Determining which of the classic seven quality tools are in the quality practitioner’s RCA tool kit

Abstract

The purpose of this study is to determine which of the classic seven quality tools are used by quality professionals during a Root Cause Analysis (RCA). The literature was assessed to determine which quality tools are considered to be part of the seven classic quality tools. The identified tools were then used in a survey of quality professionals in a manufacturing organization, who were asked to select the classic seven quality tools they use during an RCA. An email distribution list was used to send the survey to quality professionals at various locations around the world. The results were then analyzed statistically. The cause and effect diagram was the most commonly used quality tool during an RCA. This was distantly followed by the Pareto chart and then the control chart, followed by the Pareto chart, flow chart, histogram, scatter plot, and run chart. The classic seven quality tools have long been discussed in the literature; however, much of the literature consist of descriptions of their use and not an assessment of which tools are frequently used by quality professionals specifically for RCA. This study identified the seven main classic quality tools used by quality professionals during an RCA.

Keywords:

• Root cause analysis
• RCA
• problem solving
• failure analysis
• classic seven quality tools

PUBLIC INTEREST STATEMENT

The research described in this paper was conducted to identify which of the classic seven quality tools are used by quality professionals during an RCA (Root Cause Analysis). The classic seven quality tools are also known as the old seven and the seven QC (Quality Controls) tools, and they were brought together by Kaoru Ishikawa in Guide to Quality Control. Ishikawa’s original seven tools were the control chart, Pareto chart, histogram, check sheet, scatter plot, stratification, and cause and effect diagram, which is also known as a fishbone diagram, or an Ishikawa diagram. Over time, other authors have replaced tools with different tools, such as the flow chart and the run chart.

This research found that the most commonly used quality tool for RCA was the cause and effect diagram. Organizations training employees in RCA should ensure that the cause and effect diagram is included in the training.

1. Introduction

A Root Cause Analysis (RCA) is performed to identify the cause of problems and includes actions to establish a definition of the problem, analyze the problem, and find the cause of the problem, so that the problem can be prevented from occurring again (Lee et al., 2018). There are many reasons why an RCA may be needed. For example, Harjac et al. (2008) investigated corrosion of the walls of absorber towers and Cournoyer et al. (2013) investigated cut, torn, and pinched gloves in a safety glove box for a process that involved radioactive material. Alexa and Kiss (2016) describe an RCA to identify the root causes of damaged packaging and Darekar et al. (2013) explain the details of an investigation into the leakage of a vehicle fuel line.

Quality tools are needed when performing an RCA, and they are used for analyzing problems, generating ideas, and solving problems (Lee et al., 2018). Kaoru Ishikawa (1991) introduced the world to a collection of seven quality tools, which are now used by quality professionals. Ishikawa’s originals even quality tools were the cause and effect diagram, check sheet, control chart, histogram, Pareto chart, scatter plot, and stratification. However, other authors may list different quality tools (Borror, 2009; Okes, 2002).

The quality tool kit introduced by Ishikawa is now known by many names such as the classic seven, old seven, and simply the seven quality tools (Barsalou, 2017). Other names for the quality tools are the seven quality control tools, which is abbreviated as 7QC, and the basic seven quality tools (Soković et al., 2009).

This simple tool kit is applicable to a wide range of problems, and an advanced education or expert knowledge is not required to successfully apply the classic seven quality tools. For example, in the 1960s the quality guru Joseph Juran (2005) observed 18- to 23-year-old women from an assembly operation presenting the results of their quality improvements accomplished using the classic seven quality tools. According to Stamatis (2002 p. 57) “These tools can be easily learned and used by everyone in the organization; are very effective in achieving basic problem solving success; and are essential to any properly designed improvement strategy.”

This study seeks to identify which of the classic seven quality tools are used the most in industry for performing an RCA. To do so, a survey was sent to quality professionals. The survey identified the cause and effect diagram, control chart, and flow chart as the three most commonly used for RCA. This information can be useful for managers planning RCA-related training or implementing a problem-solving process.

2. Literature Review

The concept of the cause and effect diagram originated in 1943 when Dr. Karou Ishikawa was explaining to engineers how factors can be sorted in an organized way (Ishikawa, 1991). The cause and effect diagram is also known as an Ishikawa diagram or a fishbone diagram. A problem is listed at the front of the cause and effect diagram, and then category labels are placed on the main branches. These are often machines, methods, environment, people, and materials. However, the exact label names can vary. Sub-branches are then listed with potential causes (Tague, 2005). The potential causes in the major and minor branches are created through brainstorming (Metha, 2014).

