Statistical Methods for Quality Improvement

Statistical methods for quality improvement involves the use of collected data and quality standards to find new ways to improve products and services. These methods characteristically involve attempts to infer the properties of a large collection of data.

OriginPro provides the following key tools for quality improvement:

Statistical Process Control (SPC)
Design of Experiments (DOE)

The above tools are available as free Apps that can be added to OriginPro.

Statistical Process Control (SPC)

The Statistical Process Control (SPC) App, for OriginPro brings together five key SPC features:

Control Charts

The SPC App offers a wide variety of configurable control charts:

For Subgroups

Xbar-R
Xbar-S
I-MR-R/S
Xbar
R
S
Zone

For Individuals

I-MR
Z-MR
Individuals
Moving Range

Attributes Charts

P/U Chart Diagnostic
P/U Chart
Laney P' / U'
NP
C

Time-Weighted Charts

Moving Average
EWMA
CUSUM

Process Overview

The Process Overview tool in the SPC app offers reports with charts and capability statistics. It helps you to identify problem areas and gain insights into your data:

Xbar / Individuals charts
Probability plot
Capability histogram
Capability plot
Capability statistics

Capability Analysis

Capability analysis is used to detect whether a process meets specifications. The SPC app provides 5 types of capability analysis tools for different data types:

Normal
Between/Within
Non-normal
Binomial
Poisson

Data Transformation

The SPC app provides 3 transformation functions to make data approximately normal:

Box-Cox Transform
Johnson Transform
Yeo-Johnson Transform

The reports of Process Overview generates a combination report with multiple charts and statistics. It enables a user to assess the quality or capability of a process in a glance

The reports of Capability Analysis include control charts and capability reports with Cp, Cpk, Pp and Ppk indices, to help the user to access whether the process meet the specifications

Identify Data Distribution

The SPC App provides 14 distributions and 3 transformation functions to help to identify your data distribution:

Normal
Logistic
Lognormal / 3-Parameter-Lognormal
Log-logistic
Weibull / 3-Parameter Weibull
Gamma
Rayleigh
Smallest Extreme Value
Largest Extreme Value
Exponential / 2-Parameter Exponential
Gaussian Mixture
Folded Normal
Box-Cox Transformation
Johnson Transformation
Yeo-Johnson Transformation

Origin provides 14 distributions and 3 transformation functions to help to identify your data distribution

Start from the decision map, choose a proper tool basing on your data type and generate powerful reports.

Design of Experiments (DOE)

The Design of Experiments (DOE) App for OriginPro enables researchers to walk through each step of the design of experiments process.

Power and sample size

Determine the minimum sample size required to achieve a certain level of power for detecting a specified effect size
Find the right balance that ensures robust and valid experimental outcomes

Design

Factorial designs, including 2-level factorial, Plackett-Burman and general full factorial
Response surface design, including central composite and Box-Behnken

Analysis

Design evaluation tools including coefficients, standard error and p-value.
Build modes with linear+interaction+square, or customize terms in regression model.
Statistics table to examine the goodness-of-fit, and analysis summary with ANOVA
Fits and residual diagnostics
Find Y from X to extract fitted data.

Graphics

Fitted Plot
Residual Plot
Effects Plot
Main Effects Plot
2-way Interaction Plots
Contour Plot
Surface Plot

Optimize

Search for optimal responses based on the goal:minimize, maximize and target.
Constraint factors to a region or hold at a fixed value.
Display the optimization plot, in which you can tune factor settings to improve the solution.