Online Senior Assessment 2009: General Report Table of Contents EXECUTIVE SUMMARY .......................................................................................................................... 2 INTRODUCTION ........................................................................................................................................ 4 RESPONDENTS ........................................................................................................................................ 4 RESULTS .................................................................................................................................................. 10 COMPARISON: 2008 vs 2009 ............................................................................................................... 18 LIMITATIONS ........................................................................................................................................... 18 APPENDIX ................................................................................................................................................ 20 Office of Planning and Assessment, September 2009 Page 1 of 41 Online Senior Assessment 2009: General Report EXECUTIVE SUMMARY In 2009, the final sample size is 704. That means about 10% of the total population of seniors at Texas Tech are included in this analysis. This sample contains many more female students than would be expected. On the other hand, the sample seems to represent the different ethnicities in the senior population very well. There are slightly more Arts & Science students than would be expected, but overall the sample also represents the population fairly well in regards to which college the students come from. Furthermore, the differences between sample and population are statistically significant at the .05 level for GPA and age. The sample has a slightly higher GPA as well as age. The differences are not statistically significant for total transfer credits and total credits earned. There were a total of 32 so-called knowledge questions from Humanities, Multicultural, Mathematics, Natural Sciences, and Social and Behavioral Sciences on the Online Senior Assessment 2009 (i.e., questions that can be answered either correctly or incorrectly). For 26 of those questions, students who took their course for core requirement at Texas Tech, performed better than those who took the course somewhere else. The latter group performed better on only 6 questions. The difference was statistically significant at the .05 level only in 8 cases. All of these 8 cases are situations where TTU students outperformed the transfer students. When averaging the results of the knowledge questions by area first and then calculating the overall average, the average score is 62.77%. The lowest score is 14.86% and the highest score is 98%. The same results by area can be split up between students who took the respective courses for core requirement at TTU and those who took it elsewhere (“TTU” vs “Else”). This should provide some information as to if students who took the core at Texas Tech perform better or worse on average than those who took it elsewhere. Running a 2-sample t-test (a statistical test often used to check for differences in means of two different groups) shows that the means of those average scores are different from each other (statistically significant at the .05 level). The mean scores show that participants who took their core at TTU do a little bit better on average (63.19% for “TTU” vs 60.79% for “Else”). A paired t-test was run for only those students who have taken some classes for core requirements at Tech and some elsewhere (382 students). This allows us to compare if the same student performs any different depending on where the course was taken (at TTU or elsewhere). Again, the average is slightly higher for the “TTU” group, but the results below show that this difference is not significantly different at the .05 level. The average per area was compared between “TTU” and “Else” per core area (including Humanities, Multicultural, Mathematics, Natural Sciences, Social and Behavioral Sciences, and VPA). Except for Humanities, the average score is higher when the core requirement was taken at TTU. This difference is significant in all cases except for Mathematics. The difference between the slightly higher score for “Else” in Humanities is not significant as well. This means that “TTU” performs significantly better than “Else” in Multicultural, Natural Sciences, and Social and Behavioral Sciences (at the .05 level). Office of Planning and Assessment, September 2009 Page 2 of 41 Online Senior Assessment 2009: General Report Natural Sciences and Mathematics have the lowest average scores while Multicultural and Humanities have some of the highest average scores. In almost all cases the variance of performance is higher for the “Else” group. This means that the spread is wider between bad and good scores for students who took the course for core requirement elsewhere in almost all areas. The correlations between overall score of participants and both GPA and total credits earned are significant at the .05 level. Students who did well in their classes also did well on this assessment. This connection is by far the strongest. The correlation between total credits earned and overall percentage of correct answers is positive as well which means that students who have taken more classes do better on this assignment than students with fewer courses under their belt. Interestingly the correlation between performance and transferred credits is negative which would mean that the more credits a student transfers in, the worse his performance on the OSA. However, this correlation is not quite