# [ Hypothesis Testing ] ( CheatSheet )

1. Basic Hypothesis Testing

● One-Sample T-Test: [Link].ttest_1samp(data, popmean)

● Two-Sample T-Test (Independent): [Link].ttest_ind(sample1,
sample2)
● Paired T-Test: [Link].ttest_rel(sample1, sample2)
● One-Way ANOVA: [Link].f_oneway(sample1, sample2, sample3)

2. Normality Tests

● Shapiro-Wilk Test: [Link](data)

● D'Agostino's K-squared Test: [Link](data)
● Anderson-Darling Test: [Link](data, dist='norm')

3. Correlation Tests

● Pearson Correlation Coefficient: [Link](x, y)

● Spearman's Rank Correlation: [Link](a, b)
● Kendall's Tau: [Link](x, y)

4. Comparing Variances

● Levene's Test: [Link](sample1, sample2)

● Bartlett’s Test: [Link](sample1, sample2)

5. Non-parametric Tests

● Mann-Whitney U Test: [Link](sample1, sample2)

● Wilcoxon Signed-Rank Test: [Link](x, y)
● Kruskal-Wallis H Test: [Link](sample1, sample2,
sample3)
● Friedman Test: [Link](sample1, sample2,
sample3)

By: Waleed Mousa

6. Proportion Tests

● Z-test for Proportions:

[Link].proportions_ztest(count, nobs)
● Chi-Squared Test for Proportions:
[Link].chi2_contingency(contingency_table)

7. Regression Tests

● Linear Regression Test: [Link](y, X).fit().summary()

● Logistic Regression Test: [Link](y,
X).fit().summary()

8. Time Series Analysis

● Augmented Dickey-Fuller Test (Stationarity Test):

[Link](data)
● Granger Causality Tests:
[Link](data, maxlag)

9. Post-hoc Tests

● Tukey's Honest Significant Difference Test:

[Link].pairwise_tukeyhsd(endog, groups)

10. Effect Size Calculations

● Cohen's d for T-Test: CohenEffectSize(sample1, sample2)

● Eta Squared for ANOVA: EtaSquared(ANOVA_Result)

11. Power Analysis

● Sample Size for T-Test:

[Link].tt_ind_solve_power(effect_size, alpha,
power)
● Sample Size for ANOVA:
[Link]().solve_power(effect_size,
alpha, power)

By: Waleed Mousa

12. Multiple Comparisons Correction

● Bonferroni Correction: [Link](pvals)

● False Discovery Rate (FDR) Correction:
[Link](pvals, method='fdr_bh')

13. Visualization of Statistical Tests

● QQ Plot for Normality Check: [Link](data, line='s')

● Boxplot for Comparing Groups: [Link](x="group", y="data",
data=df)
● Histogram for Data Distribution: [Link](data, bins)

14. Data Preparation for Hypothesis Testing

● Removing Outliers: df[([Link]([Link](df)) < 3).all(axis=1)]

● Log Transformation for Normality: np.log1p(df['column'])

15. Advanced Statistical Modeling

● Mixed Linear Models:

[Link].mixed_linear_model.MixedLM(endog, exog,
groups).fit()
● Survival Analysis: [Link]().fit(df,
duration_col='T', event_col='E')

16. Handling Missing Data for Tests

● Impute Missing Values:

[Link](strategy='mean').fit_transform(data)

17. Distribution Fitting

● Fit Distributions to Data: [Link].<distribution>.fit(data)

By: Waleed Mousa

18. Extracting Test Statistics

● Extract P-Value and Test Statistic: result =

[Link].ttest_1samp(data, popmean); p_value = [Link]

19. Bayesian Statistics

● Bayesian Model Comparison: [Link]({model1: trace1, model2:

trace2})

20. Structural Equation Modeling

● SEM Model Fitting: [Link](sem_model, data)

21. Multivariate Statistics

● MANOVA: [Link].from_formula('y1 +
y2 ~ x1 + x2', data).mv_test()

22. Advanced Non-parametric Tests

● Permutation Test: [Link].permutation_test(x, y, func,

method='approximate')

23. Custom Hypothesis Testing Functions

● Custom Test Function: def custom_test(data): /* implement test

logic */; custom_test(data)

24. Specialized Plots for Hypotheses

● Scatter Plot with Regression Line: [Link](x='x', y='y',

data=df)

25. Meta-analysis

● Meta-Analysis: metafor::rma(yi, vi, data=meta_data)

By: Waleed Mousa

26. Testing Assumptions for Parametric Tests

● Homogeneity of Variances (Levene's Test):

[Link](sample1, sample2)

27. Reporting Results

● Formatted Result Output: def report_result(pval, alpha): if pval <

alpha: print('Reject null hypothesis') else: print('Fail to reject
null hypothesis')

28. Working with Distributions in Hypothesis Testing

● Drawing Samples from a Distribution: [Link](loc=0,

scale=1, size=100)

29. Cross-Validation in Statistical Tests

● Cross-Validation for Model Testing:

sklearn.model_selection.cross_val_score(model, X, y, cv=5)

30. Interpretation of Test Results

● Interpreting Effect Sizes: def interpret_effect_size(d): /* logic

to interpret Cohen's d */

31. Advanced Correlation Analysis

● Point-Biserial Correlation: [Link](x, y)

● Partial Correlation: pingouin.partial_corr(data)

32. Effect Size Analysis

● Cohen's d for Independent Samples: Cohens_d(group1, group2)

● Cohen's d for Paired Samples: Cohens_d_paired(sample1, sample2,
paired=True)

