Platform Analytics

Live, citizen-generated data on Ghana's public appointees - aggregated, anonymised, and updated continuously.

- votes cast - active appointments - avg rating
ApexCharts simple-statistics D3 scales
Total Votes
-
Registered Users
-
Avg Platform Rating
-
out of 5.0
Vote Withdrawal Rate
-
Analytics Interpreter

Plain-language reading of the platform data, backed by statistical summaries and confidence cues.

Explainer Layer
Loading
--

Preparing interpretation.

Monthly Voting Activity & Rating Trend

Vote volume and average rating over the last 18 months, with 3-month linear projection.

Linear Regression
Projection uses ordinary least-squares linear regression on the 6 most recent data points (y = mx + b). Shaded region represents +/-0.3 confidence band.
Rating Distribution

How citizens distributed their scores across 0.0-5.0 intervals.

Frequency Histogram
National Performance Categories

Aggregate average scores across all 15 governance dimensions, nationally.

Daily Voting Activity (Last 30 Days)

Day-by-day vote count showing engagement peaks.

Time Series
Top 10 Performers
Loading...
Governance Index Spread
Histogram
Governance Index = 70% avg rating + 20% vote confidence + 10% trend score. Scale: 0-100.
Sector Performance
Voter Age Groups
Regional Participation

Academic Insights

Detailed statistical analysis, methodology, and comparative tools for researchers and policy analysts.

Advanced
Descriptive Statistics - Rating Distribution
Sample size (n): total active votes | mu: arithmetic mean | sigma: population standard deviation | Skewness: Pearson's median skewness = 3(mu - median)/sigma
Evidence Notes

Automatically generated notes that explain what the charts are saying and where the data is still thin.

Vote Volume vs Rating Correlation

Each point = one appointment. Tests whether more popular figures rate higher.

Pearson r: -  |  Interpretation: -
Methodology
Governance Index = 0.70 x (mu/5) + 0.20 x C + 0.10 x T (x100)
where:
  • mu = mean general rating (0-5)
  • C = vote-volume confidence, log10(n+1) / log10(5001), capped at 1
  • T = trend score from last 30 vs 31-60 days, normalised to 0-1
Confidence tiers: n < 50 = low; n < 500 = moderate; n < 5000 = strong; n >= 5000 = very strong.
All vote-level data is anonymised before aggregation. No PII appears in any analytics endpoint.
Compare Tool - Combine Datasets

Select up to 5 sectors to overlay their 15-category performance for comparison.

Pick at least one sector above to draw the overlay.

Confidence-Weighted Top 10

Same as Top 10, but rankings adjusted by sample-size confidence to penalise small-n inflations.

Score Dispersion by Category

Higher sigma indicates more polarised public opinion in that dimension.

Weight-Tuning Sandbox Hypothesis Mode

Adjust the Governance Index constants to test alternative weightings. Recalculates the Top 10 instantly - raw data is never modified.

00.701
00.201
00.101
GI = 0.70*(mu/5) + 0.20*log10(n+1)/log10(5001) + 0.10*T  x 100
Sum weights (should sum to 1.0)
1.00
Rank Movement (vs default)
0 changes
Re-ranked Top 10 (using your weights)
Monthly Rating Percentile Bands

Decomposes each month's votes into 10th, 50th (median), and 90th percentiles. Wide bands = high disagreement.