Skip to content

Latest commit

 

History

History
703 lines (489 loc) · 20.6 KB

DASHBOARD_VISUALIZATIONS.md

File metadata and controls

703 lines (489 loc) · 20.6 KB

Dashboard Visualizations Guide

Bishop State Community College Student Success Analytics & Predictive Models

Dataset: bishop_state_student_level_with_predictions.csv Students: 32,800
Date: October 28, 2025
Purpose: Comprehensive visualization guide for retention, graduation, and student success metrics

📊 EXECUTIVE DASHBOARD - KPI CARDS (Top Row)

1. Key Metrics Scorecard

Display as large, prominent cards at the top of dashboard:

┌─────────────────┬─────────────────┬─────────────────┬─────────────────┐
│ Overall         │ Avg Predicted   │ Students at     │ Avg Course      │
│ Retention Rate  │ Retention Prob  │ High/Critical   │ Completion Rate │
│    XX.X%        │     XX.X%       │  Risk: XXX      │     XX.X%       │
└─────────────────┴─────────────────┴─────────────────┴─────────────────┘

Data Fields:

  • Retention (actual retention rate)
  • retention_probability (predicted retention)
  • retention_risk_category (count of High + Critical)
  • course_completion_rate (average)

Purpose: At-a-glance institutional health metrics

🚨 RISK & EARLY WARNING VISUALIZATIONS

2. Risk Alert Distribution

Chart Type: Donut Chart or Pie Chart

Data Field: at_risk_alert

Categories:

  • 🟢 LOW: 4,146 students (12.6%)
  • 🟡 MODERATE: 19,823 students (60.4%)
  • 🟠 HIGH: 8,344 students (25.4%)
  • 🔴 URGENT: 487 students (1.5%)

Display: Show both percentages and counts

Insight: Immediate view of how many students need intervention

3. Retention Risk Funnel

Chart Type: Horizontal Bar Chart

Data Field: retention_risk_category

Categories (ordered by severity):

  1. Critical Risk (242 students - 0.7%)
  2. High Risk (15,755 students - 48.0%)
  3. Moderate Risk (15,202 students - 46.3%)
  4. Low Risk (1,601 students - 4.9%)

Color Scheme: Red → Orange → Yellow → Green

Insight: Retention risk pipeline visualization

4. Risk Score Distribution

Chart Type: Histogram with color gradient

Data Field: risk_score (0-100)

Bins:

  • 0-25 (Green)
  • 25-50 (Yellow)
  • 50-75 (Orange)
  • 75-100 (Red)

Insight: Shows concentration of risk across student population

5. At-Risk Students by Program

Chart Type: Horizontal Bar Chart (Top 10)

Data Fields:

  • Program_of_Study_Year_1 (x-axis)
  • Count of students where at_risk_alert = 'HIGH' or 'URGENT' (y-axis)

Sort: Descending by count

Insight: Identifies programs needing most support resources

🎓 ACADEMIC PERFORMANCE VISUALIZATIONS

6. Retention Rate by Cohort

Chart Type: Line Chart with dual lines

Data Fields:

  • X-axis: Cohort (2019-20, 2020-21, etc.)
  • Y-axis: Average retention rate
  • Line 1: Actual Retention
  • Line 2: retention_probability (predicted)

Insight: Year-over-year retention trends and prediction accuracy

7. GPA Performance Distribution

Chart Type: Box Plot or Violin Plot

Data Fields:

  • Groups: gpa_performance (On Track / Above Expected / Below Expected)
  • Values: average_grade

Insight: Grade distribution across performance categories

8. Course Completion Rate vs Retention

Chart Type: Scatter Plot with trendline

Data Fields:

  • X-axis: course_completion_rate
  • Y-axis: retention_probability
  • Color: at_risk_alert (LOW/MODERATE/HIGH/URGENT)
  • Size: Optional - total_courses_enrolled

Add: Linear regression trendline

Insight: Relationship between completing courses and staying enrolled

9. Credits Earned Progress

Chart Type: Stacked Bar Chart

Data Fields:

  • Number_of_Credits_Earned_Year_1
  • Number_of_Credits_Earned_Year_2
  • Number_of_Credits_Earned_Year_3
  • Number_of_Credits_Earned_Year_4

Group by: Cohort or Program_of_Study_Year_1

Insight: Academic progress tracking over time

10. Gateway Course Completion Impact

Chart Type: Grouped Bar Chart

Data Fields:

  • Group 1: CompletedGatewayMathYear1 (Yes/No)
  • Group 2: CompletedGatewayEnglishYear1 (Yes/No)
  • Y-axis: Average retention_probability

