2. Success for All Delivery Plan Data Evidence – Final Report June 2005

3. 1. Provision mix

4. The majority (71%) of GFEC learners are adults studying part-time The majority (56%) of SFC learners are under 19 year olds studying full time GFEC and SFC learners in 2003/04 by mode of study and age Source: ILR/SFR05 (14 December 2004)

5. The number of short courses (<24 weeks duration) in GFECs has increased in recent years Source: DfES analysis of ILR data

6. GFEC provision mix based on qualification volume - 2003/04 Based on qualification volumes in 2003/04, the GFEC provision mix is 46% short courses, 27% adult long courses and 27% 16-18 long courses Allocation of a course to the short/long categories is based on duration – anything less than 24 weeks is defined as a short course GFEC - % qualifications by course length and student age (03/04) Source: DfES analysis of ILR

7. GFEC provision mix based on guided learning hours - 2003/04 Based on GLH in 2003/04, the GFEC provision mix is 16% short courses, 36% adult long courses and 48% 16-18 long courses Long vocational courses (16-18 and adult) account for 74% of provision on this measure GFEC - % GLH by course length and student age (03/04) Source: DfES analysis of ILR

8. GFEC provision mix over time (03/4 v 97/8): Guided Learning Hours (GLH) Source: DfES analysis of ILR

9. GFEC learner numbers over time (03/4 v 97/8) – by main aim type Source: DfES analysis of ILR

10. GFEC provision mix over time (03/4 v 97/8): Qualification aims Source: DfES analysis of ILR

11. Sixth form college provision mix in 2002/03 – on the basis of GLH and qualification aims Source: DfES analysis of ILR data

12. 2. Learner characteristics

13. Learner characteristics: age; mode of attendance; and gender LSC funded learners (000) by age, mode of attendance and gender Source: ILR/SFR05. Includes learners in external institutions.

14. Learner characteristics: ethnicity FE College learners (000) on LSC funded FE provision (2003/04) by ethnicity Source: ILR/SFR05

15. A greater proportion of GFEC learners come from relatively disadvantaged social backgrounds Source: DfES analysis of Youth Cohort Survey (Spring 2002) * Mainly independent school and college Per cent of 16 year olds in each institution type who come from the lower 3 socio-economic groups (2002)

16. A significant proportion of FE college learners receive Widening Participation (WP) uplift 37% of learners were eligible for widening participation uplift in 2003/04 (additional funds are paid to the college, not the learner) WP uplift is payable for a variety of reasons Most typically, the learner is resident in a post code deemed to be relatively disadvantaged The uplift is payable if the learning aim is basic skills Source: DfES analysis of ILR data

17. GFECs have a higher proportion of disadvantaged learners than sixth form colleges and school sixth forms We can compare the proportion of learners resident in a Widening Participation (WP) post code across institutions The proportion of GFEC learners resident in a WP postcode is 29%, compared to 25% of the population The sixth form college and school sixth form figures are 25% and 19% respectively Source: DfES analysis of ILR data

18. 3. College performance

19. 3a. Looking beneath the headline success rate

20. The FE college headline success rate has increased significantly since 1997/98 Source: LSC Benchmarking Data Note: Figures exclude external institutions (including EIs, the 02/3 figure is 67% and the 03/4P figure is 71%)

21. The recent improvement in success rates has been most marked for short courses and 16-18 long courses (a high proportion of which are ‘A’ levels) Source: LSC Benchmarking Data. Excludes external institutions

22. Looking at retention and achievement separately helps us to understand better the drivers of change … Source: LSC Benchmarking Data. Excludes external institutions

23. … which vary according to length of course and age: The improvement in the short course success rate (03/04P v 00/01) was driven by higher achievement. The retention rate increased only slightly over the period. The improvement in A level success rates for 16-18 year olds occurs mostly in the first year (00/01 to 01/02) and is almost entirely driven by an increase in retention rate following the introduction of AS levels in 00/01 Long vocational course success rates improve for both 16-18 year olds and adults, driven by increases in achievement rates. The retention rate on vocational courses showed a greater improvement in 03/4 (based on provisional data)

24. FE college retention and achievement rates on long courses by qualification type and age (00/01-02/03) Source: LSC Benchmarking Data

25. We can attribute the increase in headline success rate to the following factors (02/3 on 99/00) Source: DfES analysis of benchmarking data. Includes external institutions There is a widely held view that part of the historic increase in success rates, particularly for short courses, is a result of better reporting by colleges. Prior to the targeting of success rates, colleges had less incentive to report a success on a short course. Our indicative analysis suggests that better reporting could account for around +2% points of the 12% points increase over the period

26. 3b. Success rate performance gap

27. For GFECs, and SFCs, the success rate performance gap narrowed only slightly between 98/9 and 02/3 Variation in success rate (all courses) by provider Source: DfES analysis of Benchmarking Data

28. For long courses only, the success rate performance gap has not narrowed at all at GFECs Variation in success rate by provider – long courses only Source: Department for Education and Skills analysis of Benchmarking Data

29. The success rate performance gap has actually widened for NVQ2 (GFEC only) 10% of GFECs have an NVQ 2 success rate of 32% or below 10% of GFECs have an NVQ 2 success rate of 61% or above Source: Department for Education and Skills analysis of Benchmarking Data (only includes colleges with more than one hundred of the relevant qualification aim

30. Achievement rate performance gap 94/5 to 00/01 (all providers, old definition of achievement) Source: DfES analysis of ILR

31. Achievement rate performance gap 98/9 to 02/03 (all providers, new definition of achievement) Source: Department for Education and Skills analysis of benchmarking data

32. The gap analysis does not track providers over time – individual providers move up and down the performance distribution. When we do track colleges over time, we find that a number of colleges have been able to make significant improvements to success rates and that this is not a result of shifts in provision mix….

