Links
Fischer Family Trust
Do not delete cell
Do not delete cell

Fischer Family Trust

The Fischer Family Trust is an independent non-profit organisation that undertakes and supports projects to develop education in the UK. Its public website is:

• www.fischertrust.org

It also operates a secure website, for which schools have individual passwords. This website provides analysis of school performance data. This website is:

• www.fftlive.org

The FFT produces a variety of reports to assist schools with self-evaluation and target-setting. Schools can access all these reports via the secure website (FFT Live). See below for guidance in using FFT Live.

The range of reports is as follows. Click on the report name for more information.

• Pupil-level Estimates - looking forwards to inform target-setting
• School-level Estimates - looking forwards to inform target-setting
• Value Added Pupil  - looking retrospectively at pupils’ results in comparison with estimates
• Significant Areas Grid  - looking retrospectively at the last 3 years’ results – showing the 3-year average value added and indicating the direction of any trends
• Analyses to Support Self-Evaluation - a detailed booklet comprising retrospective analysis (including the Significant Areas Grid) and school level estimates of future performance.

These reports are explained below.

Pupil Estimates

These can be produced via FFT Live. The website allows you to generate the traditional range of pupil estimates (A, B and D – explained below) or choose your own parameters. Estimates are produced for the pupils on roll at the school at the time of the previous January school census and for whom prior attainment data is known. The ‘traditional’ models used by FFT are as follows.

Each set of estimates shows the results that can be expected at the end of the Key Stage if pupils make the following rates of progress:

Type A – the same progress as similar pupils nationally
(‘similar pupils’ here refers to pupils with the same prior attainment, gender and month of birth)

Type B – the same progress as similar pupils in similar schools nationally
(‘similar schools’ here refers to schools with similar levels of prior attainment and economic deprivation)

Type C – the progress required for the Local Authority to meet its target

Type D – the same progress as similar pupils in the top quartile of similar schools

The new style of pupil estimates, however, allows users to generate a wider range of pupil estimates, by selecting the desired estimate rank (from median to 5th percentile) and estimate basis (PA or SE – Prior Attainment or Socio Economic). As a guide, to connect this new approach and the previous approach:

Type ‘A’ estimate = PA model, rank 50
Type ‘B’ estimate = SE model, rank 50
Type ‘D’ estimate = SE model, rank 25

Therefore, if desired, schools could create estimates that are more challenging than type D e.g. by selecting the 10th percentile.
 
The report shows a range of probabilities, indicating the likelihood of achieving each level, based on the current national patterns and using the options you have selected.  The idea is that these likelihoods can then form the basis of a discussion with staff and with your SIP, so that challenging and realistic targets can be set. (See example below.)

Key:

FFT Live also includes a pupil estimate summary report which shows probabilities and estimated levels based upon the models A, B and D (see below).

How to interpret probabilities:
If a group of similar pupils all have a 25% probability of achieving Level 4, don’t simply assume that it is too unlikely for any of them to achieve Level 4. Instead, interpret 25% as meaning that it is likely that 1 in 4 of those pupils will achieve Level 4.

• top

School Estimates

The amalgamation of the pupil-level estimates produces school-level estimates, which are broken down by gender, prior attainment, ethnicity etc, as shown in the example below.

There are many pages to this report – showing probabilities relating to a variety of measures, including the ‘2 levels progress’ measure and attainment at level 4+ in both English and maths.
The reports show – in ascending order – the estimates based upon models A, B and D and an estimate (in green) based upon the school’s own historical rate of progress.
School Improvement Partners are likely to focus their discussions with schools around Type D estimates, as these represent the greatest level of challenge.

• top

Value Added Pupil Report

After the cohort has taken the end of Key Stage tests, these reports compare the actual results obtained with the pupils’ estimated results. These estimates are based on the average performance for ‘similar’ pupils nationally.
This report lists the pupils in the order of their estimated results (from highest expected result to lowest). The report highlights any actual results that are more than half a level above or below the estimates – in other words, results showing significantly high or low Contextual Value Added.
Three different versions of this report can be produced, each using a slightly different definition of ‘similar pupils’. These are called the ‘PA’ model, the ‘SE model’ and the ‘SX model’.

The PA (Prior Attainment) model produces the pupil estimates on the basis of pupils’ prior attainment, gender and month of birth.

The SE (Socio-Economic) model adjusts the PA model estimates by school-level context factors (e.g. % pupils on SEN register, % on FSM etc.)

The SX (School-EXtended) model produces the estimates using all of the above factors plus pupil-level contextual factors (SEN, EAL, FSM, socio-economic background etc.).

The PA model is equivalent to the Type A estimate produced for target-setting purposes and the SE model is equivalent to the Type B model – however please note that the estimates are re-calculated using the most recent national patterns of progress and may therefore differ from the pupil estimates produced for the cohort in the previous academic year.

The SX model is only used when analysing how pupils have performed, never for producing targets for future performance. This is because pupil-level context factors should not be used when target-setting, as to do so could cause under-expectation of particular pupil groups to become embedded.

• top

Significant Areas Grid

The significant areas grid reflects the past 3 years’ performance. Actual results are compared against the FFT estimate for each indicator (e.g. % attaining Level 4 and above).  As the FFT estimates are based on contextual factors and prior attainment (SX model), this provides a measure of contextual value added.

Where the actual result significantly exceeds the estimate across the average of the 3 years, the box is green. 

Where the actual result is significantly lower than the estimate across the average of the 3 years, the box is blue.

Where the actual result is not significantly different to the estimate across the average of the 3 years, the box is white.

Arrows are used to indicate improving, declining or varying trends across the 3-year period.

NB there is a Significant Areas Grid for Key Stage 1 in the Analyses to Support Self-Evaluation. However in this case it does not represent value added, as there is no measure of prior attainment included. Instead this compares the pupils’ outcomes with the outcomes attained by similar pupils nationally – using a range of contextual factors.

• top

Analyses to Support Self-Evaluation

This report brings together into one document several of the analyses described above, providing very detailed analysis of end of Key Stage performance, indicating which groups of pupils within the school have performed significantly better or worse than similar pupils performed nationally.
 
For more information regarding calculation of Contextual Value Added, and reasons why the FFT measures of CVA may differ from the Ofsted measures shown in RAISEonline or PANDA, see this page

• CVA explained

• top

Online Analysis (www.fftlive.org)

To access FFT Live, log in using the username and password that have been sent to your Headteacher.  Please take a moment to read the Data Sharing Protocol, available from the main log-in screen.

For guidance in using this tool and interpreting the analysis, there is a wide range of resources available from the FFT public website.

Go to www.fischertrust.org then click on Data Analysis Project, and then on Downloads.
There are also resources within the Help menu of FFT Live.

Printing reports from FFT Live
Please note that the easiest way to print reports from FFT Live is to first export them to pdf format. Printing directly from within your browser may not result in all the coloured boxes being printed – depending on your browser’s settings.

• top