WHY? Path 2 Stats. The Challenge. Nationwide Data 256,672 first-time degree-seeking students from 57 colleges participating in Achieving the Dream

Transcription

1 11/1/2012 The Challenge Path 2 Stats The more levels of developmental courses a student must go through, the less likely that student is to ever complete college English or Math. Page 0 Bailey, Thomas. (February 2009). Rethinking Developmental Education. CCRC Brief. Community College Research Center. Teachers College, Columbia University. Page 1 Nationwide Data 256,672 first-time degree-seeking students from 57 colleges participating in Achieving the Dream Students initial math placement in developmental sequence % of students who successfully complete college level gatekeeper course in math 1 Level Below College 27% WHY? 2 Levels Below College 20% 3 Levels Below College 10% Referral, Enrollment, and Completion in Developmental Education Sequences in Community Colleges (CCRC Working Paper No. 15). By: Thomas Bailey, Dong Wook Jeong & Sung-Woo Cho. December New York: Community College Research Center, Teachers College, Columbia University. (Revised November 2009). Page 2 Page 3 1

2 11/1/2012 The Multiplication Principle Let s assume 75% success and persistence rates 100 take pre-algebra 75 pass pre-algebra 56 take algebra 42 pass algebra 32 take alg2 24 pass alg2 18 take stats 14 pass stats Every extra course a student takes adds two Page 4 Page 5 Increasing success/persistence rates does not help If this was the student s initial placement And these were the rates at which they passed each class and persisted to the next class in the sequence 70% 80% 90% 1 level below transfer 34% 51% 73% Stupifiny Shorten the pipeline! 2 levels below transfer 17% 33% 59% 3 levels below transfer 8% 21% 48% Page 6 Page 7 2

3 11/1/2012 Course Themes Assumptions Working Backward Prepare students for statistics and statistics only Learning Through Playing with Data Exploratory versus discovery Tinker Plots by Key Curriculum Press Project-Based Learning Every unit begins and ends with a project Growth Versus Fixed Mindset Brainiology: Transforming Students Motivation to Learn, Dweck Page 8 Passionate, open-minded teachers Extensive professional development A predetermined minimum skill set At least 2 hours/week in a computer lab with Tinker Plots Page 9 Students will be able to: Formulate questions that can be addressed with data, then organize, display, and analyze relevant data to address these questions and communicate results Apply the basic principles of study design to develop and analyze the validity of simple experiments and sampling plans related to a given situation and goal Students will be able to: Demonstrate numerical and algebraic reasoning skills to support statistical analysis Construct, use, and interpret mathematical models, specifically linear and exponential functions, to represent relationships in quantitative data Use effective learning strategies for success in college Page 10 Page 11 3

4 11/1/2012 How Many Successfully Completed STATISTICS IN Two Years? Results Initial COURSE Size of COHORT Number COMPLETED Rate of COMPLETION Pre Algebra % Algebra % Algebra % Path 2 Stats % Page 12 Page 13 COURSE Compare Class SUCCESS RATES Size of COHORT Rate of COMPLETION Rate of SUCCESS Pre Algebra % 65.88% Algebra % 56.87% Algebra % 54.99% Path 2 Stats % 88.33% Thank You Lily Lum City College of San Francisco llum@ccsf.edu Hal Huntsman City College of San Francisco hhuntsmanccsf@gmail.com Tue Rust Los Medanos College trust@losmedanos.edu Page 14 Page 15 4