significant. The results of the Online Senior Assessment 2009 are similar to the administration of the Online Senior Assessment in 2008. However, the response rate was considerably lower in 2009 (about 10% versus about 20%). This could be due to the shorter surveys and/or the academic probation of Texas Tech in 2008. When comparing the averages of each area from 2008 and 2009, it looks like overall the results are very similar. The biggest change is that the VPA averages seem to have increased for both “TTU” and “Else”. However, it is important to note that the questions on this area changed, so maybe they did not capture the same knowledge as they did last year. Another reason might be that the scoring rubric changed. However, it is also possible that the students performed better. Overall, students still seem to have the lowest average scores in Mathematics and the highest scores in the Natural Sciences and the Humanities. Some limitations include (1) the participants have no incentive to do well, (2) the participants might not know where they took their courses for core requirement, (3) few questions per core area, and (4) there is no information about where students took their courses if they did not take them at Texas Tech. Office of Planning and Assessment, September 2009 Page 3 of 41 Online Senior Assessment 2009: General Report INTRODUCTION The Online Senior Assessment (OSA) was designed in 2008 to assess core-curriculum knowledge and abilities. In 2009, it was administered for the second time to all 7110 graduating seniors (i.e., students having 90 or more credit hours) between 3/4/09 and 4/10/09. A total of 758 students participated leading to an initial response rate of 10.66%. Three of the participants were randomly selected to win $1,500 toward tuition and fees. The instrument has one section for each of the following areas: ‐ Humanities (3 questions self-assessment, 4 questions) ‐ Multicultural (7 questions) ‐ Mathematics (5 questions) ‐ Natural Sciences (6 questions) ‐ Technology and Applied Sciences (4 questions) ‐ Social and Behavioral Sciences (10 questions) ‐ Visual and Performing Arts (4 questions) Please see attachment A for the full instrument. RESPONDENTS Inquisite, the survey software used, provides information about when students first started the survey and when they finally submitted it. The following graph shows how much time the participants spent on the survey graphically (hours:minutes:seconds) while the tables below provides some more details (values in seconds). Office of Planning and Assessment, September 2009 Page 4 of 41 Online Senior Assessment 2009: General Report The minimum time spent on the survey was 2 minutes and 57 seconds and the maximum time was 10 days 5 hours 58 minutes and 41 seconds. The average time participants spent on the survey is 2 hours and 57 minutes. However, as the chart above shows, this number is elevated because of the large outliers to the right. These outliers in all likelihood were students who started the survey and did not get back to it until a few days later. Restricting the range 3 hours (in the chart below) shows that most values actually cluster around 20 to 30 minutes. These participants were not actually deleted from the analysis. They are just not shown in the graph below which allows it to show more details on the range most students fall in. There were 54 participants who took less than 15 minutes for the whole instrument. They were taken out before any further analysis was undertaken since it takes about as long just to read through the instrument. Inquisite forces students to choose an answer before moving on to the next section, so in all likelihood, these students just clicked on any answer to get to the end where they could submit their name for the chance to win $1,500 toward tuition and fees. Eliminating these respondents leads to a final sample size of 704 students (i.e. 9.9%). Given that the final response rate is only about 10%, it is questionable if it is representative of the population. A look at the following variables should determine the answer to this question: gender, ethnicity, college, GPA, transfer credits, total credits earned, and age. Office of Planning and Assessment, September 2009 Page 5 of 41 Online Senior Assessment 2009: General Report The charts below compare the number of female and male students of the sample to the actual population of all seniors with at least 90 credit hours (F = female, M = male, N = no answer). As can be seen, in the population the split between male and female students is almost exactly reversed. Therefore, the sample has disproportionately many females. Sample Population The charts below compare the distribution of ethnicity of the sample and the actual population. The sample seems to mirror the population which means that in this regard our sample seems to be representative of the actual population. Sample Population Office of Planning and Assessment, September 2009 Page 6 of 41 Online Senior Assessment 2009: General Report The charts below compare which college the participants come from to the actual population. The charts look very similar which points to a good representativeness of the sample. The only difference that seems fairly large is between students from Arts and Sciences. There seem to be disproportionately many in the sample. Sample Population The chart below shows the average (mean), standard deviation (std dev), number of observations (N), minimum, and maximum for the sample for the following variables: (1) cumulative GPA, (2) total transfer credits, (3) total credits earned, and (4) age. The chart below provides the same information for the population. Office of Planning and Assessment, September 2009 Page 7 of 41 Online Senior Assessment 2009: General Report Sample and population seem to be very similar along those four characteristics. The average GPA is a little higher in the sample and the participants of the sample have a lower average of transfer credits. The participants have a slightly higher average number of total credits earned and are a little bit older. The charts below compare the distribution of GPA of the sample and the actual population. The main difference is that the largest group of students in the sample is the one with the highest GPA (around 3.9) while the largest group in the population is has a GPA around 3.075. Sample Population The charts below compare the distribution of total transfer credits of the sample and the actual population. They look almost identical. Sample Population Office of Planning and Assessment, September 2009 Page 8 of 41 Online Senior Assessment 2009: General Report The charts below compare the distribution of total credits earned of the sample and the actual population. It looks like the sample has slightly more people that have already completed more credit hours than the population. Sample Population The charts below compare the distribution of age of the sample and the actual population. It shows that the distributions are very similar. Sample Population A one-sample z-test (a statistical test often used to compare the sample to the population when the actual population is known) was performed to see if these differences are significant (see attachment B for details). The differences between sample and population are statistically significant at the .05 level for GPA and age. The differences are not statistically significant for total transfer credits and total credits earned. The sample size is very large here and so even small differences can be statistically significant. However, practically speaking, even the statistically significant differences look very small. Overall, the sample seems to represent the population fairly well along the chosen criteria. The biggest differences are between gender and GPA. Some measures should Office of Planning and Assessment, September 2009 Page 9 of 41 Online Senior Assessment 2009: General Report be put into place to encourage more male students to participate in the assessment next time. Also, students that do not have top 10% GPA’s should be encouraged to participate. Maybe there is a way to phrase the invitation in a way to make it less intimidating to those students who don’t perform at the top of their class. RESULTS One of the main questions the Online Senior Assessment can help answer is if students who take their classes for core requirement at institutions other than Texas Tech (i.e., the transfer students) perform similar to students who took their classes for core requirement at Texas Tech (i.e., the resident students). Whenever those two groups are compared in the analysis in this report, the group “Else” contains the students who have transferred in credits for core requirements and the group “TTU” stands for all the students who took their core requirement courses at Texas Tech. The data was summarized to show the number of people who had correct and incorrect answers to each of the 32 knowledge questions from the different areas. A Chi-squared test for differences in probabilities (a statistical test used to test for differences when the variables are categorical) was run on each question to test if the difference of performance between “TTU” and “Else” is statistically significant at the .05 level, i.e. if there is a relationship between where students took the course for core requirement and the number of correct answers to the questions (see attachment C for details). The chart below shows the difference between the percentage of the “TTU” group that had the correct answer and the percentage of the “Else” group that had the correct answer. Positive numbers (i.e., the bars to the right) represent all questions where “TTU” had a higher proportion of correct answers (for 26 questions) and negative numbers (i.e., the bars to the left) represent where “Else” had the higher proportion of correct answers (only 6 questions). The columns in red show where the difference was statistically significant, i.e. where a relationship exists between where the student took the course and the number of correct answers for that specific question. It is interesting that quite a few instances where “TTU” performed better are significant while the few instances where “Else” outperformed “TTU” are all not significant. The instances where the difference was significant could point to questions that are better at distinguishing between the two populations than the other questions. Office of Planning and Assessment, September 2009 Page 10 of 41 Online Senior Assessment 2009: General Report Social and Behavioral Sciences 2 Mathematics 5 Social and Behavioral Sciences 3 Multicultural 5 Natural Sciences 5 Multicultural 6 Natural Sciences 2 Multicultural 2 Social and Behavioral Sciences 1 Mathematics 4 Social and Behavioral Sciences 10 Social and Behavioral Sciences 8 Natural Sciences 3 Natural Sciences 6 Mathematics 1 Natural Sciences 4 Mathematics 2 Multicultural 1 Social and Behavioral Sciences 9 Multicultural 7 Humanities 4 Social and Behavioral Sciences 5 Humanities 3 Natural Sciences 1 Social and Behavioral Sciences 6 Humanities 1 Multicultural 3 Social and Behavioral Sciences 7 Social and Behavioral Sciences 4 Mathematics 3 Multicultural 4 Humanities 2 ‐10% ‐5% 0% 5% Office of Planning and Assessment, September 2009 Page 11 of 41 10% 15% Online Senior Assessment 2009: General Report The tables below summarize how students performed overall in the different areas (i.e., the percentage of correct answers). Only the 32 questions from