Skip to main content

Histogram in Quality Control

Histogram in Quality Control

What is histogram?

Histogram is a spc technique. It is graphical tool that represent the data values with the help of vertical rectangular bar. Height of the bars is corresponding to the frequency of the data values.
  • A frequency distribution shows different value in data occurs
  • A Histogram most commonly used graph to show frequency distribution.
  • It looks very much like a bar chart, but there is important difference between them.

Why does the histogram use as a QC tool?

Showcases the large amount of data using vertical bars, It is help in analysis the properties of data in statistical process control such as.
Distribution of the data
Spread of the data
Variation in the process
Skewness of the data
Help summaries data from process that has been collected over a period of time.
Help graphically represent data frequency distribution in bar chart.

What does it to do?

  • Reveals centering, variable and underlying distribution of data
  • It is help to the process capable to meeting customer requirement

Difference in histogram and bar chart

Histogram is similar to bar charts but there are two main differences.
  • There are no gaps between the bars in histogram.
  • The area of each bar is proportional to the frequency that is represented. Hence all area is proportional to the all frequency.

When to use a histogram

  • When the data are numerical.
  • When you want see data shape distribution, especially when determining whether the process output is distributed approximately normal.
  • When analyse whether a process can meet the customer`s requirement.
  • When analyse what the output from a supplier`s process looks like.
  • When process parameter change from one time period to another time period.
  • When determine whether the outputs of two or more process are different.
  • When you wish to communicate the distribution of data quickly and easily to other.
Histogram Construction 
  • Collect at least from 30 to 100 consecutive data point from a process.
  • Use the histogram work sheet to set up the histogram. It will help you determine the number of bar, the range of number that go into each bar and the labels for the bar edges. After calculating W in step 2 of the work sheet, use your judgment to adjust it to a convenient number. For Example, you might decimal to round 0.9 to an even 1.0 the value of W must not have more decimal place than the numbers you will be graphing.
  • Draw x- and y-axis on graph paper. Mark and label the y-axis for counting of data values. Mark and label the x-axis with the help of L values from worksheet. Spaces between these numbers will represent the histogram bars. Do not permit for spaces between these bars.
  • For each data point, mark off one count on the top of suitable bar with an x or by shading that portion of the bar.
How does a histogram help to the analysis the data?

Symmetrical distribution or Normal distribution 

Symmetrical distribution or Normal distribution

A common pattern is the bell-shaped curve knows as the “Normal distribution.” In a normal distribution, points are as likely to occur on one side of the average as on the other. Be alert that the other distributions looks like similar to the normal distribution. Statically calculations must be used to prove a normal distribution.

Skewed distribution 
  • The skewed distribution is asymmetrical because a natural limit prevent outcome on one side.
  • The distribution peak`s off center toward the limit and a tail stretches away from it. For example, a distribution of analysis of very pure products would be skewed, because the products cannot be more than 100 percent pure.

Other lesson of natural limits are holes that cannot be miniature than the diameter of the drill bit or call- keeping times that cannot be less than zero. These distributions are called right skewed when its tail are right position and when tail are left position it is called left skewed. It is depending on the direction of the tail.
Skewed rightSkewed left
                                   Skewed  Right                                        Skewed  Left

Doubled peaked or bi-modal distribution 

The bi-modal distribution shape looks like the back of a two humped camel. The outcomes of two processes with different distribution are combined in one set of data. 
Doubled peaked or Bi-modal distribution
                                                         Doubled peaked or bi-modal distribution

For example, a distribution of quality data from a two shift analysis might be bi-modal, if each shift produces a different distribution of results. Stratification often reveals this problem. 

Plateau Distribution 

Plateau distribution
                                                              Plateau Distribution

The plateau might be called “multi-modal distribution.” Several process with normal distribution are combined because there are many peaks close together, the top of the distribution resembles a plateau.

Edge peak distribution
Edge peak distribution

                                                      Edge peak distribution

The edge peak distribution shape looks like the normal distribution besides that it has a large peak at one tail.
Usually this is caused by faulty construction of histogram, with data lumped together into a group labelled “greater than...”


Popular posts from this blog

To determine loss on drying (LOD)

To determine loss on drying (LOD) Definition:   Loss on drying compares the weight of product sample before and after drying. The result is the percentage of moisture in a product. % Moisture = (Begging weight-Ending weight) X 100                         -----------------------------------------------                                    Begging weight  Loss on drying determine by two method which is given below: Method (1):   Weigh accurately a dry empty glass Petri dish. Put the sample (about 0.5 to 5 gm as par requirement, Its mean if sample having less wet take 5 g weight. If sample having more wet than take 0.5 g weight ) in dish and weigh. Note down the reading. Distribute the sample in Petri dish by gentle shaking. Place the loaded dish (without cover) in the drying chamber for two hours. Maintain the oven temperature 105 +/- 5 0 C.  After drying is completed, open the drying chamber, cover the dish and allow it to cool at room temp.  Keep it in the desiccator for 15

Analytical Method Validation

Analytical Method Validation   The validation is performance, demonstration of any procedure, process, equipment, material, activity or any system performance as expected under given set of condition. In the most basis form, validation is proving that the performance is as intended. When extended to an analytical procedure,depending upon the application, it means that the method works reproducibaly, when carry out by same or different persons, in same or different laboratories, using different reagents, different equipment, etc. The analyst normally tends to assume that if an analytical method or process performs reproducibility in an limited number of experiment, it will continue doing so forever, which is too optimistic to assume. The Benefits:   The biggest   advantage  of method validation is that it builds a degree of confidence, not only for the developer but also to the user. Although the validation exercise may appears costly and time consuming. It is results in expens

Handling of Laboratory Incidents

Handling of Laboratory Incidents What is Incident: Any unplanned or uncontrolled event in the form of non-compliance to the designed systems or procedures at any stage of testing, and storage of drug product due to system failure or equipment breakdown or manual error. A laboratory Incident is an event in the laboratory that occurs for two primary reasons either due to analyst error or instrument error. Few example of laboratory incidents: Use of expired reagents/working standard/reference standard System suitability fail. Improper peak shape or peak split. Relative retention time shift. Communication failure between software and hardware during analysis. Power failure during analysis.  Temperature of laboratory found out of acceptance criteria. Out of calibration instrument used for analysis. To be broken or damage part of instrument during analysis. Responsibilities of the Personnel in case of incidents: Analyst:   Reporting of an incident to Head Q