Insights Overview
Zion’s Insights module automatically analyzes your school’s data to surface actionable patterns and trends. Using statistical analysis, Zion identifies opportunities for intervention before problems escalate.
What are Insights?
Insights are automatically-generated, data-driven observations about patterns in your school’s performance data. Unlike raw reports that show what happened, insights explain what the data means and what you should do about it.
Traditional Analysis (Manual):
- Review reports and look for patterns
- Mentally compare data across time periods
- Requires statistical expertise
- Time-consuming and error-prone
- Easy to miss subtle patterns
Zion Insights (Automated):
- Automatic pattern detection
- Statistical validation of findings
- Confidence scoring for reliability
- Relevance ranking by impact
- Actionable recommendations
How Insights Work
Automatic Detection
Zion analyzes your data continuously, looking for:
- Patterns: Recurring behaviors in the data
- Anomalies: Values outside normal ranges
- Correlations: Relationships between different metrics
- Trends: Changes over time
Statistical Validation
Each potential insight is validated:
- Sample size requirements ensure reliability
- Statistical tests confirm significance
- False positives are filtered out
- Only meaningful patterns surface
Confidence Scoring
Every insight includes a confidence level:
| Level | Sample Size | Reliability |
|---|
| High | Large samples | Very reliable |
| Medium | Moderate samples | Reliable |
| Low | Small samples | Interpret with caution |
Relevance Scoring
Insights are ranked 0-100 by relevance:
- Impact Score (0-40): How significant is the finding?
- Confidence Score (10-30): How reliable is the data?
- Sample Score (0-30): Is the sample size adequate?
Higher scores indicate more actionable insights.
Types of Insights
Day-of-Week Patterns
What It Detects:
Performance variations by day of week
Example Insight:
“Goal achievement is 25% lower on Mondays compared to the weekly average”
What to Do:
- Investigate Monday-specific challenges
- Consider adjusted goals for Mondays
- Address weekend-to-school transitions
Subject Correlations
What It Detects:
Students struggling in multiple related subjects
Example Insight:
“Students struggling in Math 1042 are also struggling in Science 1040 (r=0.78)”
What to Do:
- Identify shared learning challenges
- Check for prerequisite gaps
- Consider paired tutoring
What It Detects:
Learning centres performing above or below average
Example Insight:
“Centre A is performing 18% below the school average”
What to Do:
- Investigate centre-specific factors
- Consider supervisor support
- Review resource allocation
Seasonal Trends
What It Detects:
Performance changes within a term
Example Insight:
“Achievement has declined 15% in the second half of this term”
What to Do:
- Address term fatigue
- Review workload distribution
- Consider engagement activities
Student Velocity
What It Detects:
Individual students speeding up or slowing down
Example Insight:
“12 students have shown >25% improvement in the last 4 weeks”
What to Do:
- Celebrate and reinforce positive momentum
- Identify what’s working for these students
- Watch for students slowing down
Subject Difficulty
What It Detects:
PACEs with unusually low completion rates
Example Insight:
“Math 1048 has a 62% completion rate, significantly below average”
What to Do:
- Review PACE content difficulty
- Provide additional support resources
- Consider teaching approach changes
The Insights Dashboard
Accessing Insights
- Navigate to Analytics in the sidebar
- Click Insights submenu
- View the insights dashboard
Dashboard Layout
Summary Cards:
- Total active insights
- High confidence count
- Medium confidence count
- Average relevance score
Filter Options:
- By confidence level (High/Medium/Low/All)
- By insight type (6 types or All)
Actions:
- Refresh: Update insight list
- Generate: Trigger new insight analysis
- Export: Download insights report
Insight Cards
Each insight displays as a card showing:
- Type Icon: Visual indicator of insight category
- Title: Brief description of finding
- Description: Detailed explanation
- Confidence Meter: Visual progress bar
- Relevance Score: 0-100 ranking
- Created Date: When insight was generated
Working with Insights
Viewing Details
Click View Details on any insight card to see:
- Full description and context
- Affected students list (up to 20)
- Student achievement data
- Export option for deeper analysis
Dismissing Insights
If an insight is not relevant:
- Click Dismiss on the insight card
- Insight is hidden from view
- Recorded as dismissed (with timestamp)
- Won’t appear in future until new data warrants
Dismissing: Dismissed insights are permanently hidden. Only dismiss if you’ve addressed the issue or determined it’s not actionable.
Generating New Insights
Insights generate automatically weekly, but you can trigger manually:
- Click Generate Insights button
- System analyzes current data
- New insights appear within minutes
- Old non-dismissed insights are replaced
Administrator Dashboard Integration
The main administrator dashboard includes an Insights Widget:
- Shows top 3 most relevant insights
- Quick preview format
- Click to view full details
- Link to complete insights page
Real-Time Updates
- Widget refreshes with dashboard
- Shows freshest insights
- Highlights new insights since last view
Confidence Meter
Understanding the Meter
The visual confidence meter shows:
- Full (100%): High confidence
- Two-thirds (66%): Medium confidence
- One-third (33%): Low confidence
Color Coding
- Emerald (Green): High confidence
- Blue: Medium confidence
- Gray: Low confidence
Sample Size Requirements
| Insight Type | High Confidence | Medium | Low |
|---|
| Day-of-Week | >40 samples | 20-40 | <20 |
| Correlations | >30 students | 20-30 | <20 |
| Centre Performance | >60 goals | 40-60 | <40 |
| Seasonal Trends | >150 goals | 100-150 | <100 |
Best Practices
Regular Review
- Check insights weekly at minimum
- Prioritize high-confidence findings
- Act on high-relevance scores first
Taking Action
- Analyze: Understand what the insight means
- Investigate: Look deeper into affected students/areas
- Plan: Develop intervention strategy
- Implement: Take concrete actions
- Monitor: Track improvement over time
- Dismiss: Once addressed, dismiss the insight
Building History
- Insights build on historical data
- More data = more reliable insights
- New schools see limited insights initially
- Full intelligence emerges over 4-6 weeks
Insight Validity
Expiration
Insights expire after 7 days:
- Data continues to change
- Old insights may no longer be relevant
- System generates fresh insights weekly
Staleness Prevention
- Expired insights auto-hide
- Only current, relevant insights shown
- Manual generation creates fresh analysis