33. The 10 th percentile (worst 10% of colleges in terms of success rate) in 98/9 includes 18 colleges. In 02/3, only 5 of these colleges remain in the 10 th percentile. Two had moved up to the 90 th percentile. Source: DfES analysis of Benchmarking Data

34. 3c. Performance by institution type – success rates and value added

35. Comparison of GFEC and SFC success rates Source: LSC Benchmarking data

36. Distribution of ages for 16-18 year-olds taking AS levels in 2003/04 by college type Source: ILR F04 2003/04

37. A2 performance data for 02/3 by institution type (16-18 year olds) Sources: 1 From qualification awarding body data 2 From FE benchmarking data derived from the ILR

38. Value added for 16-18 year olds studying at level 3 Value added is a measure of achievement that takes into account the individual student’s starting point. But it depends on the existence of graded qualifications. Both GCSEs and A levels provide quite finely differentiated grades. Many level 3 qualifications have a degree of grading. The existence of a strong relationship between prior attainment and outcome is helpful. Note that in post-16 qualifications, value added is only currently available for young people studying at level 3 For adult learners and for 16-18 year olds studying below level 3, qualifications are not sufficiently differentiated to allow value added scores, and also there is not a strong relationship between prior attainment and outcome It should also be noted that VA looks only at results in exams entered

39. Raw attainment scores for 16-18 year old students who are entered for level 3 exams, are much higher on average for school and sixth form college students compared to GFEC students. Average point score by candidates achieving GCE/VCE A/AS and Key Skills at Level 3 qualifications (2004 results) Source: SFR38/2004 Note: based on UCAS points system: A=120; B=100; C= 80; D=60; E=40

40. Is this difference due to differences in student in-take? Prior attainment at level 2 is the single major determinant of a 16-18 year-old’s performance in GCE/VCE A/AS level qualifications We need to investigate whether differences in raw attainment scores by institution type are to some extent a result of differences in average prior attainment of student intake. It is possible that some institution types are able to attract higher ability students This is in fact the case. In particular, GFECs attract a higher proportion of students with relatively low prior attainment at level 2. See over

41. GFECs attract more low prior attainment students (cumulative percentage of students below given prior attainment thresholds) Source: LSAD analysis of 2003 data in the Department for Educations and Skills Bulletin on Student Progress between GCSE/GNVQ and GCE/VCE A/AS levels (Issue number 01/04, May 2004)

42. What if we look at students with similar prior attainment in different institution types? Department for Education and Skills data allows us to do this for students taking level 3 exams in 2003 If we group students according to level 2 prior attainment, the ‘average’ student score at GCE/VCE for each prior attainment band is similar for different types of institution For most prior attainment bands, SFCs slightly outperform maintained schools, which in turn slightly outperform GFECs

43. For most prior attainment bands, SFCs outperform maintained schools, which in turn outperform GFECs Note: GCSE points are allocated to grade as follows: A*=8; A=7; B=6; C=5; D=4; E=3; F=2; G=1. A level points are allocated to grade as follows: A=120; B=100; C= 80; D=60; E=40

44. College performance relative to maintained schools

45. What can we say about a ‘typical’ 3-A level student? For most levels of prior attainment, a SFC student scores on average 2 points more than a student in a maintained school per A level entered A GFEC student scores on average 3 points less than a school student per exam entered So for a given level of prior attainment a student doing a 3 ‘A’ level package in a school sixth form will achieve about one third of a grade (i.e. 3 x 2 pts = 6 pts) less than one in a sixth form college and about half a grade (i.e. 3 x 3 pts = 9pts) more than one in a general FE college Note that there is an element of conjecture here – as we are taking the average results per entry and assuming that we can go from this to a total points difference for a 3 ‘A’ level student

46. Does this tell the whole story? This analysis ignores the volume of qualifications taken by students in different institutions. What it shows is that students with the same prior attainment tend to score slightly higher average grades if they attend a sixth form college than if they attend a school or GFEC. It says nothing about the number of exams that students in different institution types are entered for. It should be noted that the differences in value added by institution type could be the result of differences in numerous factors by institution type, for example: learner characteristics (other than prior attainment); funding levels; institutional effectiveness; subject mix (it has been shown that the relationship between average prior attainment at GCSE and ‘A’ level grade varies by subject type) Finally, note that this analysis is based on 02/3 Department for Education and Skills data of individual student average performance. It is not based on actual institution VA scores. Department for Education and Skills/LSC are currently developing a measure of institutional value added which will give a more robust assessment of whether value-added varies systematically according to type of institution

47. 3d. College inspection

48. Inspection grades: FE colleges Inspection reports contain a wealth of performance data. As well as an overall assessment, grades are awarded for ‘management & leadership’ and ‘teaching & learning’ by subject on a scale of 1-5, where: 1= Outstanding; 2 = Good; 3= Satisfactory; 4 = Unsatisfactory; 5 = Very Poor SFCs tend to score highly on inspection assessment, perhaps reflecting their relatively homogenous learner base (high prior attainment 16-18 year olds taking A levels) Most GFEC’s have satisfactory or better management and leadership But pockets of poor provision (graded 4 or 5) at the majority of GFECs – hence a large number of GFECs have undergone partial re-inspection A small number of GFECs have achieved excellence throughout the institution and this has not been at the expense of ‘harder to reach’ learners

49. GFEC performance in inspection: Management & Leadership grades 171 GFECs inspected 01/2 – 03/4

50. GFEC performance in inspections: pockets of poor provision 101 of 171 GFECs inspected have at least one subject area with T&L grade 4 or 5 Almost three quarters of GFECs have been subject to some degree of re-inspection