DfE / NCER CIC Data Sharing Agreement - Strict Limitations for the Use of NPD Data in Nexus and Perspective Lite

This is an NCER member authority briefing on the strict limitations regarding the use of DfE National Pupil Database (NPD) data in Nexus and Perspective Lite, including early DfE Children Looked After (CLA) and Children in Need (CIN) data.

All member Local Authorities (LAs) who access DfE NPD data through Nexus and Perspective Lite must read and comply with all restrictions and data suppression as outlined in this document. Failure to do this may result in the DfE withdrawing access to the DfE NPD data for all NCER Members Authorities.

The restrictions and data suppression as outlined in this document come from the DfE / NCER CIC Data Sharing Agreement (DSA) and are also included in the Nexus and Perspective Lite Terms and Conditions of Use.

Please note: This document refers to DfE NPD data in Nexus and Perspective Lite and not Local Authority and school data loaded directly into Nexus by local authorities.

Please also see with regard to re-use of data in external analysis: https://ncer.zendesk.com/hc/en-gb/articles/8473325325981 

Content

1. Permitted use of DfE NPD data extracts and matched datasets in Nexus and Perspective Lite
2. Statistical disclosure control
3. Early Access to DfE NPD data
   • DfE CLA and CIN NPD Data
   • DfE Unamended NPD Data
4. Further DfE CLA and CIN NPD data rules
5. Further publication rules
6. Local Authority compliance
7. KS4 and /or KS5 2019/20 DfE NPD data

Appendix A –DfE / NCER CIC Service Level Agreement, table from pages 254-30 - Important technical information on restrictions and data suppression

1. Permitted use of DfE NPD data extracts and matched datasets in Nexus and Perspective Lite

NCER has permission to use the following DfE National Pupil Data (NPD) in Nexus (for LAs) and Perspective Lite (for schools): - Key Stage 1, Key Stage 2, Key Stage 4, Key Stage 5, Children in Need (CIN) and Children Looked After (CLA), Absence and Exclusions data.

Using the DfE NPD data enables NCER to provide comprehensive analysis to local authorities through Nexus, and schools via Perspective Lite, to enable the evaluation of strengths and opportunities for improvement. Analysis provided by NCER extends from pupil level reports to local authority, regional and national comparators. Key pupil group analysis is also provided using filters and aggregations within the analysis system.

The public benefit of this work is that it provides the identification of variations in outcomes and where appropriate enables interventions to be put in place to support all pupils, groups of pupils and individuals to achieve their full learning potential. Local authorities (using Nexus) and Schools (using Perspective Lite) can examine where schools, pupil groups or pupils need additional support and triangulate this information with their own professional knowledge and judgement to ensure effective planning for continued improvement.

2. Statistical Disclosure Control

Standard disclosure rules apply in Nexus and Perspective Lite to all DfE NPD data (including DfE CLA and CIN NPD data) that includes pupils from outside a LA’s or virtual school's domain (e.g., regional or national aggregations).

Please refer to Appendix A for technical information on disclosure rules that are required by Nexus, Perspective Lite and also LA use of any exported DfE NPD data from Nexus. These statistical disclosure controls are also included in the Terms and Conditions for Nexus and Perspective Lite.

NCER can confirm that all DfE standard disclosure rules as specified for NPD data in the DfE / NCER CIC DSA are being applied throughout Nexus and Perspective Lite.

3. Early Access to DfE NPD data

DfE CLA and CIN NPD Data - The DfE have imposed strict usage limitations for DfE CLA and CIN NPD data in Nexus. These limitations are imposed on all local authorities due to the early visibility of aggregated national data.

Limitations in place regarding use of early DfE CLA and CIN NPD data will remain in place each year until the DfE CLA and CIN outcomes statistical publications, for that particular year, have been published which is usually in March.

In addition, data or findings from data yet to be released by DfE (including national and regional comparators) such as additional tables in the Outcomes for Children in Need (including children looked after by local authorities in England) statistical publication should not be shared any more widely than the LA without prior permission from DfE.

DfE Unamended NPD Data - As unamended data has not been checked by schools, it should only be used by those with a legitimate need to use it for managerial, operational, or other appropriate decision-making purposes at an early stage.

