Statistical Data Analysis for "Experts"

Workshop Overview

Data scientists often encounter non-conforming data when analyzing multiple groups and when tracking the same group over various periods. It's also expected to have insufficient data for analysis. In such cases, data analysis techniques should focus on statistical science related to "dependent" samples and "non-parametric" tests as alternatives to parametric ones (Data Analysis for Professionals). Determining the sample size is a common question when designing an analytical project. Should we consider the Big Data solution where all the data is included? This course will address these questions, discussing their advantages and disadvantages.

Learning Outcomes

• Compare dependent samples with independent ones.
• Compare non-parametric analytics with parametric ones.
• Face the two types of errors in statistical tests with « power analysis».
• Calculate sample size with scientific methods.
• Understand factors that influence sample size.
• Explore the different errors while defining sample size.
• Manipulate multiple software solutions with the correct interpretation of results.

Duration 3 days

Day 1:

- Dependent Samples Analytics

Day 2:

- Non-Parametric Analytics

Day 3:

- Power Analysis

What will it be about?

- Colored PPT documents / Videos

- Statistical tests: Man Whitney, Correlation Rank test,  t-paired, ...

- Proprietary tools solutions

- Two Way ANOVA

- Profiling Techniques

- The All-In-One chart

- The one and unique P-Value

‍