Table of Contents
For Counselors: Admission Intelligence FAQ
We're excited to share Admission Intelligence to all Scoir customers. 🚀
Admission Intelligence—or, as we like to call it, Scoir AI—are data-informed tools to help balance college lists and demystify acceptance chances, powered by tens of millions of de-identified outcome records.
Here are a few tools you can now access:
Balanced List Scores: Streamline the creation and evaluation of college lists, quickly identify students in need of support, and effectively track their progress from early list building to application planning. Learn more
Intelligent Match: Save time and set personalized acceptance likelihood at scale for all students and colleges with Intelligent Match. Learn more
Predictive Chances: Beyond Scattergrams, you'll be able to view predicted outcomes based on tens of millions of de-identified application records from real, validated students on the Scoir network. Learn more
Now, let's dive into your questions.
General Questions
What is Admission Intelligence?
Admission Intelligence—or, as we like to call it, Scoir AI—are data-informed tools to help balance college lists and demystify acceptance chances, powered by tens of millions of de-identified outcome records.
Scoir AI does the time-consuming, manual aspect of data collection and initial analysis for you by compiling mountains of verified data for you, providing real-time insights tailored to each student’s unique academic profile.
It's designed to bring clarity and confidence to acceptance chances by providing credible, predictive insights inside the platform you're already using to support informed decision-making.
Balanced List Scores: Streamline the creation and evaluation of college lists, quickly identify students in need of support, and effectively track their progress from early list building to application planning.
Intelligent Match: Save time and set personalized acceptance likelihood at scale for all students and colleges with Intelligent Match.
Predictive Chances: Beyond Scattergrams, you'll be able to view predicted outcomes based on tens of millions of de-identified application records from real, validated students on the Scoir network.
Scoir AI does the time-consuming, manual aspect of data collection and initial analysis for you by compiling mountains of verified data for you, providing real-time insights tailored to each student’s unique academic profile.
It's designed to bring clarity and confidence to acceptance chances by providing credible, predictive insights inside the platform you're already using to support informed decision-making.
Balanced List Scores: Streamline the creation and evaluation of college lists, quickly identify students in need of support, and effectively track their progress from early list building to application planning.
Intelligent Match: Save time and set personalized acceptance likelihood at scale for all students and colleges with Intelligent Match.
Predictive Chances: Beyond Scattergrams, you'll be able to view predicted outcomes based on tens of millions of de-identified application records from real, validated students on the Scoir network.
Why is Scoir adding AI to the platform?
Students and families often encounter conflicting information from various online sources about their acceptance chances, leading to confusion and uncertainty about which data to trust.
Scoir built Admission Intelligence to bring clarity and confidence to this process and incorporated predictive insights directly within the platform.
With tens of millions of validated outcome records, and a multidimensional modeling approach, Scoir AI empowers counselors with trusted insights to guide families toward informed application strategies and decision-making.
Scoir built Admission Intelligence to bring clarity and confidence to this process and incorporated predictive insights directly within the platform.
With tens of millions of validated outcome records, and a multidimensional modeling approach, Scoir AI empowers counselors with trusted insights to guide families toward informed application strategies and decision-making.
What control do counselors have over when students can use these experiences?
These tools have been designed to fit into your unique processes. You will have the control to define balanced list requirements, and determine when you introduce balanced lists or automate their scoring.
Predictive Chances will be presented alongside Scattergrams in your view. Currently, you have the option to display or hide Predictive Chances to students and parents. However, please note this setting for Predictive Chances is subject to change after an evaluation period.
With customizable settings, you choose when, who, and which class years gain access to these new capabilities, keeping you in control while guiding students and their parents.
Predictive Chances will be presented alongside Scattergrams in your view. Currently, you have the option to display or hide Predictive Chances to students and parents. However, please note this setting for Predictive Chances is subject to change after an evaluation period.
With customizable settings, you choose when, who, and which class years gain access to these new capabilities, keeping you in control while guiding students and their parents.
What's the best way to get started with Scoir AI?
We encourage you to the review the blog For Counselors: Introducing Admission Intelligence to get a lay of the land.
Once you have a better understanding of the tools you can now access, you can begin by configuring your settings and familiarizing yourself with the tools.
Once you have a better understanding of the tools you can now access, you can begin by configuring your settings and familiarizing yourself with the tools.
How do I leave feedback?