Humanities, Multicultural, Mathematics, Natural Sciences, and Social and Behavioral Sciences where one correct answer exists (i.e., the knowledge questions) were included in this analysis. The self-assessment questions from Humanities & Technology and Applied Sciences were excluded along with the VPA questions since neither of those can be either right or wrong. There were a total of 704 participants and the average score is 61.73% which is disappointing because it would not even be a passing score. The lowest score is 18.75% and the highest score is 96.88%. The graph below displays that same information. It shows that most students have average scores between 60% and 70%. It looks like more students fall to the left of that which means their scores were actually lower than 60%. Office of Planning and Assessment, September 2009 Page 12 of 41 Online Senior Assessment 2009: General Report Since each area has a different number of questions, the results above are influenced more by those areas with more questions (e.g., Social and Behavioral Sciences and Multicultural) and less by those with fewer questions (e.g., Humanities and Mathematics). The tables below provide information on how the students performed when the information is averaged by area first. The average score is slightly higher (62.77%), the lowest score is lower (14.86%), and the highest score is higher (98%). The results are almost the same when they are averaged by area first. The graph below shows that the whole distribution seems very similar even though it is a little wider (both more lower scores as well as some higher scores). It looks like a few more people have higher scores when the results are displayed this way. Office of Planning and Assessment, September 2009 Page 13 of 41 Online Senior Assessment 2009: General Report The same results by area can be split up between people who took the respective courses for core requirement at TTU and those who took it elsewhere (“TTU” vs “Else”). This should provide some information as to if students who took the core at Texas Tech perform better or worse on average than those who took it elsewhere. Since it is possible for one and the same student to take one course for one area at Texas Tech (e.g., Humanities) and another course for another area somewhere else (e.g., Mathematics), there can be a percentage in both groups for the same student. The group “TTU” therefore averages the scores of those areas where the core requirement was taken at Texas Tech while the group ”Else” averages scores of those areas where the core requirement was taken elsewhere. The two charts below compare the percentage of correct answers for “TTU” (left) and “Else” (right). While the scores for “TTU” clearly center around 60% – 70%, the results for “Else” are much less focused. There seem to be many lower scores as well as some very high ones. The middle is not nearly as well defined as in the “TTU” group. “TTU” “Else” Running a 2-sample t-test (a statistical test often used to check for differences in means of two different groups) shows that the means of those percentages are different from each other (statistically significant at the .05 level). The mean scores show that participants who took their core at TTU do a little bit better on average (63.19% for “TTU” vs 60.79% for “Else”). See below for the details. Office of Planning and Assessment, September 2009 Page 14 of 41 Online Senior Assessment 2009: General Report Again, due to the sample size even small differences can be statistically significant. The graph below shows how similar the results are overall though. The lowest point depicts the lowest average score, the second lowest point (where the blue starts) is the lower quartile, the third point (in the middle of the blue box) shows the mean, the fourth point (where the blue ends) shows the upper quartile, and the last point is the maximum score. Even though the correlation between the two groups is very low (.16; see attachment D), there is an obvious violation of independence (participants can have scores in both groups). Since independence is such an important assumptions when using a 2-sample t-test, a paired t-test was run for only those students who have taken some classes for core requirements at Tech and some elsewhere (382 students). This allows us to compare if the same student performs any different depending on where the course was taken (at TTU or elsewhere). Again, the average is slightly higher for the “TTU” group, but the results below show that this difference is not significantly different at the .05 level. Office of Planning and Assessment, September 2009 Page 15 of 41 Online Senior Assessment 2009: General Report The table below compares the differences between “TTU” and “Else” per core area. Except for Humanities, the average score is higher when the core requirement was taken at TTU. This difference is significant in all cases except for Mathematics. The difference between the slightly higher score for “Else” in Humanities is not significant as well. This means that “TTU” performs significantly better than “Else” in Multicultural, Natural Sciences, and Social and Behavioral Sciences (at the .05 level). This could point to the possibility that taking a course for core requirement in these areas at TTU (versus taking it elsewhere) helps the students to perform better. However, this conclusion assumes that the questions in these areas accurately capture the knowledge that students are supposed to have in these areas. Core Elsewhere Core Area Humanities Multicultural Mathematics Natural Sciences Social and Behavioral Sciences VPA N 178 148 134 201 208 64 Core at TTU mean stdev N mean stdev 0.73736 0.2714 526 0.731464 0.2512 0.72008 0.1899 556 0.758736 0.1775 0.40149 0.22 570 0.441754 0.2479 0.6675 