Variances that require investigation are typically those that are deemed significant or material. These variances can be identified through various methods, such as statistical analysis or expert judgment. When evaluating variances, it is essential to consider factors such as their magnitude, frequency, and potential implications. Variances that are large, occur frequently, or have the potential to impact financial performance or compliance should be prioritized for investigation. Additionally, variances related to key performance indicators, high-risk areas, or areas where fraud or errors are suspected should also be subject to investigation.
Process Variability
Process variability is the extent to which a process produces different results. It can be caused by a number of factors, including:
- Variations in the raw materials
- Variations in the equipment
- Variations in the process itself
- Variations in the environment
Process variability can have a significant impact on the quality of the final product. It can also lead to increased costs and reduced efficiency.
| Type of Variance | Description | How to Investigate |
|---|---|---|
| Assignable Variance | Caused by specific, identifiable factors. | Use statistical process control (SPC) charts to identify the assignable causes of variation. |
| Unaassignable Variance | Caused by unknown or random factors. | Use statistical analysis to estimate the amount of unassignable variance. |
Outliers
Outliers are data points that are significantly different from the rest of the data. They can be caused by a variety of factors, such as measurement errors, data entry errors, or unusual events. Outliers can be a problem for data analysis because they can skew the results of statistical tests and make it difficult to draw meaningful conclusions from the data.
There are a number of ways to investigate outliers. One common approach is to use a box plot. A box plot shows the distribution of the data and identifies any outliers that are outside of the normal range. Another approach is to use a Grubbs’ test to determine if an outlier is statistically significant.
If an outlier is found to be statistically significant, it is important to investigate the cause of the outlier. This may involve checking the data for errors, looking for unusual events that may have caused the outlier, or consulting with an expert in the field.
| Method | Description |
|---|---|
| Box plot | Shows the distribution of the data and identifies any outliers. |
| Grubbs’ test | Determines if an outlier is statistically significant. |
- Check the data for errors.
- Look for unusual events that may have caused the outlier.
- Consult with an expert in the field.
Control Limits
Control limits are set on either side of the central line, at a distance of three standard deviations from the central line. These limits are used to signal when the process is out of control, in other words, when the variance in the process is too large. Any point that falls outside the control limits should be investigated.
There are two types of control limits:
- Upper control limit (UCL)
- Lower control limit (LCL)
The UCL is the upper limit for the process, and the LCL is the lower limit. Any point that falls above the UCL or below the LCL is considered to be out of control.
Control limits are calculated using the following formulas:
UCL = central line + 3 * standard deviation
LCL = central line – 3 * standard deviation
For example, if the central line is 100 and the standard deviation is 10, then the UCL would be 130 and the LCL would be 70.
Type of Variance Description Common Cause Variance This type of variance is caused by random factors that are inherent in the process. It is also known as background noise or inherent variability Special Cause Variance This type of variance is caused by specific, identifiable factors that are not inherent in the process. It is also known as assignable variation Thanks a bunch for hanging in there with me while I rambled on about variances. I hope you found some useful nuggets of wisdom in this little article. If you’re still curious about variances and their many quirks, be sure to swing by again sometime. I’ll be here, ready to dive deeper into the wonderful world of statistics. Until then, keep your eyes peeled for those pesky variances and don’t hesitate to investigate them when they pop up!