5 SUCCESSFUL HOW MANY SUCCESSFULLY COMPLETED COMPLETION TERM COURSE COHORT Statistics OUTCOME RATE 2009FA Pre Algebra 507 BY THE END OF SUMMER % 2009FA Algebra 598 (end of Fall 2011 for 10SP cohort) % 2009FA Algebra % 09FA, 10SP Path 2 Stats % SUCCESSFUL COMPLETION TERM Placement COHORT OUTCOME RATE PATH 2 STATS DATA 09FA 10SP Total % DATA ARE DISAGREGATED BY INITIAL MATH PLACEMENT 09FA 10SP Transfer % (8 of the 120 students did not take the assessment test) 09FA 10SP Algebra % 09FA 10SP Algebra % 09FA, 10SP Pre Alg/Arith % Calculation: Successful Completion Rate = Outcome/Cohort Class Completion Rates Comparison TERM COURSE ENROLLMENT COMPLETION COMP. RATE SUCCESS SUCC. RATE 2009FA Pre Algebra % % 2009FA Algebra % % 2009FA Algebra % % 2009FA Path 2 Stats % % 2010SP Path 2 Stats % % 2010FA Path 2 Stats % % Definition Enrollment : A student is considered to have been enrolled in a course if the student received an end of term grade notation that is displayed on their official transcript. Grade notations indicating valid enrollment: A, B, C, D, F, I*, P, NP, RD, W Completion Rate (formerly Retention Rate II) : Percentage of students who do not withdraw from class and who receive a valid grade. Numerator: A, B, C, D, F, I*, NC, NP, P, RD Denominator: A, B, C, CR, D, F, I*, NC, NP, P, RD, W Success Rate: Percentage of students who receive a passing/satisfactory grade. Numerator: A, B, C, IA, IB, IC, IPP, P Denominator: A, B, C, CR, D, F, ID, IF, INP, P, NC, NP, W

6 LMC-UC Statistics Articulation Issue I. What is the rationale for developing a Statistics pathway for non-stem majors in lieu of the traditional algebra sequence? Why did Los Medanos College initiate such an apparently disruptive innovation? A. Attrition in the Remedial Math Pipeline in Community Colleges The more levels of remedial courses a student must go through, the less likely the student will ever complete a college math course. National Data: 256,672 first-time degree-seeking students from 57 community colleges participating in the Achieving the Dream initiative Students initial enrollment in remedial math sequence % of students who successfully complete a college-level math course 1 level below college (Intermediate Algebra) 27% 2 levels below college (Elementary Algebra) 20% 3 or more levels below college (Pre-algebra/Arithmetic) 10% Source: Referral, Enrollment, and Completion in Developmental Education Sequences in Community Colleges (CCRC Working Paper No. 15). By: Thomas Bailey, Dong Wook & Sung-Woo Cho. December New York: Community College Research Center, Teachers College, Columbia University. LMC data: Fall 2009 remedial math enrollments tracked through Spring 2012 Students initial enrollment in remedial math sequence % of students who successfully complete a college-level math course 1 level below college (LMC Math 30) 37% (121 of 325) 2 levels below college (LMC Math 25) 18% (57 of 315) 3 or more levels below college (LMC Math 4,9, and 12) 10% (64 of 619) Source: Basic Skills Cohort Tracker, DataMart Data includes repeaters. Attrition in the remedial math pipeline has a disproportionate impact on students of color. For example, in California, black and Latino students are much more likely to be placed 3-4 levels below college math. According to a 2010 EdSource study Course taking patterns, policies, and practices in developmental education in California Community Colleges by Perry, Bahr, Rosin, and Woodward, 61% of Black students and 53% of Latino students placed 3-4 levels below college math in California community colleges, compared to 34% of White students. At LMC on average 80% of incoming students assess below college-level math, but placement data disaggregated by ethnicity (as done

7 by Perry, et. al.) is not available. But some insight can be gained from a further analysis of the 2009 LMC cohort enrolled in remedial math. Of those enrolled in remedial math, 74% were enrolled in a course 2 or more levels below college, with 72% of remediating Black students in math courses 3-4 levels below college math, compared to 51% of remediating Latinos and 40% of remediating Whites. High rates of attrition are inevitable given the length of the remedial pipeline. For students enrolled in Elementary Algebra, the pipeline to and through a transferable math course has 5 exit points and we lose significant numbers of students at each point. This is illustrated with the LMC Fall 2009 Elementary Algebra cohort (tracked through Spring 2012). 68% (215 of 315) pass Elementary Algebra Of those who pass, 83% (178 of 215) enroll in Intermediate Algebra* Of those who enroll, 66% (118 of 178) pass Intermediate Algebra** Of those who pass, 73% (86 of 118) enroll in a college-level math course* Of those who enroll, 72% (62 of 86) pass the college-level course** * Enrollment counts are taken at 1 st census. ** Passing is a C or better. As a result, only 20% of the cohort passed a college-level math course in 3 years: (0.68)(0.83)(0.66)(0.73)(0.72) = 0.20, which is 62 of 315 Can we improve the college course completion rate to an acceptable level by improving course success rates and course-to-course persistence rates? Unfortunately, the answer is no. For example, if success and persistence rates were an unprecedented 85%, the completion rate would only be 44%, 0.85^5, which is a huge improvement but an unacceptable general outcome. Our conclusion: If the goal is to increase the college math completion rate for underprepared students, we must shorten the remedial course pipeline AND still prepare students adequately for the rigors of the college math course that is appropriate for their major. B. Student Pathways Most LMC students want a 4-year degree, but they are not STEM majors. Of those completing the remedial math sequence and enrolling in a college math course, on average 80% take Statistics (and no other college math course) to meet the mathematical concepts and quantitative reasoning requirement for IGETC.

