Skip to content

BdM-15/capture-insights / skills/data-analyzer

data-analyzer

Verified

Advanced data analysis, pattern detection, and insight generation from structured and unstructured datasets. Use when the user wants to analyze data, perform statistical analysis, find insights, detect patterns, identify anomalies, compare segments, test hypotheses, or generate data-driven recommendations. Triggers on phrases like 'analyze data', 'data analysis', 'find insights', 'analyze dataset', 'statistical analysis', 'find patterns', 'compare groups', 'test hypothesis', 'correlation analysis', or 'trend analysis'.

FileKindSize
SKILL.mdskill md8.6 KB
assets/report_template.mdmarkdown1.7 KB
evals/evals.jsonconfig2.0 KB
references/pitfalls.mdmarkdown4.0 KB
references/statistical-methods.mdmarkdown4.8 KB
references/visualization-guide.mdmarkdown3.6 KB

references/statistical-methods.md

View on GitHub
# Statistical Methods Reference

## Descriptive Statistics

### Central Tendency

| Measure | Use                     | Notes                 |
| ------- | ----------------------- | --------------------- |
| Mean    | Symmetric distributions | Sensitive to outliers |
| Median  | Skewed distributions    | Robust to outliers    |
| Mode    | Categorical data        | Can be multimodal     |

### Dispersion

| Measure            | Formula / Notes                     |
| ------------------ | ----------------------------------- |
| Range              | max − min                           |
| Variance           | Average squared deviation from mean |
| Standard Deviation | √Variance; same units as data       |
| IQR                | Q3 − Q1; robust to outliers         |

### Distribution Shape

- **Skewness** — positive (right tail), negative (left tail)
- **Kurtosis** — leptokurtic (heavy tails), platykurtic (light tails); excess kurtosis = kurtosis − 3

### Percentiles

- Quartiles: Q1 (25th), Q2 (50th / median), Q3 (75th)
- Outlier bounds (Tukey): below Q1 − 1.5×IQR or above Q3 + 1.5×IQR

---

## Inferential Statistics — Test Selection

### Comparing Means

| Scenario                   | Test               | Assumptions                               |
| -------------------------- | ------------------ | ----------------------------------------- |
| One sample vs. known value | One-sample t-test  | Normality                                 |
| Two independent groups     | Independent t-test | Normality, equal variance (Levene's test) |
| Two paired groups          | Paired t-test      | Normality of differences                  |
| 3+ independent groups      | One-way ANOVA      | Normality, homogeneity of variance        |
| 3+ groups, 2 factors       | Two-way ANOVA      | Same as one-way                           |
| Non-parametric 2-group     | Mann-Whitney U     | No normality required                     |
| Non-parametric 3+ groups   | Kruskal-Wallis     | No normality required                     |

### Comparing Proportions

| Scenario                            | Test                            |
| ----------------------------------- | ------------------------------- |
| Single proportion vs. expected      | Chi-square goodness of fit      |
| Two categorical variables (large n) | Chi-square test of independence |
| Two categorical variables (small n) | Fisher's exact test             |

### Correlation

| Type                | Test       | When to Use                         |
| ------------------- | ---------- | ----------------------------------- |
| Linear (continuous) | Pearson r  | Both variables normally distributed |
| Rank-based          | Spearman ρ | Non-normal, ordinal, or monotonic   |
| Concordance         | Kendall τ  | Small samples or many ties          |

### Distribution Tests

| Test               | Purpose                           |
| ------------------ | --------------------------------- |
| Shapiro-Wilk       | Test normality (n < 50 preferred) |
| Kolmogorov-Smirnov | Compare distribution to reference |
| Levene's test      | Test homogeneity of variance      |

---

## Effect Size Measures

Effect size answers "how big is the difference?" — independent of sample size.

| Measure          | Formula              | Context                 | Thresholds (Cohen)                      |
| ---------------- | -------------------- | ----------------------- | --------------------------------------- |
| Cohen's d        | (μ1 − μ2) / SDpooled | Mean differences        | Small: 0.2, Medium: 0.5, Large: 0.8     |
| Eta-squared (η²) | SSeffect / SStotal   | ANOVA                   | Small: 0.01, Medium: 0.06, Large: 0.14  |
| Odds Ratio       | (a/b) / (c/d)        | Categorical association | OR=1 (no effect), OR>1 (increased odds) |
| Pearson r²       | r²                   | Correlation             | % of variance explained                 |
| Cramér's V       | √(χ²/n·min(r−1,c−1)) | Chi-square              | 0–1 scale; > 0.3 = moderate             |

**Rule of thumb:** Always report effect size alongside p-value. A statistically significant result with a tiny effect size may not matter practically.

---

## Regression Quick Reference

| Type            | Use                                      | Key Output                |
| --------------- | ---------------------------------------- | ------------------------- |
| Simple linear   | One predictor → continuous outcome       | Slope, R²                 |
| Multiple linear | Multiple predictors → continuous outcome | Coefficients, adjusted R² |
| Logistic        | Predictors → binary outcome              | Odds ratios, ROC/AUC      |

**Check:** Residual plots for linearity, homoscedasticity, normality of residuals; VIF for multicollinearity (VIF > 10 is problematic).
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 197kUpdated Claude CodeCodexVerifiedInstall