Social Class, Ethnicity, Gender and Patterns of Educational Achievement : Data
|
Update: Educational Attainment and School Closures due to COVID 19
Learning objectives for this document
|
Social Class Differences in Educational achievement:
Introduction
Sociological studies in advanced industrial countries including the UK have shown that educational achievements (as measured mainly by educational qualifications achieved) are closely related to social class background and that upper and middle class students on average out perform working class students at all levels of the education system. For example, sociologists from the 1950s onwards have regularly pointed to the progressive under-representation of working class students in:
higher streams in primary (i.e. middle schools)
numbers passing the 11+ examination
numbers in higher streams in grammar schools and subsequently in comprehensives
numbers remaining in school after the minimum school leaving age
numbers passing O levels, gaining high grade GCSE passes and passing A levels
numbers enrolled on undergraduate courses
numbers involved in post graduate study.
Despite a wide range of government educational policy initiatives such as the introduction of free secondary schooling for state educated students and apparently fair, objective methods of selection (the 11+ examination) for the different types of school (Grammar, Technical, Secondary Modern) in the new Tripartite system in the 1944 Education Act, the expansion of state expenditure on education, the subsequent recognition of the limitations of the Tripartite Secondary System and its replacement almost everywhere by Comprehensive Secondary Education, the raising of the school leaving age to 15 and subsequently 16, the development of Education Priority Areas , the schools initiatives too numerous to mention of the Thatcher, Major, Blair and Brown , Coalition and Conservative administrations and the expansion of opportunities for further and higher education, it can still easily be demonstrated that working class students are less likely to be educationally successful than are their middle class peers counterparts.
Social Class Differences in Educational achievement:
- Social Class and GCSE Level Attainment
- Long term trends 1989-2006
- GCSE Attainment and Free School Meal Eligibility 2008/9- 2016/17
- The Coalition’s Record on Schools: Policy, Spending and Outcomes 2010-2015 by Ruth Lupton and Stephanie Thomson
Source 1: Youth Cohort Studies of the educational achievements of 16 year olds 1989-2006 published on the Department for Education and Skills DfES [subsequently DCFS and now DfE] website.
[The 1989-1999 data refer to GCSE Examination results in for England and Wales whereas the 2001-2006 data refer to England only]
An individual’s social class position is often approximated by his/her occupation but the occupational classification schema used by the UK Government were altered in 2000 so that the statistics for 1989-1998 are not comparable with the statistics for 1999- 2006. Nevertheless the data do show that children whose parents are in professional and managerial occupations are more likely than children whose parents are in manual occupations to gain 5 or more GCSE A*-C grade passes.
For a diagrammatic presentation of these data for years 1999-2006 - Click Here to Download , once you have reached the diagram [chart two] you may also click Sheet 1 or scroll over the diagram itself if you wish to revisit the actual statistics.
Activity
|
Source 2: GCSE Attainment and Free School Meal Eligibility
I have updated this table to include 20198-19 data. It already ovelapped the margin and now does so even more. Therefiore I suggest deleted 2010/11 and 2013/14. In future more intermediate colunms can be deleted as we add future years so that it will give us quite a nice long term trend The net effect should be that there is no overlap
The YCS series on social class differences in educational achievement has been discontinued but the DfE continue to provide data on differences in educational achievement as between students eligible and ineligible for Free School Meals . Obviously Free School Meal Eligibility is a less than prefect measure of social class membership but these data do give some indication of the correlation between adverse economic circumstances and educational achievement at GCSE Level. Also since 2015/16 educational achievement has been measured in terms of the Progress 8 criterion and or attainment of 5 Ebacc subjects rather than in terms of attainment of 5 or more GCSE pass grades A*-C
Table : Gender, Free School Meal Eligibility and Percentages of Pupils gaining 5 or more GCSE A*-C Grades including English and Mathematics 2008/9 – 2014/15 6/17 and percentages achieving the EBacc [with English and Maths Grades 9-4 in 2016/7 and other subjects graded 9-4 in 2017/18]
Sources : DFE SFR Various Years: GCSE Attainment and Pupil Characteristics:
| Activity
1. Using information in the above table on Gender, Free School Meal Eligibility and Educational Attainment answer the following questions.