8 C. Misalignment of Remedial Content with Statistics Very little algebra is needed to be successful in college-level Statistics. In our analysis of topics traditionally taught in Elementary and Intermediate Algebra about 15% were relevant to building skills used in the study of Statistics. Similarly, in our conversations with faculty teaching courses typically taken by non-stem students in other IGETC areas only a small subset of algebra skills were truly requisite to the work in those courses. II. Our Solution We shortened the pipeline to Statistics for non-stem majors by replacing the traditional remedial sequence with a single intensive pre-stats course. Since articulation with UC and CSU requires Intermediate Algebra or the equivalent as a pre-requisite, we designed backwards from Statistics and created a course (Math 27, a.k.a. Path2Stats) that the LMC Math Department and Curriculum Committee deemed equivalent to Intermediate Algebra in depth, scope and rigor. STEM students still take the traditional algebra path. Math 27 is a project-based course in which students investigate large and complex data sets to answer open-ended questions. In the context of these investigations, we develop a toolbox of data analysis techniques to analyze univariate and bivariate quantitative and categorical data. The level of rigor as defined by the type of analysis and reasoning required in the projects is higher than that found in traditional algebra courses that focus on procedural skills and template problems. The scope of the course is broad, covering topics from arithmetic, algebra, and statistics. The course delves into exploratory data analysis at a depth that exceeds the treatment of this content in college Statistics. III. The Results Significantly higher completion rates of college-level math To date, 58% of the 151 Path2Stats students have completed transferable math, compared to 19% of developmental students overall. This improvement in transferable math completion rates is seen in all math placement groups. In a study of the 119 students in the 1 st two cohorts, Path2Stats students completed transferable math at 2.2 to 4.3 times the rates of students with comparable math placement who took the traditional algebra sequence. See disaggregation by placement level in the next table.

9 Student placement in traditional math sequence LMC Path2Stats: % of Math 27 students passing Statistics (in 1 year) LMC Traditional Path: % of remedial students passing any college math course (in 3 years) National Data Transfer-level 100% (3 of 3) Intermediate Algebra 82% (18 of 22) 37% (121 of 325) 27% Elementary Algebra 78% (25 of 32) 18% (57 of 315) 20% Pre-algebra or Arithmetic 38% (21 of 55) 10% (64 of 619) 10% Unknown Placement 57% (4 of 7) Overall Completion Rate 60% (71 of 119) 19% (242 of 1259) 20% Source: CCCCD Office of Institutional Research Comparable performance in Statistics as measured by course success rates Student performance in Statistics (Math 34) From Path2Stats (Math 27) From Intermediate Algebra (Math 30) No remediation at LMC Pass (A, B, C, or P) 73% (74 of 101) 74% (772 of 1041) 69% (1054 of 1523) Fail (D or F) 13% (13 of 101) 10% (100 of 1041) 11% (162 of 1523) Withdraw (W) 14% (14 of 101) 16% (169 of 1041) 20% (310 of 1523) Source: CCCCD Office of Institutional Research Achievement of learning outcomes as measured by departmental final exam assessment In college Statistics, the first Path2Stats cohort outperformed the Honors section on the departmental final exam. In the most recent assessment of learning outcomes in college Statistics, 100% of Path2Stats students were rated proficient or better on 2 of 3 learning outcomes; 82% were proficient on the 3rd. On items from the Comprehensive Assessment of Outcomes for a first course in Statistics (CAOS), Path2Stats students overall performance was within 3% of the national average. A window into the Path2Stats classroom For a 5-minute window into the Path2Stats (Math 27) classroom, go to This video was created by an LMC student, Jose Reynoso with footage from the 6 th week of an 18-week semester in the pre-stats course. The students featured in the opening segment placed into arithmetic, elementary algebra or intermediate algebra. Can you tell who is who?