Student retention at university The Social Market Foundation 1 Why - - PowerPoint PPT Presentation

Student retention at university

The Social Market Foundation

1

Why worry about student retention?

• Participation rates at university have grown remarkably in past decades
• There has been significant attention on widening the pool of applicants to and

enrolments in university, but less focus on whether those that attend stay on

• Teaching Excellence Framework may raise this up the agenda.
• Those from specific demographic groups may be more likely to drop out. This

is an important social mobility agenda.

2

• The difference in dropout rates between London and the rest of the country is much less

pronounced than two years ago.

• However, the region is still performing poorly.

Regional dropout rates over time

4% 5% 6% 7% 8% 9% 10% 11/12 12/13 13/14 14/15

EAST MIDLANDS EASTERN NORTH EAST NORTH WEST SOUTH EAST SOUTH WEST WEST MIDLANDS YORKSHIRE AND HUMBERSIDE GREATER LONDON TOTAL

Proportion of UK-domiciled students that drop out of higher education by end of first year 3

5.0% 6.7% 5.9% 7.8% 5.1% 4.6% 6.5% 6.2% 7.7% 9.9% 13.3% 11.8% 11.1% 10.1% 10.3% 11.9% 11.3% 13.1%

0% 2% 4% 6% 8% 10% 12% 14%

Young Dropout Mature Dropout

• Mature first-degree entrants have significantly worse retention rates than their young

counterparts in all regions of England. Mature is defined by HESA as aged 21 or over on September 30th on the year of entrance.

• There is imperfect correlation between the retention rates of the two groups regionally.

The factors influencing their dropout rates are likely to be different.

• Our analysis focuses premodinantly on young students.

All cross- sections from 2014/15

Significant differences between younger and older students

Proportion of UK-domiciled students that drop out of higher education by end of first year 4

Principal retention factors

A sense of belonging

• Building relationships with other students and academic staff; feeling that the

institution is suited to your needs Engagement

• Participating in lectures, tutorials and extra-curricular groups

Financial Constraints

• Being able to afford the costs of living, the costs of study material and of

socialising and participating in university life Examples of how such factors may be manifested

• Those from poorer backgrounds, or ethnic minorities, may find it more difficult to

engage with their studies (differences in culture, attitudes).

• Poorer students may also find it difficult to afford the necessaries.
• Those living further from university, and especially not in halls or on campus, may

find it more difficult to engage in student life generally.

5

10 20 30 40 50 60 70 5 10 15 20

NS-SEC 4-7 (%) Dropout Rates (%)

• London’s rate of students

whose parents are classed as NS-SEC 4-7 is around the median for universities. The capital’s figure (35%) is much lower than the West Midlands’ 41%.

Socio-economic background

0% 5% 10% 15% 20% 25% 30% 35% 40% 45%

Socioeconomic background by region (NS-SEC 4-7)

• NS-SEC 4-7 is a measure of whether a

student’s parents are / were in specific

• ccupations (including routine jobs, lower

supervisory roles and small employers) indicative of a lower socioeconomic class. There is high correlation between the proportion of student’s who fall into this group at a university and dropout rates.

Relationship between socio-economic score and drop-out rates 6

Prior attainment of students

• Evidence shows a negative correlation

between prior attainment and dropout rates.

310 320 330 340 350 360 370 380 390 EAST MIDLANDS EASTERN GREATER LONDON NORTH EAST NORTH WEST SOUTH EAST SOUTH WEST WEST MIDLANDS YORKSHIRE AND HUMBERSIDE

Average UCAS points

• London has one of the highest average

UCAS tariffs

• But, data on north east suggests that

prior attainment is unlikely to be attributing to the regional differences

Relationship between UCAS score and drop-out rates Average UCAS score by region 7

• Certain ethnic groups have higher

drop out rates than others. Given the differences in regional ethnicity, this could explain regional differences.

• Black students are more likely to

drop out than other ethnic groups.

The importance of ethnicity

0% 5% 10% 15% 20% 25% 30% 35% 0.0% 5.0% 10.0% 15.0% 20.0% Black Students (%) Dropout (%)

0% 2% 4% 6% 8% 10% 12% 14% 16% Ethnicity (%)

Regional Ethnicity Breakdown

Black Asian

• London has by far the highest proportion
• f Black students. Other regions with a

high proportion also have high dropout rates: Eastern England and the West Midlands come third and fourth bottom for retention.

