I. Basic concepts.
- Hypothesis testing (I). Null hypothesis. Alternative hypothesis.
It all spins around the null hypothesis. - Hypothesis testing (II). Alpha. Beta. Power. Sample size.
Size and power. - Hypothesis testing. Fragility index.
The fragility of the EmPress. - Confidence intervals.
Always look for an interval, but of confidence. - Confidence intervals. Prediction intervals. Tolerance intervals.
The poor relations. - Hypothesis contrast. Statistical significance. Tipo 1 error. Tipo 2 error.
To p or not to p…is that the question? - Statistical significance. Clinical relevance.
Even non-significant ps have a little soul. - The meaning of p-value.
Worshipped, but misunderstood. - Fragility, stability and clinical relevance.
Take off your blinders. - Confidence intervals. Statistical significance. Clinical relevance.
Life is not rosy. - Inverse fallacy. Conditionals’ transposition fallacy.
The fallacy of small p. - Sample size.
Size does matter. - Sample size in surveys.
With very little we fine-tune a lot. - Sample size and precision.
Having a large n, who needs a samll p?. - Factor that condition the sample size.
A strenuous effort, though unnecessary. - Sample size for estimating a proportion.
If you doubt, better in the middle. - Sample size for estimating a mean.
A measure with a pedigree. - Sample size for the comparison of two means.
A pair of means. - Sample size calculation in survival studies.
By your actions they will judge you. - Study of qualitative variable dispersion.
Like a forgotten clock. - Location parameters. Arithmetic mean. Median. Mode. Geometric mean. Harmonic mean.
Virtue is the happy medium between two extremes, but… - Choosing between mean and median.
Meat or fish? - Robust location parameters.
A very robust family. - Harmonic mean.
A trickly riddle. - Geometric mean.
Counting bugs, - Robust scale parameters.
Do not be misled by the outliers. - Biased and unbiased estimators.
Why spare one? - Dispersion parameters. Variance. Standard deviation.
Not all deviations are perverse. - Dispersion parameters. Minimum. Maximum. Quartile. Decile. Percentile. Interquartile range.
The most wished statistical for a mother. - Standardization. z score. Frequiency distribution. Probability density curves. Histograms.
Give me a bar and I’ll move the earth. - Simmetry measures. Mean deviation. Distribution bias. Kurtosis.
Playing with powers. - Degrees of freedom. Sample size. Power.
Freedom in degrees. - Regression to the mean.
The curse. - Independency of variables. Paired data.
Independence matters. - Transforming data. Logarithm transformation. Inverse transformation.
Cheating Gauss. - Graphic representation of qualitative variable. Pie chart. Bar chart. Pareto’s graph.
Columns, sectors, and an illustrious italian. - Histogram. Bar graph.
As and egg to a chestnut. - Box plot.
A box with whiskers. - Quantile graph. q-q plot.
Some comparisons are not odious. - The pie chart.
Do not eat too many pies. - Outliers.
Black sheep. - Frequentist vs Bayesian’s statistics.
Rioja vs Ribera?
II. Probability.
- Normal distribution. Standardization.
The most famous of bells. - Probability distributions. Stdent’s t distribution. Chi-squared distribution. Snédecor’s F distribution.
The big family. - Student`s t distribution.
The brewer. - Discrete probability distributions.
The sleepless naturalist. - Binomial probability (I). Binomial distribution. Bernouilli’s events.
The cooker and his cake. - Binomial probability (II). Survey errors. Lie. Obfuscation factor.
Lie to me. - Binomial probability (III). Calculating number of successes.
Do not gamble. - Binomial probability (IV). The law of small numbers.
The oodities of small towns. - Yates’ continuity correction.
Horror vacui. - Conditioned probability. Bayes’ rule.
The stigma of guilt. - Conditioned probability. Bayes’ theorem.
A case of misleadind probability. - Conditioned probability (II). Bayes’ rule.