LA data can be shared with that LA ONLY. Any sharing of data or publishing beyond that would require NCER to seek approval from the DfE. School level CLA and CIN NPD data for a specific LA is for that LAs internal use only and should not be published in any form.

In addition, unamended NPD data (including CLA and CIN NPD data) should NOT be used for

• Issuing public reports / statements based on the information before the official statistics are published
• Supporting direct action on schools or dictating school funding
• Publication or public pronouncement of any analyses at aggregate or school level without permission from the DfE

No school-level figures must be published based on unamended data without prior DfE permission.

4. Further DfE CLA and CIN NPD data rules

Pupil level DfE CLA and CIN NPD data and the matched results are only visible to the LA that has corporate parenting responsibility for the pupil. Regional and National aggregations are calculated in Nexus for comparison purposes only with DfE suppression rules applied to all early calculated fields (see section 2: Standard Disclosure Control for more information). School Level data for CLA and CIN data should not be produced.

Summary reports in Nexus provide outcomes for an individual Local Authorities’ Virtual School cohort, and comparison data, with DfE suppression applied.

This data SHOULD NOT be published in anyway. The purpose of this early analysis is to support effective interventions at the earliest opportunity for some of the most vulnerable pupils nationally.

5. Further publication rules:

LAs should not release anything from the NPD dataset publicly.

NCER should contact and gain agreement from the DfE before releasing anything from the NPD dataset publicly, for example in conference presentations. Any data to be publicly released will be shared with the DfE at least one week in advance of planned release in order to ensure there is time for DfE to comment.

In relation to Standard Disclosure Control of outputs, the term “published” relates to any DfE NPD data that is shared with an audience beyond the LA with corporate parenting responsibility for the pupil. Any DfE NPD data needs to be suppressed by a LA, according to the DfE statistical disclosure controls (as specified in Appendix A), if exported from Nexus and added to reports or presentations for employees that are made available to an external audience beyond the LA, including but not limited to publication of scrutiny meeting papers and minutes*, regional meetings and national presentations where a LA is using DfE NPD data.

*Papers for scrutiny meetings have been included here as they are public documents accessible publicly via external websites.

6. Local Authority compliance

NCER agree that all local authorities who can access DfE NPD data through NCER services offered must read and comply with all restrictions as outlined in the DfE /NCER CIC Data Sharing Agreement as summarised in this document. The restrictions as outlined in the DfE / NCER CIC Data Sharing agreement are also listed in the Nexus and Perspective Lite Terms and Conditions. LA Nexus Users need to tick the box to agree to the Nexus Terms and Conditions on log in to Nexus.

Local Authorities have the ability to export a flat file of their pupils educated within their LA, School. A Virtual School flat file can be exported “only where corporate parenting responsibility exists” or there is a safeguarding issue. Under no other circumstances should a LA have access to another LA’s data.

Failure to do this may result in the DfE withdrawing access to the DfE data and in future years.

7. KS4 and / or KS5 2019/20 and 2020/21 DfE NPD data

Under the NCER / DfE Service Level Agreement the provision of KS4 and KS5 2019/20 and 2020/21 DfE NPD data has been permitted. No directly identifiable establishment IDs will be provided within the KS4 and/or KS5 2019/20 and 2020/21 data extract.

As the DfE have made a commitment to not hold schools and colleges to account on the basis of exams and assessment data from summer 2020 and 2021, NCER and member local authorities agree that they will not attempt to use the KS4 and/or KS5 2019/20 and 2020/21 DfE NPD data for accountability purposes, and furthermore NCER and local authorities will not attempt to re-identify, extract or publish any establishment level data from the data they hold.

NCER Board – March 2023

Appendix A - DfE / NCER CIC Service Level Agreement, table from pages 254-30 - Important technical information on restrictions and data suppression

Please note: The information as shown in Appendix A also forms part of the Nexus and Perspective Lite Terms and Conditions for which NCER users need to tick the box agreeing to the Terms and conditions before log in and use of Nexus.