0.2258 503 0.717362 0.1999 0.48894 0.60208 0.1706 0.1743 496 135 0.53004 0.658683 0.1544 0.162 pvalue 0.791 0.0207 0.0846 0.0041 0.0019 0.0258 The following chart shows the mean with a 95% confidence interval for each of the areas above as well as an added area, Visual and Performing Arts (VPA). The mean is shown as the line in the little box and the confidence interval is depicted as the little blue box. For all areas except VPA, the percentage of correct answers is shown while for VPA the percentage grade averaged for all three questions is shown. For each area the data is shown for “Else” and “TTU”. Natural Sciences and Math have the lowest average scores while Multicultural and Humanities have some of the highest average scores. In almost all cases the variance of performance is higher for the “Else” group (as indicated by a wider confidence interval). This means that the spread is wider between bad and good scores for students who took the course for core requirement elsewhere in almost all areas. The “TTU” group performs much more coherently which could point to the possibility of the education at TTU in the different courses being of similar standards regardless which course was taken for the core. Again, this assumes the questions accurately capture knowledge students are supposed to achieve in the various core areas. Office of Planning and Assessment, September 2009 Page 16 of 41 Online Senior Assessment 2009: General Report VPA Else VPA TTU Social and Behavioral Sciences Else Social and Behavioral Sciences TTU Natural Science Else Natural Sciences TTU Mathematics Else Mathematics TTU Multicultural Else Multicultural TTU Humanities Else Humanities TTU 0.00% 20.00% 40.00% 60.00% 80.00% 100.00% The table below shows the correlation between the overall score of participants with some of the variables used to determine if the sample was representative. The correlations are displayed below. The correlations with both GPA and total credits earned are significant. Students who did well in their classes did so as well on this assessment. This connection is by far the strongest of all four variables. The correlation between total credits earned and overall percentage of correct answers is positive as well which means that students who have taken more classes do better on this assignment than students with fewer courses under their belt. Interestingly the correlation between performance and transferred credits is negative which would mean that the more credits a student transfers in, the worse his performance on the OSA. However, this correlation is not quite significant. Office of Planning and Assessment, September 2009 Page 17 of 41 Online Senior Assessment 2009: General Report COMPARISON: 2008 vs 2009 The response rate in 2008 was about 20% which is much higher than 2009’s 10%. The incentives were the same and the survey was open almost 2 weeks longer in 2009 than in 2008. In 2008, students were invited to participate in the end of February while in 2009 invitations went out 5 days later (in the beginning of March). This difference is not so large that it should cause such a difference. Spring Break was March 16-20 in 2009 while in 2008 it was March 17-21. So, the difference there is minimal again. One of the obvious differences is the length of the survey. Since students only answered questions from two separate subject areas in 2008 as opposed to questions from all areas in 2009, the instrument was much longer in 2009. The estimated time to take it might have scared some students off that would have otherwise participated. Another reason could be that when this assessment was offered in 2008, students were very aware that Texas Tech was on academic probation by SACS-COC (Southern Association of Colleges and Schools – Commission on Colleges). They more than likely felt more pressure to help Texas Tech with its assessment efforts. When comparing the number of times “TTU” outperformed “Else” on the different questions in the 2008 and 2009, it is interesting to note that while in both years “Else” performed better than “TTU” on exactly 6 questions, these questions were not actually the same. In fact, some are even from different areas. However, this is not surprising since none of these differences were significant in 2009. When comparing the averages of each area from 2008 and 2009, it looks like overall the results are similar. The biggest change is that the VPA averages seem to have increased for both “TTU” and “Else”. However, it is important to note that the questions on this area changed, so maybe they did not capture the same knowledge as they did last year. Another reason might be that the scoring rubric changed. However, it is also possible that the students performed better. Overall, students still seem to have the lowest average scores in Mathematics and the highest scores in the Natural Sciences in the Humanities. LIMITATIONS There seem to be two major limitations with the way the Online Senior Assessment is administered. First of all, students really have no incentive to do well. And as the analysis of how much time students spend on the assessment shows, many students seem to click through rather fast and do not seem to take much time to read through the questions and think about the answers. It would be very helpful if in the future this could be changed somehow. The incentive structure could be changed for example. While now, every participant is entered into a drawing for the tuition waver, it could be structured to where only the students who pass a certain threshold (e.g., a “passing grade” of 70%) are entered into the drawing. Until this is changed, there is no way of knowing if students did better because they know more or because they cared more about giving the right