We encourage you to leave feedback about the direction we’re taking Admission Intelligence in Scoir, and you don’t have to wait until it’s released. Feel free to share your thoughts about what you’ve learned so far by filling out this form.
Data Questions
What data is driving these predictions?
Tens of millions of de-identified application outcome records from real, validated students across the Scoir network.
Scoir leverages machine learning models that go beyond ’s models use a broad approach that goes beyond taking in just academic factors like GPA and test scores. These models also take into account characteristics such as first-generation status, geographic location, race/ethnicity, sex, and high school profile, ensuring a more comprehensive prediction.
Scoir considers the student’s and college’s location, including in-state and out-of-state admission trends. These trends are not specifically programmed into the model but are a result of analyzing historical outcomes that include the applicant’s state. This approach allows the model to naturally reflect location-based patterns in admissions without relying on preset rules.
Scoir leverages machine learning models that go beyond ’s models use a broad approach that goes beyond taking in just academic factors like GPA and test scores. These models also take into account characteristics such as first-generation status, geographic location, race/ethnicity, sex, and high school profile, ensuring a more comprehensive prediction.
Scoir considers the student’s and college’s location, including in-state and out-of-state admission trends. These trends are not specifically programmed into the model but are a result of analyzing historical outcomes that include the applicant’s state. This approach allows the model to naturally reflect location-based patterns in admissions without relying on preset rules.
How is Scoir protecting student data and privacy?
We continue to protect student PII. The models used to inform Admission Intelligence de-identify the student by stripping any personally identifiable information.
Intelligent Match Questions
What is Intelligent Match?
Intelligent Match will assist in scoring college lists by automatically assigning acceptance likelihoods based on each student’s Predictive Chances for every college.
What is acceptance likelihood?
Acceptance likelihood represents the overall selectivity or competitiveness of that college for each student, and is the primary value used to assess list balance.
There are 5 acceptance likelihood categories: Far Reach, Reach, Level, Likely, Very Likely. These labels can be adjusted within high school settings.
There are 5 acceptance likelihood categories: Far Reach, Reach, Level, Likely, Very Likely. These labels can be adjusted within high school settings.
How can acceptance likelihood be assigned to a college on a student’s list?
Intelligent Match will automatically assign acceptance likelihood for colleges on a student’s list based on their overall Predictive Chances, and the mapping of those chances to each acceptance likelihood value, as defined in the high school settings. Alternatively, acceptance likelihood can be manually assigned by a counselor or student (if the high school allows students to do this). Manually assigned values will be preserved over any predictions made by Scoir AI. Learn more
What if I disagree with the predictions from Intelligent Match?
A student’s chances of acceptance may deviate from predictions for a number of circumstances, including individual hook or application to a particular major or program. In such cases the counselor (and student if configured to allow) may override the acceptance likelihood set by Intelligent Match.
What if I still want to use the Match Level approach to assigning acceptance likelihood?
You can continue to leverage the Match Level approach by adding a Match Level on the student’s profile and choosing not to use Intelligent Match.
List Score Questions
What is a List Score?
A List Score is a personalized grade for a student's college list allowing counselors to easily identify students with balanced or imbalanced lists. Each student’s List Score is calculated by evaluating the student's college list to ensure it includes an appropriate mix of "Reach," "Level," and "Likely" schools.
Can I adjust “balanced list” requirements?
Yes. Scoir’s default is 2 likely or very likely, 3 level, and 2 reach or far reach schools. The high school may choose to adjust the minimums within each of those three buckets. Learn how
Can I control who sees balanced list scores?
Yes. The high school can choose if and when balanced list experiences are displayed to students and parents. We do encourage that high schools enable the features for all parties and leverage it as an accelerator to build a list, and a tool to inform consultation. Coupled with tasks and assignments, counselors could assign list building tasks to encourage all students to build an A-grade list by a deadline. Learn how
Can I adjust what colleges (Following, Applying, Applied) are considered in balanced list scores?
By default, list scores consider colleges in Following, Applying or Applied states, within a student’s My Colleges list.
However, counselors can decide when it’s time to focus on building an application list, at which point the student’s list score will be calculated based on those schools in Applying and Applied states within MyColleges. Learn how
However, counselors can decide when it’s time to focus on building an application list, at which point the student’s list score will be calculated based on those schools in Applying and Applied states within MyColleges. Learn how
Predictive Chances Questions
How does Scoir derive Predictive Chances?
Scoir has created machine-learning models for a large percentage of colleges based on de-identified historical application records created on Scoir or imported into Scoir as part of the high school onboarding process.