• HESA provides only two minority

divisions when reporting institutional-level dropout rates: Asian and Black. We know that retention levels among different Asian groups varies significantly.

Relationship between ethnicity and drop-out rates 8

• We find that, overall, students at campus universities have a non-continuation rate of 5.9%,

considerably lower than the average of 6.8% for non-campus students.

• In regions outside of the North and London, the population of campus universities have at least a
• ne percentage point advantage over non-campus universities.
• In London and the North this picture is reversed – drop out rates for campus universities are

higher than for non-campus universities.

• When looking at campus retention rates by university tariff type the pattern remains for low and

high tariff institutions.

• Medium tariff campus universities have higher dropout rates than their non campus counterparts

Does being on a campus make a difference?

0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% Dropout rates Universities MEAN

University non-continuation rates: campus versus non-campus

Non-campus Campus

9

Student satisfaction

Relationship between student satisfaction and drop out rates:

75 80 85 90 95 100 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20%

Proportion 'satisfied' NSS Institutional dropout rate

LONDON ELSEWHERE Linear (LONDON) Linear (ELSEWHERE) 81 82 83 84 85 86 87 88 89

Average proportion satisfied by region

• On average London has the lowest

proportion of satisfied students based on figures from the NSS

• However, the trend line is flatter within

London suggesting less of an influential effect

10

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 0% 5% 10% 15% 20% 25% 30% 35% Proportion living at home Young Dropout

• An important metric of student living is the proportion of students living at home. Our hypothesis

was that students living at home would have higher dropout rates, given the increased difficulty of engaging socially and academically in the student experience. It is possible that this is also be a product of cultural factors and / or socio-economic factors.

• The data looks to support this hypothesis, given the obvious correlation over regions.

Student Living – Living at home

Relationship between living at home and drop out rates 11

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 1 2 3 4 5 6 7 8 9 Commute to university (miles) Young Dropout

• Commuting distance, may capture the effects of those living at home being less engaged, and

tending to live further away.

• However, it may also represent an isolated effect of living far away limiting the amount of time

that may be spent in university, whether studying or otherwise engaging in the experience.

Student Living – commuting distances to study

Relationship between commuting distance to university and drop out rates 12

Institutional cost of living

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 EAST MIDLANDS EASTERN GREATER LONDON NORTH EAST NORTH WEST SOUTH EAST SOUTH WEST WEST MIDLANDS YORKSHIRE AND HUMBERSIDE

Average cheapest accomodation cost by region

• London universities have the

highest average cheapest room costs available to students.

• There is difficulties in

representing the regional value as institutions will have varying accommodation policies.

• At an institutional level there appears to be

no correlation between the cheapest accommodation costs and the dropout rate.

13

Regression analysis Variable Coefficient (*sgf) P-value

% of Black students 0.139* 0.007 % of Asian students

• 0.116*

0.036 % Female

• 0.407*

0.061 % NS-SEC 4-7 0.763* 0.000 % POLAR3 0.299* 0.000 Campus dummy

• 0.005

0.938 UCAS

• 0.001*

0.003 % Satisfied (NSS)

• 1.184*

0.057 University population 5.71e-07 0.877 Distance moved by region 0.002 0.317 London dummy 0.262* 0.008

To understand better whether there is a ‘London effect’ present we conducted a regression analysis that controlled for a range of factors that we believe could influence student retention. The first thing to note when looking at these results is that whilst we have a London effect present, this result does not hold when we do not control for POLAR3. This is due to the low level of POLAR3 scores amongst London institutions. The London effect is unlikely to be robust. However, the results add substantial evidence to the debate around the importance of student characteristics. A significant amount

• f variation in institutional student dropout

rates is attributed to the characteristics of their student population. Some of the variables that are insignificant at an institutional level may be of importance to certain subgroups of students.

* Statistically significant 14

Conclusion:

Variables found to be significant in our regression:

• Ethnicity
• Gender
• Socio-economic background (NS-SEC and POLAR)
• NSS scores

15

Factors not significant but of further interest:

• Campus vs. non campus
• Commuter students
• Those studying within the same region they grew up in
• Financial constraints

The London effect:

• Negative and significant effect on retention when including POLAR3
• Given the participation rate of students attending university in London we

would expect retention rates to be higher

• Based on our significant variables the London retention issue is likely to be

attributed to student demographic and NSS scores We put forward a number of policies for universities, government and local mayors

Questions?

16