Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a va

Measuring performance results is a great thing to do, but understanding the causes of those results is at least as worthwhile. Understanding causes means you have information about how to exercise more influence (or control) over those results. If you want your results to improve, you've got to change the right things about the process or activity or function that produces those results.

Understanding the real causes of performance results means taking a more rigorous approach than knee-jerk reacting to hearsay, opinion or gut feel. Here are some basic techniques to help you navigate through the stages of cause analysis:

* find the likely causes, and measure the incidence of each

* assess the nature and size of the cause's impact

* check for interaction with other causal factors

Technique #1: flow charting

It's impossible to do any kind of serious cause analysis unless you can actually trawl through all the factors that have some kind of potential impact on your performance result, and sift out those factors that have the most dominant impact. Flow charting the process or activity or function whose results you are measuring, is a great way to systematically trawl through all the potential causes of those results. There is software available for flow charting, but hand-drawn charts are quick and easy.

Technique #2: cause-effect diagrams

After flow charting your process and identifying what can sometimes be dozens of potential causes, you can have long lists that contain duplicates and related causes. Cause-effect diagrams (or fishbone diagrams) are a great way to collate and organise potential causes as you identify them, clustering related causes together so you can more clearly see the themes, and more easily discuss the most likely causes. There is software available for cause-effect diagrams, but again hand-drawn diagrams can do the job well enough.

Technique #3: Pareto charts

When you then go and count or measure how often or how much each likely cause is associated with your results, Pareto charts can then help you rank the causes and highlight those that have the biggest impact. You're now getting the stage where you have between 2 and 5 (roughly) causal factors you may wish to learn even more about. In Microsoft Excel, just use a vertical bar chart on your data, after sorting it from biggest to smallest.

Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a val

the stages of cause analysis:

* find the likely causes, and measure the incidence of each

* assess the nature and size of the cause's impact

* check for interaction with other causal factors

Technique #1: flow charting

It's impossible to do any kind of serious cause analysis unless you can actually trawl through all the factors that have some kind of potential impact on your performance result, and sift out those factors that have the most dominant impact. Flow charting the process or activity or function whose results you are measuring, is a great way to systematically trawl through all the potential causes of those results. There is software available for flow charting, but hand-drawn charts are quick and easy.

Technique #2: cause-effect diagrams

After flow charting your process and identifying what can sometimes be dozens of potential causes, you can have long lists that contain duplicates and related causes. Cause-effect diagrams (or fishbone diagrams) are a great way to collate and organise potential causes as you identify them, clustering related causes together so you can more clearly see the themes, and more easily discuss the most likely causes. There is software available for cause-effect diagrams, but again hand-drawn diagrams can do the job well enough.

Technique #3: Pareto charts

When you then go and count or measure how often or how much each likely cause is associated with your results, Pareto charts can then help you rank the causes and highlight those that have the biggest impact. You're now getting the stage where you have between 2 and 5 (roughly) causal factors you may wish to learn even more about. In Microsoft Excel, just use a vertical bar chart on your data, after sorting it from biggest to smallest.

Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a va

matically trawl through all the potential causes of those results. There is software available for flow charting, but hand-drawn charts are quick and easy.

Technique #2: cause-effect diagrams

After flow charting your process and identifying what can sometimes be dozens of potential causes, you can have long lists that contain duplicates and related causes. Cause-effect diagrams (or fishbone diagrams) are a great way to collate and organise potential causes as you identify them, clustering related causes together so you can more clearly see the themes, and more easily discuss the most likely causes. There is software available for cause-effect diagrams, but again hand-drawn diagrams can do the job well enough.

Technique #3: Pareto charts

When you then go and count or measure how often or how much each likely cause is associated with your results, Pareto charts can then help you rank the causes and highlight those that have the biggest impact. You're now getting the stage where you have between 2 and 5 (roughly) causal factors you may wish to learn even more about. In Microsoft Excel, just use a vertical bar chart on your data, after sorting it from biggest to smallest.

Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a va

kely causes. There is software available for cause-effect diagrams, but again hand-drawn diagrams can do the job well enough.

Technique #3: Pareto charts

When you then go and count or measure how often or how much each likely cause is associated with your results, Pareto charts can then help you rank the causes and highlight those that have the biggest impact. You're now getting the stage where you have between 2 and 5 (roughly) causal factors you may wish to learn even more about. In Microsoft Excel, just use a vertical bar chart on your data, after sorting it from biggest to smallest.

Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a va

Technique #4: scatter plots

When you arrived at the few causes that have the biggest impact on your performance result, it can be useful to know just how big that impact is. Scatter plots are an easy and visual way to explore when the cause variable changes, how much and in which direction the performance result changes. Scatter plots are one of the charts available in Microsoft Excel.

Technique #5: correlation coefficients

To get more a quantitative measure of the impact of a causal factor on your result, you can calculate a correlation coefficient which will give you a value between 0 and 1 indicating the strength of the relationship between your causal factor and the result. A postive value means that an increase in your causal factor will likely lead to an increase in your result, and a negative value means that an increase in your causal factor will likely lead to a decrease in your result. In Microsoft Excel, use the CORREL function to calculate your correlation coefficient.

Technique #6: regression analysis

Regression analysis goes a step further, and builds a mathematical model you could use to predict a result based on a change in your causal factor. Knowing this can help you set achievable targets for improvement, and estimate realistically what resources you're really going to need to get that improvement. In Microsoft Excel, create your scatter plot between your causal factor and performance result, then add a trend line, with the options of showing the equation and R-squared value on the chart (the R-squared value is a measure the reliability of the equation).

There are certainly more statistical techniques that can help you with cause analysis (such as multi-variate regression, experimental design and analysis of variance or ANOVA), but those provided above will still bring some valuable rigour to your performance improvement efforts.