dramatically is not going to be particularly significant. The increase would be of significance only if it had an impact on the mean. This is unlikely to happen.
Number of DAT versions.
In all likelihood, only a limited number of ADA staff knows the number of versions that are available. However, it is inconceivable that an organization such as the ADA, with all the resources at its disposal, cannot come up with virtually an endless number of versions. If, for general chemistry alone, Schaum’s can come up with close to 3000 questions for the Solved Problems Series and over 1100 questions for the Schaum’s Outlines, surely the DAT come up with more than a few versions of the exam.
ADA will make the DAT harder.
It is a mystery why ADA would need to make the DAT harder. While the frequency distribution of some sections of the DAT may fluctuate from year to year, AA scores are less likely to fluctuate. Making an argument for the ADA supporting making the DAT harder is an unenviable task. Keep in mind that for TS each missed questions up to 6 decreases the STD score by a point each. (5) For a score of 24 in TS, a candidate only missed 6 questions. If the ADA wanted to bring the mean down, it could easily do so by changing the range of correct answers needed for a mean of 17 or 18.
Applicants/Enrollees Pool.
There seems to be some confusion regarding which DAT scores are under consideration. As it has been pointed out in a previous post, a distinction needs to be made between the applicant pool and the enrollee pool scores. The applicant pool mean scores are generally lower than that of the enrollee pool. As the number of applicants increases there will be an increase in the number of applicants that have scores above the mean. The net effect of choosing candidates with higher scores will be a shift of the mean to the right, or if you will, an increase in the mean scores for enrollees. In order to reach an AA mean of 20 for enrollees, the applicant pool would have to be ~25K.
Applicant Pool Enrolee Pool
Year AA PA AA PA
1998 17.3 16.5 18.2 17.2
1999 17.2 16.5 18.1 17.1
2000 17.8 17.1 18.5 17.7
2001 18 17.6 18.8 18.1
2002 18 17.7 18.8 18.1
2003 17.8 17.3 18.5 17.5
2004 17.8 16.7 18.7 17.3
2005 17.7 16.7 18.9 17.4
2006 17.97 17.46 19.29 18.45
Scores at 24 and above (Percentage)
Year Count QR RC Biology G. Chem O Chem TS AA PA
1988 2631 0.3 2.5 0.4 0.9 1.2 0.4 0.1 0.8
1994 9323 2.11 2.7 0.7 3.1 3.4 0.5 0.4 0.4
1999 6645 3 8.1 2.3 4.8 5.8 1.2 1.3 0.8
2004 11363 2.5 7.1 1.2 4.3 4.6 1.5 1.1 0.2
06 pre 2600 4.4 8.5 2.4 5.5 8 2.5 0.9 1.4
Dental Admission Test
Standard Score Analysis
2004
# of Items Mean S.D.
Quantitative Reasoning 40 16.31 3.22
Reading Comprehension 50 18.89 3.05
Biology 40 17.24 2.64
General Chemistry 30 17.15 3.47
Organic Chemistry 30 16.99 3.71
Survey of Natural Sciences 100 17.12 2.68
Perceptional Ability 75 16.35 2.44
(1) R. Weaver, U.S. Dental School Applicants and Enrollees: A Ten Year Perspective, J. Dent Educ. 2000 Sep; 64(12): 867-874.
(2) R. Weaver, U.S. Dental School Applicants and Enrollees: 2003-2004, J. Dent Educ. Sept 2005:69(9):1064-72.
(3) www.ada.org/prof/ed/testing/dat/dat_users_manual.pdf
(4) www.ada.org/prof/ed/testing/dat/dat_score_frequency.pdf
(5) www.ada.org/prof/ed/testing/dat/dat_estimate_score.pdf