Description of use of the DfE Data Extracts and the Matched Datasets (as applicable) permitted under this Agreement to be used in the details published on GOV.UK	The National Consortium of Examination Results (NCER) wishes to use the following National Pupil Data: - Key Stage 1, Key Stage 2, Key Stage 4, Key Stage 5, Children in Need (CiN) and Children Looked After (CLA), Absence and Exclusions data. With this data they will provide a comprehensive analysis system to Local Authorities and schools to enable them to evaluate their strengths and opportunities for improvement. Analysis provided by NCER will range from pupil level reports through to Key Stages on a Page. Key pupil group analysis will also be provided using filters and aggregations within the analysis system. The public benefit of this work will be identification of variations in outcomes and where appropriate enable interventions to be put in place to support all pupils, groups of pupils and individuals to achieve their full learning potential.   Local Authorities (using Nexus data) and Schools (using Perspective Lite data) can examine where schools, pupil groups or pupils need additional support and triangulate this information with their own professional knowledge and judgement to ensure effective planning for continued improvement. All 152 Local Authorities are members of NCER and use the NCER online analysis system Nexus to support their education and children’s services provision and to carry out their statutory functions to provide a good education to all pupils.
Permitted Use of the DfE Data Extracts	DfE Data Extracts may be used by the Requester for research in respect of the following project and to produce the Permitted Intended Outputs for the DfE Data Extracts: To provide a comprehensive analysis system to Local Authorities and schools to enable them to evaluate their strengths and opportunities for improvement. Analysis provided by NCER will range from pupil level reports through to Key Stages on a Page. Key pupil group analysis will also be provided using filters and aggregations within the analysis system.
DfE Lawful Basis under the Data Protection Legislation	Public Task
Statistical Disclosure Control	Standard disclosure rules will apply to all data that includes pupils from outside a local authority's or virtual school's domain (e.g., regional or national aggregations): •       Any published NPD figure(s) at a local level must be based on no fewer than 3 cases or else suppressed (e.g. replaced with an X, with a note “* in this table means the figures have been suppressed due to small numbers”. Zero is allowed). •       Any published NPD figure(s) at a school level from 2011 onwards must be based on no fewer than 6 cases or else suppressed (e.g. replaced with an X, with a note “* in this table means a figure less than 6”. Zero is allowed). •       Any published figures at school level for years prior to 2011 must be based on no fewer than 11 cases or else suppressed (e.g. replaced with an X, with a note “* in this table means a figure less than 11”. Zero is allowed). •       Even if only one cell requires cell suppression, you must suppress at least one other component cell (the next smallest) to avoid calculation of suppressed values from the totals. •       No suppression or rounding is required on any pupil, school, virtual school or local authority-based figures within a local authority's/virtual school's domain. Children Looked After (CLA) data: Local level (refers to local authority level) must be based on no lower than 6 cases or else suppressed (e.g. replaced with an c, with a note “c in this table means the figures have been suppressed due to small numbers”). National and regional figures should be rounded to the nearest 10 and Local Authority figures should be rounded to the nearest 5. Percentages that would round to zero (eg less than 0.5%) but are not zero should be replaced with a note to say the % is negligible).  Percentages should be rounded to whole numbers but where the numerator is five or fewer or the denominator was 10 or fewer, they should be suppressed and replaced by  a (c). For CLA data, zero is allowed.  For CLA matched with NPD data, zero is not allowed. Even if only one cell requires cell suppression, you must suppress at least one other component cell (the next smallest) to avoid calculation of suppressed values from the totals. CLA percentages should be rounded to 0 decimal places. See CLA methodology document  https://explore-education-statistics.service.gov.uk/methodology/children-looked-after-in-england-including-adoptions-methodology#content-section-3-content-5. For CLA data matched with NPD – it isrounded to the same number of decimal places as in the national publications for each outcome. (so for example KS2 data is rounded to 0dp) https://explore-education-statistics.service.gov.uk/methodology/outcomes-for-children-in-need-including-children-looked-after-by-local-authorities-in-england-methodology Percentages that would round to zero (eg less than 0.5%) but are not zero should be replaced with a note to say the % is negligible).  