![Sample our Education journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5932/TOC-Subject-Sample-2021-Education.jpg"})
Abstract
Given the financial implications for studying at a higher education institution, students are faced with many decisions both in terms of the topics they wish to study but also the decision of whether to remain at home or move away. The aim of this article is to compare the experiences of South Asian (Indian, Bangladeshi and Pakistani) female students to White females, to identify similarities and differences between women from the three Asian groups and also compare changes over time. This article is focused on the term-time accommodation status of these students, i.e. the decision to move away or stay residing in the parental/guardian home during their first year of university. Understanding students’ accommodation choices is important as it can limit the other choices they have to make such as which course to choose and students who are limited to their local institutions can only study what is offered by those particular institutions. We examine this issue by using binary logistic regression to analyse administrative enrolment data on HE students in England for academic years 1998/99 and 2005/06 from the Higher Education Statistics Agency (HESA).
Acknowledgements
Thanks are given to the Economic and Social Research Council (ESRC) for providing funding for this project and the Higher Education Statistics Agency (HESA) for supplying the data.
Notes
1. No date mentioned.
2. It should be noted that students enrolled at a London-based institution are not present in the data set as they have asked HESA not to release individual level data.
3. English domiciled students are defined as those whose normal residence is in England.
4. No date mentioned.
5. What is reported as highest qualification on entry is not necessarily what was required for entry on to the course being studied.
6. The authors were required to adhere to HESA’s rounding strategy. Throughout this paper, raw figures ending in 0, 1 or 2 are rounded to 0. All other numbers are rounded to the nearest 5. Total figures are also subject to this rounding methodology; the consequence of which is that the sum of numbers in each row or column will rarely match the total shown precisely.
7. Tables can be found in the Appendix.
8. The method used for constructing the IMD is explained further on in the paper.
9. Each HEI varies on whether or not they wish to include results for General Studies examinations in Tariff score. Information on which HEIs include them is not available in the data sets. However the implications of this will not be great. Some universities consider this A level in confirming a student’s place at the university if they have not met their grade requirements in their other subjects as it can demonstrate that a student has more potential than just within the boundaries of the other subjects they are studying. However, only a small proportion of students study this subject as on the whole universities are unwilling to include the subject in their offers (TSR 2009). Therefore, the variation amongst universities in their willingness to accept the subject has little ramifications for this research as it does not apply to most students.
10. See: http://www.ucas.com/students/ucastariff/tarifftables/ for a full explanation of methods used for calculating points.
11. If a student is in temporary accommodation but has permanent accommodation agreed, they areinstead asked to give information on their planned accommodation.
12. There may be students registered at institutions but who are studying in regions other than that of the administrative centre. No identification of these students is available in the data sets.
13. As we are analysing those students that do leave home, those that remain living in the parental/guardian home are not included in this table.
14. These domains are income, employment, health deprivation and disability, education skills and training, barriers to housing and services, crime and living environment.
15. This information was missing for 3.2% (2,080) of students in 1998, and 3.1% (3010) in 2005. Overall, there were a very small percentage of IMD missing values.
16. The score for each LSOA in England is available from the Communities and Neighbourhoods government website in Excel format. This file was converted into a data file and then merged with each of the data sets so that IMD scores were linked in with LSOAs. This new variable was the only (geographical) socio-economic indicator in the 1998 data set. In 2005, it was used as one of two socio-economic variables as information was also available on SEG on the basis of parental occupation.
17. If this household member is retired or unemployed, the student is asked to state their most recent occupation. Those not in full-time education and/or are aged 21 or over are asked to state their own occupation.
18. There is an important characteristic that needs to be considered when discussing this variable. The quality of the socio-economic data depends on the accuracy with which an applicant records their parental occupation. What is recorded may be vague and loosely linked to the sector and position of their parent(s), so some degree of error is introduced whenoccupation is recorded. This is however clearly not an important feature of the data as overall 0.0% of the sample is in the ‘Not Classified’ category.
19. In regression modelling, a set of baseline characteristics are held constant in order that factors associated with a particular outcome can be identified.
20. Regression results are presented as odds ratios.
21. Not statistically significant.
22. These prices were correct at the time of writing.
23. The results for IMDs 2 and 3 are not statistically significant.
24. The multiplicative interaction effect was calculated by multiplying together the various terms for each of the explanatory variables and related interaction. It is illustrated by the following formula: (Agresti Citation1996, 110).
25. The effect for Indian students can be calculated by multiplying the main effect for Tariff Z-Score (0.51) and the interaction between Indian and Tariff score (1.23). The result is 0.63 indicating a 37% decrease in odds of living at home.
26. The result for the third quintile also indicates a greater likelihood of living in the parental home, but it is not statistically significant.