2. Notice change of date to 2018/19
Which factor seems to be the more significant influence on educational achievement : Gender or Free School Meal Eligibility? 3. Which factor seems to be the more significant influence on educational achievement: gender or social class? 4. However it is important to consider interrelationships between gender, free school meal eligibility and ethnicity. It can then be shown that using EBacc attainment as a measure of achievement in 2016/17 the overall size of the free school meal eligibility/all other students gap is heavily influenced by the size of the gap among white British students who are a large majority of the total student cohort. For White British students the FSM eligibility/all other students is far greater than the Gender gap but the reverse is true in the case of Asian students, Pakistani students, Bangladeshi students, [but not Indian students], Black African students, Black Caribbean students and Any Other Black students. For Chinese students the Gender gap was 12% but Chinese FSM pupils actually outperformed all other Chinese students. |
Source 3
The Coalition’s Record on Schools: Policy, Spending and Outcomes 2010-2015 by Ruth Lupton and Stephanie Thomson . This is a first class paper providing detailed, insightful information on Coalition schools policies. Scroll down to page 45-47 for detailed data and diagrams on educational attainment and free school meal eligibility. There are one or two technicalities here that you may need to discuss with your teachers!
2.Social Class and GCE Advanced Level Attainment
Source 4 and 4b: Attainment at Level 2 [ =attainment of 5 or more GCSEs at Grades A*-C or equivalent qualifications ] and Level 3 [attainment of 2 or more A levels or equivalent qualifications]
Click here for a DFE document entitled Level 2 and Level 3 attainment by young people in England. You may then scroll down to page 8 of this document for the Section on Free School Meal Eligibility which shows attainment by FSM and the attainment gap between the FSM group and their peers at Levels 2 and 3 between 2009 and 2013. It will be seen that both of these attainment gaps are substantial and have changed little in recent years. You may summarise these results here.
Click here for the 2018 version of this document.
For data on attainment at Level 3 by Free School Eligibility and Ethnicity in 2018. It will be seen that a substantial attainment gap exists between pupils eligible and ineligible for free school meals and that this gap varies substantially as between different ethnic groups
Click here BBC item on technicalities of Attainment 8 and progress 8 and here for a very technical article
Click here for calculation of Attainment 8 and here for calculation of Progress 8
- Social Class and Access to Higher Education
Source 5: Focus on Social Inequality [Edited by P. Babb J. Matin and P.Haezewindt ONS 2004] provides data on relationships between social class [measured by parental occupation] and participation in Higher Education .
The above data indicate that, for example, in 1960 27% of the children of parents in non-manual occupations participated in Higher Education compared with 4% of the children of parents in manual occupations. Also in 1960 only 5% of all 18-21 year olds participated in Higher Education.
The final columns of the table are s taken from a 2009 Department of Business, Innovation and Skills Paper distinguishing between Higher Education Participation Rates of male and female students from NS SEC Classes1, 2 and 3 and NS SEC Classes 4,5,6 and7. Clearly the 2007/8 data are based upon different social class schema and are therefore not fully comparable with the previous data.
For a diagrammatic presentation of these data. [However I have not included the more recent DBIS statistics in this diagram since they are not fully comparable with the earlier data]
More Recent Data
Click here for DFE publication December 2019: Widening Participation in Higher Education Also once you reach the SFR you may click on National Tables and then on Table 1 to find the source of the tables which I have presented below. Download full Document in PDF - Click Here
Eligibility and ineligibility for free school meals are obviously inadequate measures of social class inequality but pages 1, 4 and 5 of the report data do nevertheless indicate that social class inequality , although imperfectly measured, does have a significant impact on access to Higher Education.
- Clearly The Gap in access to Higher Education between pupils eligible and ineligible for free school meals remains substantial
- However the are also significant regional differences in this access gap. Note especially the performances of London pupils eligible and ineligible for free school meals
- Access to more selective Higher Education Institutions is more likely for independent school students than for state school students
- There are also significant social class differences in post-graduation employment in high skill occupations
Click here and scroll to pages 4-5 for similar data provided by UCAS
Click here for a Channel 4 Fact Check which provides very useful information on Social Class and Access to HE
Click here for useful article 2015
ActivityWhat has happened to total participation in Higher Education since 1960?
|
Elitist Britain : the educational background of the Britain’s leading people [Sutton Trust Report June 2019]
The Educational Backgrounds of the UK Professional Elite [Sutton Trust Report March 2016]
Approximately 7% of UK pupils are educated in private schools and privately educated pupils[ especially those from prestigious Public Schools such as Eton and Harrow are disproportionately more like to secure employment in elite occupations. The above report provides very detailed information on this issue . It also has a clear concise Executive Summary.
