In this chapter you'll use bootstrapping techniques to estimate a single parameter from a numerical distribution.

Welcome to the course!

Generate bootstrap distribution for median

Review percentile and standard error methods

Calculate bootstrap interval using both methods

Which method more appropriate: percentile or SE?

Doctor visits during pregnancy

Average number of doctor's visits

SD of number of doctor's visits

Re-centering a bootstrap distribution

Test for median price of 1 BR apartments in Manhattan

Conclude the hypothesis test on median

Test for average weight of babies

Bootstrapping for estimating a parameter

In this chapter you'll use Central Limit Theorem based techniques to estimate a single parameter from a numerical distribution. You will do this using the t-distribution.

t-distribution

When to t?

Probabilities under the t-distribution

Cutoffs under the t-distribution

Estimating a mean with a t-interval

Average commute time of Americans

Average number of hours worked

t-interval for paired data

t-interval at various levels

Understanding confidence intervals

Testing a mean with a t-test

Estimate the median difference in textbook prices

Test for a difference in median test scores

Interpret the p-value

Introducing the t-distribution

In this chapter you'll extend what you have learned so far to use both simulation and CLT based techniques for inference on the difference between two parameters from two independent numerical distributions.

Hypothesis testing for comparing two means

Evaluating the effectiveness of stem cell treatment

Evaluating the effectiveness of stem cell treatment (cont.)

Conclusion of the hypothesis test

Evaluating the relationship between smoking during pregnancy and birth weight

Bootstrap CI for difference in two means

Quantifying the relationship between smoking during pregnancy and birth weight

Median lengths of pregnancies for smoking and non-smoking mothers

Comparing means with a t-test

Hourly pay vs. citizenship status

Estimating the difference of two means using a t-interval

Inference for difference in two parameters

In this chapter you will use ANOVA (analysis of variance) to test for a difference in means across many groups.

Punteggio del vocabolario: variazione tra classi sociali (autoidentificate)

EDA: punteggio di vocabolario vs. classe sociale

Confrontare molte medie, visivamente

ANOVA

ANOVA per punteggio di vocabolario vs. classe sociale (auto-dichiarata)

Condizioni per l'ANOVA

Verificare la condizione di normalità

Verificare la condizione di varianza costante

Test post-hoc

Calcola alpha*

Confronta le medie a coppie

Congratulazioni!

Comparing many means

Chp1-vid1-boot-dist-noaxes-parantheses

Chp1-vid1-bootsamp-bootpop.001

Chp1-vid1-manhattan-rents

Chp1-vid2-boot-dist-withaxes

Chp1-vid2-perc-method.001

Chp1-vid2-perc-method.002

Chp1-vid3-boot-test.001

Chp3-vid3-hrly-rate-citizen-smaller

Chp3-vid3-hrly-rate-citizen

Chp4-vid1-class-bar

Chp4-vid1-wodrsum-hist

Gss moredays

GSS data

Manhattan rent data

Runners.001

Tdistcomparetonormaldist

In questo corso imparerai a usare tecniche statistiche per trarre inferenze e fare stime a partire da dati numerici. Useremo due approcci per affrontare questi compiti comuni. Il primo sfrutta bootstrapping e permutazioni per creare test e intervalli di confidenza basati su ri-campionamenti. Il secondo utilizza risultati teorici e la distribuzione t per ottenere lo stesso obiettivo. Imparerai come (e quando) eseguire un t-test, costruire un intervallo di confidenza e fare un'ANOVA!

Foundations of Inference in R

Impara tecniche come il bootstrapping e la distribuzione t per inferenze statistiche su dati numerici.

Inferenza per dati numerici in R

In questo corso imparerai le tecniche per fare inferenze statistiche sui dati numerici.

Inferenza statistica in R

Verificare la condizione di normalità

Inferenza per dati numerici in R

Esercizio pratico interattivo