Another about coins. - Combinatory and probability.
The deception of intuition. - Probability value when numerator equals zero.
When nothing bad happens, is everything okay?
III. Statistic tests.
- Hypothesis testing. One-tailed test. Two-tailed test.
The tails of p. - Statistical significance. Power. Type I error. Type II error.
The false coin. - Multiple comparisons. Type I error.
Even a blind donkey… - Statistical test by variable type.
Brown or blond, all bald. - Contingency tables. Residuals. Pearson’s residuals. Standardized residuals. Adjusted residuals. Chi-square.
Residuals management. - Chi-square test (I). Goodness of fit.
Solomonic decisions. - Chi-square test (II). Homogeneity test.
Counting sheep. - Chi-square test (III). Dependent variables.
Studying or working?. - Chi-squared-test (IV). Calulating expected values.
The why and wherefores. - Comparison of two proportions. Chki-square.
All road lead to Rome. - Fishjer’s exact test.
A history of tea and numbers. - Confidence interval for a mean.
The error of confidence. - Concordance. kappa’s interobserver concordance coeficient.
A good agreement?. - Concordance. Bland-Altaman’s method.
Another stone with which nor to trip over. - Stratification. Mantel-Haenszel’s test. Confusion. Adjusted measures of association.
Dividing to conquer. - Choosing the statistical test.
Pairing. - Normality análisys. Contrast and graphical methods.
A picture is worth a thousand words. - Normality analysis. Analytic methods.
Moments. - Goodness of fit to test a normal. Shapiro-Wilk’s test. Kolmogorov-Smirnov’s test.
Behind the scenes. - Comparison of two means. Student’s t test.
The same old story. - Analysis of variance of one factor. ANOVA.
More than two is a crowd. - Bonferroni’s correction.
When the zeroes of p are important. - False discovery rate. False positive. Type 1 error.
Percy Fawcett and the Lost City. - Linear correlation. Pearson’s lineal correlation coefficient. Covariance. Independence.
An open relationship. - Missing data. Imputation.
…there are not all those who are. - Management of non-normal data.
Everything is not normal. - Non-parametric tests. Sign test.
At a rough guess. - Survival analysis. Survival tables. Censored data.
Censorship. - Correlation. Correlation’s coeficients.
Its bark is worst than its bite. - Simple regression. Lineal regression. Logistic regression. Cox regression.
A simple relationship. - The least squares method.
The shortest distance. - Simple lineal regression. Model diagnostics.
Don’t leave things half done. - Collinearity. Multiple lineal regression.
Three feet of a cat. - Quality of a linear regression model.
The mistery of the different colored socks. - Regularization regression techniques. Ridge regression. Lasso regression. Elastics net regression.
An epic dance. - Discontinuity regression. Multiple lineal regression.
In search of causality. - Analysis of logistic regression models.
The paradox of the air. - Resampling techniques. Bootstrapping.
An impossible task. - Propensity score.
You can’t make a silk purse… - Meta-analysis (I). Heterogeneity. Metaregression.
Apples and pears. - Meta-analysis (II). Heterogeneity. Cochran’s Q. I2. H2.
The polisemy of Q. - Meta-analysis (III). Publication bias. Forest plot.
Achilles and Effects Forest. - Fisher’s z. Meta-analysis with small sample sizes.
The Palace of Probabilities. - Trim and fill method. Publication bias. Meta-analysis.
A seance. - Vote counting in systematic reviews.
The failure of democracy. - Evidence of effect in meta-analysis. The curve of p.
The vanishing hitchhiker. - Effect size based in differences between means.
I am Spartacus. - Multivariate analysis techniques.
Exoteric or esoretric? - Incorrect uses of statistics.
The cheaters detector. - Artificial intelligence. Machine learning.
Hasta la vista, baby. - Artificial neural networks.
Science behind the magic. - Decision trees.
The tree and the labyrinth.