Training Sessions in Statistics

×

Warning

The RokSprocket Module needs the RokSprocket Component enabled.

Statistics for Engineers This session discusses key concepts in statistics. Classical and more recent exploratory data analysis techniques to efficiently summarise data and to detect outliers are presented. Statistical testing and decision-making in the presence of variation are also discussed.

Learning Objectives

Learning Objectives:

Upon completion of this module, participants will be able :

• To distinguish descriptive from inferential statistics
• To appreciate the value of exploratory data analysis methods in preliminary data analysis & in experimental design construction
• To explore, characterise and identify problems and trends in data using plots
• To understand the concepts of hypothesis testing, confidence intervals, risk & power
• To understand the importance of sample size calculations and the required input parameters to estimate a sample size
• To perform simple data more quickly and more accurately
• To interpret results reliably and confidently

Target Audience

Target Audience:

This session is aimed at engineers who must make decisions based on that data.

Prerequisite

Prerequisite:

This session introduces the important ideas in statistics and data analysis. It assumes that participants either have no previous knowledge of statistics or that they have not used statistics for a long time.

Course Outline

Course Outline:
• Why Do we Need Statistics?
• Descriptive or Exploratory Data Analysis "EDA"
• Overview and Goals
• Importance of Identifying the Type and Role of Variables in Studies
• Visualizing and Summarizing Data: The Concept of a Distribution
• Graphical Tools: histogram, Box-plot
• Numerical Tools: mean, median, standard deviation, standard error, etc.
• Exploring the relationship between two (2) variables
• Frequency tables for categorical variables
• Pearson's correlation coefficient for continuous variables
• Plots: Scatter plots, box-plots, etc.
• Statistical inference or hypothesis testing
• Overview: What is statistical inference?
• Statistical Inference with Hypothesis Testing:
• Null and alternative hypotheses
• One-tailed vs. two-tailed tests
• Test statistics: t-test, F-tests, etc.
• Observed significance level or "p-value"
• Statistical significance and decision rules
• Risk involved in hypothesis testing
• Risks or type I and II errors
• Confidence level of a test
• Power of test
• The importance of sample size calculations and the required input parameters to estimate a sample size
• Statistical inference with confidence Intervals: interpretation and usage
• Statistical Inference for a Single Sample or Group: Hypothesis Testing vs. Confidence Interval Approach
• Summary

Practical Info

Practical Info:

Recommended Course Duration: 1 day

Course materials :

• Course notes on the statistical techniques
• Sample datasets
• Related Sessions

• Efficient experiments must be large enough to detect meaningful scientific differences and maximize the use of available resources. Learn about sample size and power calculations.

• Variation is present in every experiment. Learn about DOE techniques to control variation, and to maximise data quality. Commonly used experimental designs are discussed as well as the statistical data analysis tools.

• Learn about key biostatistical concepts and efficient tools for summarising and plotting data as well as outlier detection. Demystify the statistical testing approach used to make decision in the presence of uncertainty: p-values, power, and so on.

• This module covers key biostatistical concepts required to review and interpret findings published in the medical literature. Selected scientific publications are reviewed, discussed and criticised in terms of bias, uncertainty and scope.