Office of National Statistics Data 2011 indicate that individuals ’ educational qualifications have a major impact on their earnings potential.
Click here for the relevant ONS publication
The above data illustrate that there are very significant social class differences in educational achievement and also that higher educational achievements are associated with higher earnings.
- The Youth Cohort Study data indicate strong relationships between parental social class and educational achievement at GCSE level.
- Pupils eligible for free school meals are on average less successful at GCSE level than pupils not eligible for free school meals.
- The same applies to Level 3 Qualifications [Source 4]
- There are strong relationships between social class membership and participation in Higher Education .
- There are strong relationships between educational qualifications and earnings for both men and women
Taken in combination these findings mean that many working class children are themselves unlikely to earn high incomes in adult life because of their limited educational qualifications. Class advantage is to some extent transmitted from generation to generation although many working class children are successful in education and socially mobile in their employment careers.
Ethnic Differences in Educational Achievement
The single most useful source of data on Ethnic Differences in Educational Achievement is without doubt Ethnicity : Facts and Figures: Education, Skills and Training.This document provides information on ethnic patterns of educational achievement at all levels of the English education system for 2017-18
The sources which appear below provide information on Longer term trends and a little information on Examination results in 2018/19. Also there is a little information on Average Point Scores at GCE Advanced Level in 2017/18 and 2018/19 whereas Ethnicity: Facts and Figures provides information on Ethnicity and High Achievement at GCE Advanced Level in 2017/18. With regard to Access to Higher Education there is information on Ethnicity, Free School Meal Eligibility and Access to Higher Education.
Ethnicity and GCSE Level Attainment: Longer term trends
Ethnicity and GCSE Level Attainment
The following information is extracted from the Youth Cohort Study of 16 Year olds published in Feb 2005 and amended in June 2006. It illustrates trends in educational achievement at GCSE level [as measured by attainment of 5 or more GCSE grades A*-C in Year 11] between 1989 and 2004 according to the ethnicity of the students estimated on the basis of samples which range between 24922 and 13,698.
Subsequently much greater emphasis came to be placed on the percentages of students achieving 5 or more GCSE subjects including English and Mathematics and in the two final columns I have now included 2014/15 data for students achieving 5 or more GCSE grades A*-C [Column A] and for students achieving 5 or more GCSE Grades A*-C including English and Mathematics [Column B].
Subsequently achievement levels have been measures in terms of the Progress 8 statistic and the attainment of 5 EBacc subjects at grades A*-C and most recently grades 9-4 or 9-5. Achievement levels based upon these most recent criteria are shown below.
Attainment of 5 or more GCSE grades in Year 11 by ethnicity 1989-2004 and 2014/15
The above data indicate that the attainments of all ethnic groups listed have improved very substantially between 1989 and 2004 and between 2004 and 2014/15. The Following points should be noted.
- Data for Chinese students were not available in the YCS source but it is clear that Chinese students are the most successful of all ethnic groups at GCSE Level.
- Indian students narrowly outperformed White students in 1992 but by 2014/15 the achievement gap between Indian students and White students increased significantly.
- The relative performance of Bangladeshi students has improved significantly and by 2014/15 they were out performing White students.
- By 2014/15 White students continued continued to outperform Pakistani students but the achievement gap was narrowing.
- By 2014/15 White students continued to outperform Black students but the achievement gap was narrowing.
- By 2014/15 White students continued to outperform Black Caribbean students but the achievement gap was narrowing.
In the above table important subdivisions within the White Category and students of mixed ethnicity have not been included and the data on the various ethnic groups have not been subdivided according to gender and free school meal eligibility. By 2016/17 a new system of evaluation was in place and in the following source I include full data on ethnicity ,gender and free school meal eligibility for the 2017/8 GCSE results in England. Notice that , very importantly, data on Traveller and Gypsy Roma pupils are included in these more recent data
Exercise on Ethnicity and Educational Achievement at GCSE Level in the 2018 Examinations [Uncertainty here] Will this new table photo in black and white?MAYBE leave for Now?