Why Important: Gateway courses are critical milestones

Insight: Impact of completing foundational courses on retention

👥 DEMOGRAPHIC & EQUITY ANALYSIS

11. Retention by Demographics

Chart Type: Multiple Grouped Bar Charts (one for each demographic)

Create separate charts for:

  • Race: Average retention by racial group
  • Gender: Male vs Female vs Other
  • First_Gen: First-generation vs continuing-generation
  • Pell_Status_First_Year: Pell-eligible vs not
  • Student_Age: Age group brackets

Data Fields:

  • Demographic field (x-axis)
  • Average Retention (actual)
  • Average retention_probability (predicted)

Insight: Equity gaps and support needs across populations

12. First-Gen Students Risk Profile

Chart Type: Stacked Bar Chart

Data Fields:

  • X-axis: First_Gen (Yes/No)
  • Y-axis: Count of students
  • Stacks: at_risk_alert levels (LOW/MODERATE/HIGH/URGENT)

Insight: Disproportionate risk in first-generation population

13. Math Placement vs Retention

Chart Type: Grouped Bar Chart or Heatmap

Data Fields:

  • X-axis: Math_Placement levels (Below Transfer, Transfer Ready, etc.)
  • Y-axis: Average retention_probability

Why Important: Math placement is the #1 predictor (35.1% feature importance)

Insight: Placement test scores strongly predict outcomes

🔮 PREDICTIVE ANALYTICS VISUALIZATIONS

14. Predicted vs Actual Retention

Chart Type: Confusion Matrix Heatmap

Data Fields:

  • Rows: Actual Retention (0 = Not Retained, 1 = Retained)
  • Columns: retention_prediction (0/1)
  • Cell values: Count of students

Color: Blue gradient (darker = more students)

Insight: Model accuracy visualization

15. Time to Credential Distribution

Chart Type: Histogram with overlay

Data Fields:

  • Primary: predicted_time_to_credential (histogram)
  • Overlay: Actual Time_to_Credential for students who completed

X-axis: Years (1, 2, 3, 4, 5+)

Insight: Graduation timeline forecasting

16. Predicted Graduation Year

Chart Type: Area Chart or Vertical Bar Chart

Data Field: predicted_graduation_year

X-axis: Academic year (2023-24, 2024-25, 2025-26, etc.)
Y-axis: Count of expected graduates

Insight: Institutional planning for future cohorts

17. Credential Type Predictions

Chart Type: Stacked 100% Bar Chart or Stacked Area Chart

Data Fields (probabilities):

  • prob_no_credential
  • prob_certificate
  • prob_associate
  • prob_bachelor

Group by: Cohort or Program_of_Study_Year_1

Insight: Expected credential outcomes by program

18. Predicted GPA Performance

Chart Type: Donut Chart

Data Field: gpa_performance

Categories:

  • On Track
  • Above Expected
  • Below Expected

Insight: Academic performance forecast across student body

📍 GEOGRAPHIC & INSTITUTIONAL ANALYSIS

19. Retention by ZIP Code

Chart Type: Choropleth Map (Kentucky state map)

Data Fields:

  • Geography: zip_code
  • Color intensity: Average retention_probability

Color Scale: Red (low retention) → Green (high retention)

Insight: Geographic patterns in student success

20. At-Risk Students by Institution

Chart Type: Horizontal Bar Chart

Data Fields:

  • X-axis: Count of students with at_risk_alert = 'HIGH' or 'URGENT'
  • Y-axis: Institution_ID

Sort: Descending by count

Insight: Campus-level intervention prioritization

📈 TREND & COMPARATIVE ANALYSIS

21. Enrollment Intensity Impact

Chart Type: Grouped Bar Chart

Data Field: Enrollment_Intensity_First_Term

Groups: Full-Time vs Part-Time

Metrics to compare:

  • Average Retention
  • Average course_completion_rate
  • Average average_grade

Insight: Full-time vs part-time student outcomes

22. Online vs Face-to-Face Learning

Chart Type: Line Chart or Bar Chart

Data Field: pct_online (bucket into ranges)

Buckets:

  • 0-25% online
  • 25-50% online
  • 50-75% online
  • 75-100% online

Y-axis: Average retention_probability

Insight: Effectiveness of different delivery modalities

23. Dual Enrollment Impact

Chart Type: Comparison Bar Chart

Data Field: Dual_and_Summer_Enrollment (Yes/No)

Metrics:

  • Retention rate
  • Course completion rate
  • Average GPA
  • Average retention_probability

Insight: Impact of early college credit on success

24. Program of Study Performance

Chart Type: Horizontal Bar Chart (Top 20 programs)

Data Fields:

  • X-axis: Average retention_probability
  • Y-axis: Program_of_Study_Year_1
  • Optional: Bar width or annotation showing student count

Sort: Highest to lowest retention probability

Insight: Best and worst performing programs

🎯 INTERVENTION PRIORITY DASHBOARDS

25. High-Priority Students Table

Chart Type: Interactive Sortable Data Table

Columns to Display:

  • Student_GUID
  • risk_score
  • at_risk_alert
  • retention_probability
  • course_completion_rate
  • average_grade
  • Program_of_Study_Year_1
  • Cohort
  • Institution_ID

Interactive Features:

  • Sort by any column
  • Filter by:
    • Alert level (URGENT/HIGH)
    • Program
    • Cohort
    • Institution
    • Demographics

Export: Download filtered list to CSV for advisor outreach

Insight: Actionable student list for immediate intervention

26. Intervention Impact Simulator

Chart Type: Interactive What-If Analysis Dashboard

Input Controls (Sliders):

  • Improve course completion rate by: X%
  • Improve average GPA by: X points
  • Increase gateway course completion by: X students

Calculations:

  • Current retention probability
  • Adjusted retention probability (simulated)
  • Net change in expected graduates
  • Projected revenue impact

Display: Before/After comparison bars

Insight: ROI estimation for intervention programs

📊 RECOMMENDED DASHBOARD LAYOUTS

Page 1: Executive Overview

┌──────────────────────────────────────────────────────────────────┐
│  KPI CARDS                                                        │
│  [Overall Retention] [Predicted Retention] [At-Risk] [Completion]│
├────────────────────────────────┬─────────────────────────────────┤
│ Risk Alert Distribution        │ Retention Rate by Cohort        │
│ (Donut Chart)                  │ (Line Chart - Actual + Predicted│
├────────────────────────────────┼─────────────────────────────────┤
│ At-Risk Students by Program    │ Predicted Graduation Year       │
│ (Horizontal Bar - Top 10)      │ (Area Chart)                    │
└────────────────────────────────┴─────────────────────────────────┘

Audience: Presidents, VPs, Deans
Purpose: High-level institutional health

Page 2: Academic Performance Deep Dive

┌────────────────────────────────┬─────────────────────────────────┐
│ GPA Performance Distribution   │ Gateway Course Completion       │
│ (Box Plot)                     │ Impact (Grouped Bar)            │
├────────────────────────────────┼─────────────────────────────────┤
│ Course Completion vs Retention │ Credits Earned Progress         │
│ (Scatter Plot)                 │ (Stacked Bar by Year)           │
└────────────────────────────────┴─────────────────────────────────┘

Audience: Academic Affairs, Faculty
Purpose: Course and curriculum effectiveness

Page 3: Equity & Demographics Analysis

┌────────────────────────────────┬─────────────────────────────────┐
│ Retention by Race              │ Retention by First-Gen Status   │
│ (Grouped Bar)                  │ (Risk Profile Stacked Bar)      │
├────────────────────────────────┼─────────────────────────────────┤
│ Math Placement vs Retention    │ Retention by ZIP Code           │
│ (Bar Chart)                    │ (Kentucky Map)                  │
└────────────────────────────────┴─────────────────────────────────┘

Audience: Diversity/Equity Officers, Student Affairs
Purpose: Identify and address equity gaps

Page 4: Intervention & Advisor Tools

┌──────────────────────────────────────────────────────────────────┐
│ High-Priority Students Table (Sortable, Filterable)              │
│ [Show URGENT only] [Show HIGH only] [Filter by Program ▼]       │
├────────────────────────────────┬─────────────────────────────────┤
│ Risk Score Distribution        │ Predicted Credential Types      │
│ (Histogram)                    │ (Stacked Area)                  │
└────────────────────────────────┴─────────────────────────────────┘

Audience: Academic Advisors, Student Success Teams
Purpose: Daily operational intervention work

Page 5: Predictive Analytics & Planning

┌────────────────────────────────┬─────────────────────────────────┐
│ Predicted vs Actual Retention  │ Time to Credential Distribution │
│ (Confusion Matrix)             │ (Histogram)                     │
├────────────────────────────────┼─────────────────────────────────┤
│ Model Performance Metrics      │ Intervention Impact Simulator   │
│ (Scorecard)                    │ (Interactive Controls)          │
└────────────────────────────────┴─────────────────────────────────┘

Audience: IR/Analytics Teams, Researchers
Purpose: Model validation and strategic planning

🛠️ RECOMMENDED TOOLS & PLATFORMS

Option 1: Tableau or Power BI ⭐ RECOMMENDED

Pros:

  • Professional, interactive dashboards
  • Easy drag-and-drop interface
  • Built-in filters and drill-downs
  • Mobile-responsive
  • Easy sharing with stakeholders

Best for: Executive and advisor-facing dashboards

Option 2: Python (Plotly Dash / Streamlit)

Pros:

  • Full customization
  • Can integrate live ML model predictions
  • Open-source and free
  • Can embed complex calculations

Best for: Data science teams, custom analytics

Sample Stack:

import pandas as pd
import plotly.express as px
import streamlit as st

Option 3: Looker / Google Data Studio

Pros:

  • Cloud-based
  • Good for Google Workspace integration
  • Free tier available (Data Studio)

Best for: Budget-conscious institutions

Option 4: Excel / Google Sheets (Quick Prototype)

Pros:

  • Universally accessible
  • Quick to build
  • Pivot tables and charts

Cons: Limited interactivity

Best for: Initial proof-of-concept

📦 INTERACTIVE FEATURES TO INCLUDE

Essential Filters (Apply to all pages)

  • ☑️ Cohort: 2019-20, 2020-21, 2021-22, etc.
  • ☑️ Program of Study: Filter by major/program
  • ☑️ Institution ID: Campus selection
  • ☑️ Risk Level: URGENT/HIGH/MODERATE/LOW
  • ☑️ Demographics: Race, Gender, First-Gen, Pell Status

Interactive Actions

  1. Drill-downs: Click on a risk category → see student list
  2. Hover tooltips: Show detailed stats on hover
  3. Cross-filtering: Select a program → all charts update
  4. Export buttons: Download filtered data to CSV
  5. Refresh data: Update predictions with latest data

Alert Features

  • 🔔 Email notifications for new URGENT students
  • 📊 Weekly digest of risk category changes
  • 🚨 Dashboard alerts for sudden drops in retention probability

🎯 KEY INSIGHTS TO HIGHLIGHT IN DASHBOARD

From Model Performance:

  1. Math placement is the #1 predictor (35% importance) → Show prominently
  2. Course completion rate strongly predicts retention → Track closely
  3. Gateway courses in Year 1 are critical → Monitor completion
  4. First-gen students face higher risk → Equity focus area

Actionable Metrics:

  • 487 students need URGENT intervention (1.5%)
  • 8,344 students at HIGH risk (25.4%)
  • Average predicted retention probability: Calculate from data
  • Programs with lowest retention: Identify bottom 5

Success Stories to Track:

  • Students who moved from HIGH to LOW risk
  • Programs with improving retention trends
  • Impact of interventions (before/after comparison)

📈 SAMPLE DASHBOARD METRIC CALCULATIONS

Overall Institutional Retention Rate

Total Students Retained / Total Students * 100

Predicted Retention Rate

Average(retention_probability) * 100

Intervention Success Rate (if tracking interventions)

(Students who improved risk category) / (Students who received intervention) * 100

Program Risk Score

Average(risk_score) by Program_of_Study_Year_1

Gateway Completion Rate

(CompletedGatewayMathYear1 = Yes) / (AttemptedGatewayMathYear1 = Yes) * 100

🚀 IMPLEMENTATION ROADMAP

Phase 1: Essential Dashboard (Week 1-2)

  1. KPI cards
  2. Risk alert distribution
  3. Retention by cohort
  4. High-priority students table

Goal: Get advisors using the system

Phase 2: Academic Insights (Week 3-4)

  1. GPA and completion visualizations
  2. Gateway course impact
  3. Program performance analysis

Goal: Inform curriculum decisions

Phase 3: Equity & Demographics (Week 5-6)

  1. Demographic breakdowns
  2. First-gen analysis
  3. Geographic mapping

Goal: Address equity gaps

Phase 4: Advanced Analytics (Week 7-8)

  1. Predictive model validation
  2. Time to credential forecasting
  3. Intervention simulator

Goal: Strategic planning and ROI

📚 ADDITIONAL RESOURCES

  • Model Performance Details: See ML_MODELS_GUIDE.md
  • Data Dictionary: See DATA_DICTIONARY.md
  • Raw Data: bishop_state_student_level_with_predictions.csv

💡 BEST PRACTICES

  1. ✅ Start simple: Launch with KPIs and risk alerts first
  2. ✅ Test with advisors: Get feedback from end users early
  3. ✅ Update regularly: Refresh predictions monthly or quarterly
  4. ✅ Tell stories: Use annotations to explain significant trends
  5. ✅ Make it actionable: Every chart should suggest an action
  6. ✅ Protect privacy: Ensure student data security and FERPA compliance

Questions or need help building specific visualizations?
Contact the data analytics team or refer to visualization tool documentation.

Document Version: 1.0
Last Updated: October 28, 2025