Exploratory Data Analysis


Overview/Description
Target Audience
Expected Duration
Lesson Objectives
Course Number



Overview/Description
The Analyze phase of Six Sigma's® DMAIC (Define, Measure, Analyze, Improve, and Control) roadmap includes what is traditionally referred to as "crunching the numbers." After you have accurately defined the problem and measured the correct data in earlier phases, the Analyze phase looks at that data from all angles in an effort to precisely determine the relationships among variables. Making the data useful is the job of the Analyze phase. By organizing, quantifying, visualizing, and testing the relationships between variables in a process, the Analyze phase narrows the focus to the few key causes of error or inefficiency. In the Improve phase that follows, Six Sigma teams then have the formulas, tested hypotheses, and target areas for the changes that are needed--a statistical understanding of the problem and likely solutions. This course, Exploratory Data Analysis, discusses the initial methods for understanding the collected data, using various tools and techniques that present the data in ways that reveal both simple and complex relationships among the variables. Using statistical analysis techniques--some unique to Six Sigma, but most standard to all statistical evaluation--relationships are displayed and correlations are established. Using regression analysis, formulas can be developed to model relationships, allowing prediction and estimation. This "number crunching" can later be used to formulate and mathematically test hypotheses before actually implementing solutions. Six Sigma is a registered Trademark of Motorola Corporation, and all rights, title and interest in Six Sigma belongs to Motorola.

Target Audience
Candidates for Black Belt certification; managers/executives overseeing personnel involved in the implementation of Six Sigma in their organization; consultants involved in implementing a Six Sigma proposal; and organizations implementing a Six Sigma project

Expected Duration (hours)
2.5

Lesson Objectives

Multivariate Studies

  • recognize the benefits of visualization of the relationships between key process input and output variables.
  • match data visualization techniques with corresponding images.
  • match data visualization techniques with examples of the types of data to which they are best suited.
  • match types of multivariate analysis to descriptions.
  • match the families of variation shown by multi-vari charts with examples.
  • Simple Linear Correlation

  • identify the benefits of understanding simple linear correlation.
  • calculate the correlation coefficient for a given set of data.
  • match statements about causation and correlation to examples of their respective characteristics.
  • Modeling Relationships between Variables

  • identify the benefits of developing a regression equation.
  • perform simple least-squares regression calculations for a given scenario.
  • match the elements of a multiple least-squares linear regression equation to their descriptions.
  • match the types of residual analysis to corresponding examples.
  • match the common methods for analyzing uncertainty in a model to their correct descriptions.
  • Course Number:
    OPER0251