Let us investigate ethnic patterns of achievement using performance in EBacc subjects in 2017 as indicated in Table 2a of the 2018 publication. With regard to English and Mathematics I refer to attainment levels between 9-4 for all pupils. I have copied and pasted the relevant table here but students may also use the [replace above with followng] ] gabove link to refer to separate data for males and females and to data on Attainment 8 and Progress 8 if they so wish.
These data are updated yearly and you may find the full data for 2018/19 by clicking the follow link.
Once you reach the relevant page Click on Key Stage 4 Performance 2019; scroll down a little to National Characteristics Table and open Table CH1 .
You may then discuss this data with your teachers
https://www.gov.uk/government/statistics/key-stage-4-performance-2019-revised
Source: Key stage 4 attainment data
1. Includes entries and achievements by these pupils in previous academic years.
2. State-funded schools include academies, free schools, city technology colleges, further education colleges with provision for 14- to 16-year-olds and state-funded special schools. They exclude independent schools, independent special schools, non-maintained special schools, hospital schools and alternative provision (including pupil referral units, AP free schools and AP academies as well as state-funded AP placements in other institutions).
3. Pupils at the end of key stage 4 who are included in the measure.
4. Attainment 8 and Progress 8 are part of the new secondary accountability system that was implemented for all schools from 2016. Users should be cautious when comparing Attainment 8 scores between 2017 and 2016. In 2017, Attainment 8 scores were calculated using slightly different point score scales in comparison to 2016, in order to minimise change following the introduction of 9-1 reformed GCSEs. This means that Attainment 8 scores are likely to look different in 2017, as a result of changes to the methodology. More information on the calculation of these measures is available in the Progress 8 guidance:
https://www.gov.uk/government/publications/progress-8-school-performance-measure
5. A Progress 8 score of 1.0 means pupils in the group make on average approximately a grade more progress than the national average; a score of -0.5 means they make on average approximately half a grade less progress than average. Progress 8 scores should be interpreted alongside the associated confidence intervals. If the lower bound of the confidence interval is greater than zero, it can be interpreted as meaning that the group achieves greater than average progress compared to pupils in mainstream schools nationally and that this is statistically significant. If the upper bound is negative, this means that the group achieves lower than average progress compared to pupils in mainstream schools nationally and that this is statistically significant.
6. As a percentage of all pupils at the end of key stage 4. In 2014/15 and earlier, where the English language and English literature option was chosen in English, exams in both had to be taken and a C grade or above achieved in English language. From 2015/16, to meet the English requirement of the A*-C in English and maths attainment measure, a C in either English language or English literature counted and there was no requirement to take both. From 2017, following the introduction of the reformed 9 to 1 GCSEs in English, a grade 5 or above in either English language or English literature counts and there remains no requirement to take both. The 9-4 measure shows pupils who achieved a grade 4 or above in either English language or English literature and maths and is shown alongside the headline measure for transparency and comparability.
To Download the above Chart in excel format - Click Here
Using the above data answer the following questions
- Rank the broad ethnic categories in terms of the achievement of all pupils [Final Column]
- Which is the highest performing and the lowest performing broad ethnic group?
- Describe the ethnic differences in attainment within each ethnic category
- Notice especially the attainment levels of students in the Traveller of Irish Heritage and Gypsy Roma categories.
- Which is the highest performing and and lowest performing “narrow ethnic category”?
- Use your calculator to calculate the proportions of pupils eligible for free school meals in each ethnic category
- For all pupils what is the percentage gap in attainment between pupils eligible and ineligible for free school meals?
- In which broad ethic category is the gap in attainment between pupils eligible and ineligible for free school meals greatest?
- What is significant about the gap in attainment between Chinese pupils eligible and ineligible for free school meals?
- Can you give any reasons why the attainment gap between pupils eligible and ineligible for free school meals varies as between different ethnic groups.?
- Discuss these statistics with your fellow students. Any other insights?
- You will also find that females out perform males in every ethnic category and that it is white boys eligible for free school meals who perform particularly badly.
- The overall attainment gap between pupils eligible for free school meals and pupils ineligible for free school meals is greater than the overall attainment gap between female and male students. Numerically this occurs because the FSM eligibility gap is very high [18.2%] for white British students who account for a very large proportion of the total student population. However in several ethnic categories [Pakistani, Bangladeshi, Black Caribbean , Chinese] the gender attainment gap is greater than the FSM eligibility attainment gap. [You would need to look at other Characteristics National tables to confirm this latter point.]
