Team Reality Variance Analysis

Revealing where perceptions converge and diverge

Load Team Data

Format: Name, Function, then 32 question scores (1-5). First row should be headers. Avoid commas in names unless the field is quoted, e.g. "Compton, Rich".

Data Validation Errors

    Team Reality Variance Analysis Results

    respondents analysed

    Small Sample Size

    With fewer than 8 respondents, statistical measures like Krippendorff's Alpha and standard deviation become less reliable. The patterns are still informative, but interpret the precision of the numbers with appropriate caution.

    How to Read These Numbers

    α Krippendorff's Alpha

    0 0.67 0.80 1.0
    ≥ 0.80: Strong agreement
    0.67–0.79: Moderate agreement
    < 0.67: Weak agreement

    σ Standard Deviation (1-5 scale)

    0 0.7 1.1 2.0
    < 0.7: Tight consensus
    0.7–1.1: Moderate spread
    > 1.1: Significant divergence

    % Coefficient of Variation

    0% 15% 25% 50%+
    < 15%: Tight consensus
    15–25%: Moderate variation
    > 25%: Wide divergence

    What This Means

    Pattern Library

    Common organisational patterns. Suggested patterns are highlighted based on your team's data, but this is a starting point for exploration, not a diagnosis.

    Variance Heatmap

    Each cell shows how far an individual's score deviates from the team mean. Green indicates consensus; amber and red indicate divergence (regardless of direction).

    Near mean (±0.5)
    Moderate divergence (0.5-1.5 from mean)
    Strong divergence (≥1.5 from mean)

    Lens-Level Analysis

    Team averages with variance indicators. Wide bands suggest disagreement; narrow bands suggest consensus.

    Reading the Statistics
    SD (Standard Deviation)
    How spread out the scores are. Lower = more agreement.
    • Under 0.7 = tight consensus
    • 0.7–1.1 = normal variation
    • Over 1.1 = notable divergence
    α (Krippendorff's Alpha)
    Inter-rater reliability. Higher = more consistent ratings.
    • 0.80+ = strong agreement
    • 0.67–0.80 = acceptable agreement
    • Under 0.67 = low agreement

    Fault Lines

    Questions where the team diverges most. These are conversation starters, not problems per se.

    Individual Drift Analysis

    Who sees things most differently? Euclidean distance from the team centroid, with direction of divergence.

    Low drift (< 4): Close to team consensus
    Moderate drift (4-6): Notable differences
    High drift (> 6): Sees things quite differently

    Function Comparison

    How different functions see each lens. Gaps often reflect genuine differences in experience.

    Export Results