Introduction
This article will help users to format files properly so that they may be securely uploaded to the ECRISS system. Each of the files for the growth model has the required fields along with a short explanation of how the files are used.
Growth Model File Upload Instructions
The following files come from growth model runs and feed into the School Improvement, My Students, and PLP apps.
1. LGM_GDF – LGM_GDF-Copy.csv This file contains information on how to draw the blue growth model curves and is also used to identify student projections based on their propensity. Important fields include:
• SystemEcraId
• GYear – year the curve should be applied to (spring year is always used)
• GTestTo – Criterion Name (e.g. Spring MAP)
• GSubject – Subject
• GGrade – Grade
• Propensity
• Projection
• UpperCurve – The upper curve value at the given propensity
• LowerCurve – The lower curve value at the given propensity
• PropensityId – This is NOT in the current upload but is expected to be used in the new version of the portal to provide a link between each propensity equation and the curves it matches with.
• SubTestId – This is NOT in the current upload but is expected to be used in the new version of the portal to identify tests more specifically (instead of using the criterion name and subject)
• AnchorYears – This is NOT in the current upload but may be used in the new version to identify which anchor years are included in the model build. Not sure if there is a plan to display this.
2. LGM_STUDENT – This file contains basic roster information. This is like your enrollments file with each student having one row per year. The fields include:
• StudentId
• StateId
• LastName
• FirstName
• TestGrade
• SystemEcraId
• SchoolEcraId
• TestYear
3. LGM_GROUPS – This file contains information on student designations and programs and tends to feed the accordians and populate the designations tile in the PLP. Fields include:
• SystemEcraId
• StudentId
• Category – e.g. Gender, Income, Program,
• Value – e.g. Male, IEP, Accelerated Math
• TestYear
4. LGM_STUFULL – This file gives information on the test scores, propensity, growth scores, and benchmarks. It includes a lot of fields but many are not used. The ones that are used from what I understand include:
• YScore – The score of the test (a score of “99999” is a projection score and is not displayed)
• StudentId
• SystemEcraId
• ToGrade – The grade that is displayed next to the score in the PLP, though the accordion in the PLP filters students using the grade in the LGM_STUDENT file.
• Other1 – Contains the PL (if available)
• Other2 – Contains the NP (if available)
• Propensity – Contains the student propensity (a score of 0 means there is no propensity and the data is loaded as non-growth data
• Projected – This is the projected score but I believe Daniel pulls this from the LGM_GDF file by matching the students propensity to that file and getting the propensity there.
• SResidualB – This is the growth score for the student. However, I believe Daniel pulled the growth score from the next field in order to avoid rounding issues.
• StuVag – This is the growth score x 100 with no decimals that Dr. McConnell outputs. Daniel takes this and divides by 100. This ensures that the rounding in R and on the portal are the same.
• SLevel – a code indicating if the student had a low (-1), average (0), or high (1) propensity, though I believe this isn’t used and it’s calculated someplace else on the portal with a rule from the propensity field.
• PLevel – a code indicating if the student had low (-1), typical (0) or high (1) growth, though I believe this isn’t used and it’s calculated someplace else on the portal with a rule from the Growth field.
• Criterion – Test name (e.g. Spring MAP)
• Subject – e.g Mathematics
• Achieve – A binary variable (1/0) indicating if the student met their growth target, though I believe this isn’t used and it’s calculated someplace else on the portal with a rule from projected and actual score.
• Makebench – A binary variable (1/0) indicating if the student met the benchmark, though I know this is calculated someplace else on the portal with a rule from actual score and benchmark.
• Benchmark – Indicates the benchmark (if available)
• Benchmarkphrase – Indicates the definition of the benchmark that is shown on the scatter plots (e.g. PARCC Equated Meets Cut Score)
• VALID – Indicates the number of valid predictors that went into the propensity. If it is 0 then the test is loaded as non-growth data.
• TestYear
5. PL_LABEL – This file provides labels for the numeric values in the PL field (Other1 in the STUFULL). While we load this separately for each district, this file is the same for almost all districts and may be considered a universal file instead of having to be district specific. It might be good to have the option of having district-specific labels in the future though (e.g. have SystemEcraId as 11111111 for universal labels and if a district has a specific way they want something shown then put their systemId in the field instead as an exception).
• SystemEcraId
• Criterion – e.g. Spring PARCC
• PerformanceLevel – Numberic field that matches the Other1 field (e.g. 4)
• PerformanceLevelLabel – Label that should show for this number (e.g. 4-Met Standards)
6. LGM_SUPER_GROUP – This file is only used if a district wants some groups showing in the accordion for one app or location but not another. This is not commonly used and was mostly in place when we wanted to load some programs as filters to the School Improvement app but didn’t want them to show up in the PLP app. Fields include:
• SchoolYear – This is 8 digits (e.g. 20152016) but does NOT filter by year. It is just a place holder in case the functionality to filter by year was needed in the future. The rules are currently applied across all years.
• SystemEcraId
• GroupId – This is just a unique number that we have to make up for each group. Not sure if or how it’s used.
• GroupName – This is the name of the group that is found in the LGM_GROUPS file. E.g. Chinese Immersion, Income, etc.
• SuperGroupName – This is the app that should show the listed group in the accordion (e.g. LGM, Reports, PLP, MyStudents).
7. LGM_PLACEMENT – This file feeds the Placement app only. It looks just like the LGM_STUDENT file and likely works the same way but only for the placement app. If a district wants to allow the high school to see current 8th graders who are in middle schools in the current year, we just load all of the 8th graders into the placement file with their current year information (e.g. grade 8, year 2017, etc.) but with the SchoolEcraId being set to the high school. This way the high school can see their scatter plots and view historical data in the placement app.