- Write two paragraphs summarising the relationships between ethnicity, gender , free school meal eligibility and attainment at GCSE level
Sociologists have suggested three main types of explanation for the above relationships between ethnicity, free school meal eligibility and educational achievement.
- It is suggested that eligibility for free school meals is an imperfect measure of the relative economic deprivation suffered by pupils of different ethnic groups and that more accurate measures of economic deprivation would show that economic deprivation is a more important explanatory factor than the free school meals data suggest.[ As mentioned, several methods may be used to measure levels of economic deprivation but the comparison of these methods is technically complex and I shall not consider this point any further here. ]
- It is suggested that the differential educational achievements of different ethnic groups may be explained partly by a range of cultural factors operative for some ethnic groups so that some ethnic groups are more successful than others in overcoming the effects of material deprivation. It is widely believed Indian and Chinese families are especially likely to value education highly and that these cultural attitudes enable poor Indian and Chinese students to offset more effectively the effects of poverty.
- Note that on the basis of the above data the NFSM-FSM discrepancy is smaller in all ethnic minority groups than it is among white students.
- Others have is suggested that the cultural differences between ethnic groups have been much overstated and that instead a range of processes internal to the schools themselves may operate to the relative disadvantage of some ethnic groups rather than others. Thus for example poor Afro-Caribbean students and also poor white students may be more likely than poor Chinese and Indian students to experience negative labelling in schools.
The gaps in attainment for white pupils eligible and ineligible for free school meals are greater than for any other ethnic group and this has generated considerable concern as to the prospects of white working class students although free school meal eligibility is a less than perfect measure of social class membership. This topic will be considered in more detail presently.
Ethnicity and GCE Advanced Level/Level Three Attainment
Click here and scroll to page 11 for data illustrating ethnic differences in GCE Advanced Level achievement for broad ethnic groups
Click here for data indicating that Chinese and Indian students are especially likely to achieve very good GCE Advanced Level Results
Ethnicity and Access to Higher Education
Click here for Widening Participation in Higher Education 2019 and then on Main Tables: and then on Table 8 . Data are available on Access To HE and Access to High Tariff Universities. Notice these data at the above link refer to Access To HE from State Schools. It is important to note that large proportions of students gain access to High Tariff Universities via private schools which may well mean that the overall percentages of white students at these universities are greater that the data based only on [state funded and special schools] suggest.
Also if you use the above click here link there are particularly nice graphics on pages 12 and 13 which illustrate the points shown in the following table.
Also in the Ethnicity and Access to Higher Education section .change the last 2 sentences to the following
Relationships between ethnicity, gender, free school meal eligibility and access to Higher Education are illustrated in the following table. You may discuss these data with your teachers
For Similar Data on Ethnicity and Access to High Status Universities Click here for Entry rates to Higher Education from Ethnicity Facts and Figures: Education skills and Training and then on Higher Education. I have extracted the most relevant table below It is noteworthy that although a smaller percentage of White students than other other Ethnic groups enter HE , of those entering HE a larger % of white students than of other ethnic groups enter High Status universities. However given that other members of other ethnic groups are far more likely than whites to enter HE it is likely that as a percentage of 16-18 year old students at the end of their 16-18 studies white students are less likely than all other ethnic minority students apart from Black students to enter High Tariff Universities. Click here and scroll to pages 8-9 for similar data provided by UCAS
This table must be interpreted with care
- The data do indicate that a larger percentage of White entrants to University than of entrants to university from other ethnic groups enter high status universities.
- However from the previous table we also know among students who have completed course for 16 -18 year olds [mostly A Levels] white students are the least likely ethnic group to enter Higher Education. This means that taken as a percentage of these 18 year olds White students are not the most likely to enter high status universities. This is shown in the previous table taken from the DFE site
- There is no separate category for Chinese students who obviously out- perform all other ethnic groups Data for Chinese students are provided in the previous DFE table. We see , for example that Chines students eligible for frre school meals are more likely than all White Students to enter High Tariff Univeristieswh
- Since White students are by far the largest ethnic category they obviously make up the majority of students at all universities
Degree outcomes
Again Click here for Entry rates to Higher Education from Ethnicity Facts and Figures: Education skills and Training and then on Higher Education.