Outcomes for Children in Need including Children looked After by Local Authorities in England Data or findings from data yet to be released by DfE (including national and regional comparators) as additional tables in the Outcomes for Children in Need (including children looked after by local authorities in England) statistical publication should not be shared any more widely without prior permission from DfE. NB:- DfE publishes national and regional CIN/CLA figures rounded to the nearest 10; if LAs had the unrounded (regional) figure they may in some instances be able to calculate suppressed figures within DFE main publications. NCER & LAs should always use rounded regional and national figures. Children In Need (CIN) data: Any published CIN or CIN matched with NPD figures at a local level must be based on no lower than 6 cases or else suppressed (e.g. replaced with an c , with a note “c in this table means the figures have been suppressed due to small numbers”). National and regional figures should be rounded to the nearest 10. Percentages based on fewer than 6 cases must also be suppressed with a “c”. Percentages should be rounded to one decimal place. For CIN data zero is allowed. For CIN matched with NPD data, zero is not allowed and should be suppressed with a c. Even if only one cell requires cell suppression, you must suppress at least one other component cell (the next smallest) to avoid calculation of suppressed values from the totals. The DfE have imposed strict usage limitations for the 2020 / 21 CIN data in Nexus. These limitations are imposed on LA’s due to early visibility of aggregated national data.  Unamended data: As unamended attainment data has not been checked by schools, it should only be used by those with a legitimate need to use it for managerial, operational or other appropriate decision-making purposes at an early stage. Data for a specific LA can be shared with that LA ONLY. Any sharing of data or publishing beyond that would require NCER to seek approval from DfE. School level CIN/CLA data for a specific LA is for that LAs internal use only and should not be published in any form. In addition, nether the unamended attainment data nor the CLA should NOT be used for:- ·       Issuing public statements based on the information before the official statistics are published ·       Supporting direct action on schools or dictating school funding ·       Publication or public pronouncement of any analyses at aggregate or school level without permission from the DfE No school-level figures will be published based on unamended data without prior DfE permission.
Special conditions for DfE Data Extracts (if any)	Further CIN/CLA rules: Pupil level CIN/ CLA data and the matched results are only visible to the Local Authority that has corporate parenting responsibility for the pupil. Regional and National aggregations are calculated in the system for comparison purposes only with DfE suppression rules applied to all early calculated fields (see Standard Disclosure Control above). School level data for CIN/CLA should not be produced. Summary reports provide outcomes for an individual Local Authorities virtual school cohort and comparison data with DfE suppression supplied. This data should not be published in any way. The purpose of this early analysis is to support effective interventions at the earliest opportunity for some of the most vulnerable pupils nationally. The DfE have imposed strict usage limitations for 2020/ 2021 CIN/ CLA data in Nexus. These limitations are imposed on LAs due to early visibility of aggregated national data. Further publication rules: NCER should contact and gain agreement from DfE before releasing anything from the dataset publicly. Any data to be publicly released will be shared with the DfE at least one week in advance of planned release in order to ensure there is time for DfE to comment. In relation to Standard Disclosure Control of outputs, the term “published” relates to any DfE data that is shared with an audience beyond the Local Authority with corporate parenting responsibility for the pupil. Any DfE data needs to be suppressed by a Local Authority if exported from Nexus and added to reports or presentations for employees that are made available to an audience beyond the local authority, including but not limited to publication of scrutiny meeting minutes, regional meetings and National presentations where an LA is using DfE data.  Local authority compliance: NCER agree that all local authorities who can access DfE data through NCER services offered must read and comply with all restrictions as outlined in this Agreement. Failure to do this may result in the DfE withdrawing access to the DfE data and in future years. Local Authorities have the ability and access to export a flat file of their pupils educated within their LA / School or a Virtual School flat file can be exported “only where corporate parenting responsibility exists” or there is a safeguarding issue. Under no other circumstances should a Local authority have access to another local authority’s data. KS4 and KS5 2019/20 data: Under this Agreement, the provision of KS4 and KS5 2019/20 data has been permitted. However, whilst directly identifiable establishment IDs will be provided in accompanying datasets, no directly identifiable establishment IDs will be provided within the KS4 and/or KS5 2019/20 data extract. As the Department have made a commitment to not hold schools and colleges to account on the basis of exams and assessment data from summer 2020, NCER and member authorities agree that they will not attempt to use the KS4 and/or KS5 2019/20 data for accountability purposes, and furthermore NCER will not attempt to re-identify, extract or publish any establishment level data from the data they hold.