By: Waleed Mousa

33. Bootstrapping Methods

● Bootstrap Resampling: bootstrap = [[Link](sample,

size=len(sample), replace=True) for _ in range(n_iterations)]

34. Bayesian Hypothesis Testing

● Bayesian T-Test: [Link](x, y)

● Bayesian ANOVA: [Link](model)

35. Nonparametric Bootstrap Confidence Intervals

● Bootstrap CI for Mean: bootstrapped_CI(sample, [Link], alpha=0.05)

● Bootstrap CI for Median: bootstrapped_CI(sample, [Link],
alpha=0.05)

36. Power and Sample Size Calculation

● Calculate Sample Size for T-Test:

[Link].tt_solve_power(effect_size, alpha, power)
● Calculate Power for T-Test:
[Link]().solve_power(effect_size,
nobs1, alpha)

37. Multiple Testing Correction

● Benjamini-Hochberg Procedure:
[Link](pvals, method='fdr_bh')
● Holm-Bonferroni Method:
[Link](pvals, method='holm')

38. Exploratory Data Analysis for Hypothesis Testing

● Pairplot for Visual Exploration: [Link](data)

● Heatmap for Correlation Analysis: [Link]([Link](),
annot=True)

By: Waleed Mousa

39. Working with Categorical Data

● Chi-Squared Test of Independence:

[Link].chi2_contingency(observed)
● Fisher's Exact Test: [Link].fisher_exact(table)

40. Tests for Proportions

● Proportions Z-Test:
[Link].proportions_ztest(count, nobs)
● Test for Equality of Proportions:
[Link].test_proportions_2indep(success1,
nobs1, success2, nobs2)

41. Regression and ANCOVA

● Simple Linear Regression Test: [Link](y, X).fit()

● ANCOVA: [Link].anova_lm(ols_model, typ=2)

42. Tests for Model Fit

● Log-Likelihood Ratio Test: [Link].anova_lm(model1,

model2, test='LRT')
● AIC and BIC Comparison: [Link], [Link]

43. Spatial and Temporal Data Analysis

● Mantel Test for Spatial Correlation:

[Link](x, y)
● Durbin-Watson Test for Autocorrelation:
[Link].durbin_watson(residuals)

44. Working with Survival Data

● Log-Rank Test for Survival Data:

[Link].logrank_test(event_times_A, event_times_B)
● Cox Proportional Hazards Model: [Link]().fit(df,
duration_col, event_col)

By: Waleed Mousa

45. Robust Statistical Methods

● M-estimators for Robust Regression:

[Link].robust_linear_model.RLM(y, X).fit()
● Robust Scale Estimators (e.g., MAD):
[Link](data)

46. Exploratory Factor Analysis

● Factor Analysis: factor_analyzer.FactorAnalyzer().fit(data)

● Rotated Factor Analysis:
factor_analyzer.FactorAnalyzer(rotation='varimax').fit(data)

47. Multivariate Statistical Tests

● Hotelling's T-squared Test: hotellings_t_square(sample1, sample2)

● MANOVA: [Link].from_formula('y1 +
y2 ~ x1 + x2', data).mv_test()

48. Graphical Representation of Test Results

● Violin Plot for Distribution Comparison:

[Link](x='group', y='data', data=df)
● ECDF Plot for Empirical Distribution:
[Link].empirical_distribution.ECDF(data)
● Scatter Plot for Association: [Link](x, y)

49. Data Transformation for Hypothesis Testing

● Box-Cox Transformation: [Link](data)

● Yeo-Johnson Transformation: [Link](data)

50. Advanced Non-parametric Methods

● Mann-Kendall Trend Test: [Link](x, y)

● Sen's Slope for Trend Analysis: mk.original_test(data)

By: Waleed Mousa

51. Hypothesis Testing in Time Series Analysis

● Granger Causality Test:

[Link](data, maxlag)
● ADF Test (Augmented Dickey-Fuller):
[Link](series)

52. Spatial Statistics

● Moran's I for Spatial Autocorrelation: [Link](data, w)

● Geographically Weighted Regression (GWR): [Link](y, X,
bw).fit()

53. Dealing with High Dimensionality

● Principal Component Analysis (PCA):

[Link](n_components).fit(data)
● Factor Analysis for Dimension Reduction:
[Link](n_components).fit(data)

54. Working with Complex Datasets

● Handling Missing Data with Multiple Imputation:

[Link]().fit_transform(data)
● Random Forest for Feature Importance:
[Link]().fit(X, y)

55. Interpretation and Reporting

● Effect Size Calculation (Cohen's d): compute_cohens_d(group1,

group2)
● Confidence Interval Calculation:
[Link].proportion_confint(count, nobs,
alpha=0.05)

56. Custom Hypothesis Tests

● Writing Custom Test Functions: def custom_test(data): # Implement

your test logic

By: Waleed Mousa

57. Simulation for Hypothesis Testing

● Simulating Data for Power Analysis: [Link](loc,

scale, size)
● Monte Carlo Simulation for P-values:
perform_monte_carlo_simulation(data, num_simulations)

58. Advanced Visualization for Test Assumptions

● Q-Q Plot for Normality: [Link](data,

plot=[Link])
● Pair Plot for Multivariate Normality: [Link](data)

59. Post-hoc Analysis

● Post-hoc Analysis for ANOVA:

[Link].pairwise_tukeyhsd(endog, groups)

60. Handling Time-to-Event Data

● Kaplan-Meier Estimator:
[Link]().fit(durations, event_observed)

By: Waleed Mousa