I have extracted this table. White students are shown to be most likely to gain First Class degrees
Again the data in this table must be interpreted with care
Click here and here and here for further information
To Download this Chart- Click Here
Question: How would you summarise relationships between ethnicity and access to HE, access to High Status He, and likelihood of attaining a First Class degree?
Gender and Educational Achievement
More detailed information on Gender and Educational Achievement may be found in my documents on Gender and Educational Achievement and Gender and Subject Choice.
However the main points in relation to gender and educational achievements in England and Wales are as listed below
- Gender and GCSE Level Attainment
- In terms of attainment of 5 or more GCSE A*-C grades in recent years 8-10 % more females than males have reached this standard.
- From 2016 onwards new criteria have been adopted to measure pupil achievement at GCSE level [EBacc, Attainment 8 and Progress 8, Grades 9-1.] Gender differences in attainment remain with these new measurement criteria.
- Females outperform males at GCSE level in every major ethnic group.
- For all pupils at GCSE level the attainment gap between pupils eligible for free school meals and all other pupils is greater than the attainment gap between boys and girls.
- However this total figure arises because of the large FSM/Other attainment gaps among British and White pupils which make up a large proportion of the total. In several ethnic minority groups the gender gap is greater than the FSM/Other Gap.
- There are some significant gender differences in subject choice at GCSE level but these occur mainly in minority subjects
- Gender and GCE Advanced Level Attainment
- Until the late 1980s males out-performed females in GCE Advanced Level Examinations but this then began to change.
- Between around 1990 and 2010 females were more likely than males to gain 3 A Level passes and also more likely to gain high grade passes although the overall differences in pass rates were smaller than at GCSE Level
- In 2017and 2018 and 2019 the overall A*-E Pass rates for females [98.3% and 98.1% and 98.0% ] are s narrowly higher than for males[97.3%and 97.1% and 97% ].
- .In 2012 for the first time since the A* grade was introduced the percentage of males gaining A* pass grades was greater than the percentage of females. This has continued from 2012 to 2019.
- In 2017 for the first time the percentage of males gaining [A*+A combined] pass grades was greater than the percentage of females gaining [A*+A combined ] pass grades. This continued in 2018 but was reversed in 2019 when 25.5% of females and 25.4% of males gained [A*+A grades.]
- You may find more DFE data here on gender differences in achievement in the 2017/18 GCE Advanced Level Examinations. Scroll down to page10 . Female students have a slightly higher points score per entry ; they also had a higher points score per entry in the 3 best A level results per pupil; however a larger percentage of males achieved 3A*-A results or better; males were more likely to achieve grades AAB or higher; and males were more likely to achieve AAB or higher when at least two A levels were “facilitating subjects.” However from 2019 onwards the DfE has ceased recording data on facilitating subjects
- Nevertheless notice that the number of girls taking Advanced Level subjects is greater than the number of boys so that despite boys’ higher percentages of top grade passes the gender difference in the number of top grade passes is small.
- Some analysts predicted that female attainment levels might fall relative to male attainment levels as the coursework components of Advanced Level subjects were discontinued but others have never believed that coursework requirements particularly favoured female students.
- The gender differences in subject choice are greater at Advanced Level than at GCSE Level. See Gender and Subject Choice for recent data
- There are significant gender differences in choice of vocational courses
Gender and Access to Higher Education
Recent Data from the Higher Education Statistics Agency 2018/19
Click here for HE Student Enrolments and Personal Characteristics 2012/13- 2018/19
Females continue to be more likely than males to enrol for Higher Education Courses. Some students self- identify as “Other” rather than male or female
The gender differences in subject choice which occur at GCE Advanced Level continue in Higher Education
Gender Differences in Degree Results 2011/12- 2016/19
These statistics may be summarised as follows. Females are now marginally more likely than males to gain First Class Honours Degrees and quite significantly more likely than males to gain Upper Second Class Degrees
Useful Links
Click here for a report from HEP1 on the underachievement of young men in higher education and here for Guardian coverage of this report May 2016
Click here for Mind The Gap: Gender Differences in Higher Education [Rachel Hewitt; 2020]
Click here for Women and STEM