answer. Office of Planning and Assessment, September 2009 Page 18 of 41 Online Senior Assessment 2009: General Report The second major limitation seems to be the assumption that students actually know where they took the classes for the different areas of the core curriculum. While this is easy for students who transferred to TTU after meeting all their core requirements or for students who took all their classes at TTU, a large group of the students take some classes at TTU and some classes at other institutions. It is questionable if they know which class counted for core curriculum requirements and which class was just part of their regular curriculum. To expect students to be able to self-identify into the groups “TTU” or “Elsewhere” is a very important assumption since all the results are based on correct selection. If students are not able to select the correct group, all conclusions drawn concerning which group does better are not valid. One potential improvement could be to ask leading questions as opposed to only one question. For example, the students could be asked “Have you taken any classes at any other institution than Texas Tech?”. If the answer is yes, the next question could be “Which area was this class in?”. The answer choices could be for example biology, physics, theater, etc. The next question could be “Did any of these classes count toward your core requirement?”. Most importantly however, the answer options to this question should include an “I don’t know”. That way students at least have the option of saying that they cannot provide us with this information. Another potential limitation is that there are very few questions per core area and those do not necessarily represent each subject a student could have taken as part of the core curriculum For example, the questions in Natural Sciences seem very focused on biology. A student could have simply taken other classes that are part of Natural Sciences and performed badly on this assessment. Also, the group “Else” is far from homogeneous. Students in this group could have taken their classes anywhere from a small community college to another major institution. It might be worthwhile to explore the differences in this group further. Office of Planning and Assessment, September 2009 Page 19 of 41 Online Senior Assessment 2009: General Report APPENDIX Attachment A: The Instrument Office of Planning and Assessment, September 2009 Page 20 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 21 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 22 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 23 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 24 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 25 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 26 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 27 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 28 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 29 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 30 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 31 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 32 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 33 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 34 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 35 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 36 of 41 Online Senior Assessment 2009: General Report Attachment B: Sample Statistics and Z-Tests Office of Planning and Assessment, September 2009 Page 37 of 41 Online Senior Assessment 2009: General Report Office of Planning and Assessment, September 2009 Page 38 of 41 Online Senior Assessment 2009: General Report Attachment C: Knowledge Questions n= 178 n= 526 Else correct TTU incorrect correct incorrect Chi-Square Probability Humanities 1 116 62 348 178 0.8094 Humanities 2 94 84 242 284 0.1163 Humanities 3 157 21 471 55 0.6181 Humanities 4 158 20 478 48 0.41 n= 148 n= 556 Else correct TTU incorrect correct incorrect Chi-Square Probability Multicultural 1 100 48 393 163 0.4621 Multicultural 2 129 19 528 28 0.0007 Multicultural 3 140 8 520 36 0.6329 Multicultural 4 71 77 255 301 0.6474 Multicultural 5 100 48 424 132 0.0312 Multicultural 6 100 48 421 135 0.0445 Multicultural 7 106 42 412 144 0.5433 n= 134 n= 570 Else correct TTU incorrect Chi-Square Probability Mathematics 1 46 incorrect 88 218 352 0.3993 Mathematics 2 95 39 423 147 0.4335 Mathematics 3 62 72 253 317 0.6933 Mathematics 4 35 99 180 390 0.2169 Mathematics 5 31 103 185 385 0.0353 n= 201 correct n= 503 Else correct TTU incorrect correct incorrect Chi-Square Probability Natural Sciences 1 88 113 226 277 0.7817 Natural Sciences 2 86 115 255 248 0.0579 Natural Sciences 3 134 67 357 146 0.2611 Natural Sciences 4 170 31 444 59 0.185 Natural Sciences 5 156 45 434 69 0.0048 Natural Sciences 6 171 30 449 54 0.1214 Office of Planning and Assessment, September 2009 Page 39 of 41 Online Senior Assessment 2009: General Report n= 208 n= 496 Else correct TTU incorrect correct incorrect Chi-Square Probability Social and Behavioral Sciences 1 115 93 305 191 0.1258 Social and Behavioral Sciences 2 136 72 395 101 <.0001 Social and Behavioral Sciences 3 71 137 213 283 0.0297 Social and Behavioral Sciences 4 21 187 41 455 0.4344 Social and Behavioral Sciences 5 80 128 201 295 0.6101 Social and Behavioral Sciences 6 112 96 272 224 0.8093 Social and Behavioral Sciences 7 62 146 139 357 0.6326 Social and Behavioral Sciences 8 103 105 269 227 0.2529 Social and Behavioral Sciences 9 199 9 489 7 0.0179 Social and Behavioral Sciences 10 118 90 305 191 0.2392 Office of Planning and Assessment, September 2009 Page 40 of 41 Online Senior Assessment 2009: General Report Attachment D: Correlation “TTU” and “Else” Office of Planning and Assessment, September 2009 Page 41 of 41