These de-identified records contain a variety of details about each applicant, their school, the round they apply to and the outcomes. When returning Predictive Chances for a student, the model considers details about that student, comparing outcomes for similar students within similar high schools and presents personalized Predictive Chances for each application round currently offered by that college.
These de-identified records contain a variety of details about each applicant, their school, the round they apply to and the outcomes. When returning Predictive Chances for a student, the model considers details about that student, comparing outcomes for similar students within similar high schools and presents personalized Predictive Chances for each application round currently offered by that college.
How does Scoir use Machine Learning to power these models?
Machine learning is a modern technology that learns patterns from historical data and makes predictions about future outcomes without relying on fixed rules or pre-defined formulas. Machine Learning systems improve their accuracy as they process more data.
To ensure confidence in our predictions, Scoir follows industry proven practices by training each model using real application outcomes (training data) and then we test each model’s predictive abilities using additional real application outcomes (testing data) that the model did not see at training time. Scoir repeats these steps to produce Machine learning models for thousands of colleges. As a result of “training” and “testing” these models on a regular basis, Scoir generates and deeply analyzes model performance metrics to determine which models are reliable enough to generate Predictive Chances for our counselors and families.
Each College has its own machine learning model, which allows each model to evaluate the unique trends of each college and how the various features of an applicant (e.g. location, academic performance, high school, personal characteristics, etc) influence the outcomes. As such the models do not require algorithms or formulas to account for certain conditions, but rather the models learn what factors have the greatest influence on the outcomes for each college.
To ensure confidence in our predictions, Scoir follows industry proven practices by training each model using real application outcomes (training data) and then we test each model’s predictive abilities using additional real application outcomes (testing data) that the model did not see at training time. Scoir repeats these steps to produce Machine learning models for thousands of colleges. As a result of “training” and “testing” these models on a regular basis, Scoir generates and deeply analyzes model performance metrics to determine which models are reliable enough to generate Predictive Chances for our counselors and families.
Each College has its own machine learning model, which allows each model to evaluate the unique trends of each college and how the various features of an applicant (e.g. location, academic performance, high school, personal characteristics, etc) influence the outcomes. As such the models do not require algorithms or formulas to account for certain conditions, but rather the models learn what factors have the greatest influence on the outcomes for each college.
Do Predictive Chances consider my school’s outcomes?
Yes, the models also take into account the outcomes of other schools on Scoir, especially those with characteristics similar to yours. This is why it’s essential to ensure your school’s outcomes are accurately recorded in Scoir
Do Predictive Chances consider just the GPA and standardized test scores?
No, Scoir’s models use a broad approach that goes beyond academic factors like GPA and test scores. The models also take into account characteristics such as first-generation status, geographic location, race/ethnicity, sex, and high school profile, ensuring a more comprehensive prediction. Additionally, GPAs are normalized behind the scenes, so the model evaluates all students on the same scale, regardless of differences in grading systems between high schools
Do Predictive Chances consider the student’s location?
Yes, student and college location are factors included in the machine learning models (see "How Does Scoir Use Machine Learning to Power these Predictions?"). As a result, the personalized Predictive Chances consider the outcome trends of the student's state for each given college. Therefore, the models are able to identify trends specific to States, including whether residing In-State or Out-of-State plays a role in admissions decisions.
Do Predictive Chances consider individual factors like course selection, rigor, test-optional, or extracurricular activities?
Not yet.
Please note: The adoption of tools like Secondary School Reports will enrich our models’ view of an applicant and afford more comprehensive predictions in the future.
Please note: The adoption of tools like Secondary School Reports will enrich our models’ view of an applicant and afford more comprehensive predictions in the future.
What years of data are included in the predictions?
Scoir has outcomes records dating back ~10 years and all of those records are currently included in the models. Scoir recognizes that admissions trends change over time and those trends can be different for each college. We are actively evaluating the performance of each model to ensure that each model is trained only with data that is consistent with recent trends for each individual college.
What’s the difference between Predictive Chances and Scattergram?
While they’re both outputs of historical outcomes data, Predictive Chances leverages millions of records to offer predicted chances of acceptance based on the outcomes of students with similar profiles. Scattergrams plot only the individual high school’s application outcomes.
Additional Resources
We hope this FAQ was useful in helping you better understand Scoir AI. Here are some additional resources to help manage your account and support students at every stage of their journey.