An advantage of the cause and effect diagram is its simplicity. An inexperienced person can quickly learn to create a cause and effect diagram in a few minutes and this brief training works even better when a real-life example known to the trainee is used (Arp, 2020). However, a cause and effect diagram may need to go down multiple levels to identify root causes (Suárez-Barraza & Rodríguez-González, 2019). Once identified, hypotheses in a cause and effect diagram must be tested using new data (Kume, 1997).

Sharma et al. (2018) created a cause and effect diagram for a pitting problem. The main branches were material, measurement, environment, method, machine, and manpower. Under environment, they listed acidic/alkaline mist, temperature, and contaminants. For the material branch, they listed chaises surface, cleaning detergent, and deionized water ion component. The remaining branches each had from 1 to 2 possible causes listed. The cause and effect diagram was also used by Desai and Johnson (2013) to investigate causes that impact the persistence of first-year students at a university.

Check sheets are a simple, yet versatile, quality tool used for data collection. Once collected, the data can then be analyzed and are frequently used in places where failures occur such as manufacturing processes that produce out of specification parts (Carita, 2014). Although check sheets can be electronic, they have the advantage of simplicity in not requiring advanced skills and can be used in a paper format (J. D. Smith, 2019).

Bothe (2001) describes the use of check sheets for a shipyard that was confronted with excessive crane downtime; therefore, a checklist was implemented to collect data to better understand the problem, so that the root cause could be identified and corrected. This was first done to identify which of the five cranes was experiencing problems. Upon determining that all five cranes experienced downtime, a new check sheet was used to identify downtime through work shift, and the results indicated only one shift was experiencing a problem. This information sufficiently localized the problem for the root cause to be identified. J. D. Smith (2019 p. 43) describes the use of a check sheet with categories such as “system moving slow” and “internet connection issues” as well as “other” for unwanted events that do not fit the listed categories.

The term checklist is sometimes used to describe a version of the check sheet, which is a list of steps, operations, or tasks to perform. This serves as either a reminder that something must be done or a verification that it was done (Blank, 2014). Chakravorty (2016) presents an example of this type of check sheet when describing a jet repair operation that was frequently plagued with aircraft moving forward in the repair process with unfinished operations. The on-time delivery of aircraft increased from 31.3% to 93.2% with the introduction of checklists to ensure all operations are carried out before an aircraft leaves a repair station.

There is a plethora of control charts available, and the correct control chart must be selected based on the type of data and the size of the sample. Control charts may be for individual measurements, multiple measurements together, defective parts, or defects. Control charts plot the position of the sample relative to the mean of the process and there are control limits three standard deviations above and below the process mean. Values that fall within the control limits display the normal variability that is inherent to the system. Values that are above the upper control limit or below the lower control limit are special cause variability, which is an indication that something has happened (B. Anderson, 2007).

Jukka Rantamäki et al. (2013) describe the implementation and use of control charts in a Finnish paper mill where machines run almost continuously and are only rarely shut down for maintenance purposes. Employees would review the control chart and address problems. To facilitate root cause identification by individual operators, generic cause and effect diagrams were created with known causes listed. Any problems that could not be immediately solved by operators were passed on to problem solving teams that would seek to identify root causes.

Flow charts use symbols to graphically depict the steps in a process or series of operations (Neyestani, 2017). For example, a sideways oval can depict the beginning and end of a process, a rectangle depicts a process step, a diamond depicts a decision, and a transportation operation can be depicted using an arrow. A flow chart may provide only a high-level view of a process or a low-level view that includes all steps and movements within the process (Villarreal & Kleiner, 1997).

Johnson et al. (2006) present a simple high-level flow chart for a paper helicopter process. The process starts with the need for more paper helicopters and then material is requisitioned. The materials are then fabricated and assembled. The paper helicopter is then tested and a decision is made; rejected helicopters are replaced and the process starts over with the requisition of material, or the process ends if the paper helicopter is acceptable.

Histograms are useful for graphically observing the frequency at which something occurs. The shape of the histogram can be used to observe the amount of variability in a process, and the shape can provide an indication as to what is happening in the process (Blank, 2014). Values are placed in bins with the bin size often equal to the square root of the number of values used. The bins are on the x-axis, and the number of occurrences is listed on the y-axis (Sleeper, 2006).

Borrows (2005) describes the use of a histogram to assess data from a bottle filling process that should fill bottles with a 5.5% bleach solution. The process had a mean of 5.504%; however, production data were observed in a histogram and found to be slightly left-leaning. The values were spread around the mean, which was close to the target value, but the data could be seen to have more values than on the high side than would be expected with a symmetrical distribution. This has led to hypothesizing potential causes.

The Pareto principle is attributed to Juran, who noticed the concept while consulting at General Motors in the 1930s and named the concept in the late 1940s. However, Juran named the concept after the Italian economist Vilfredo Pareto, who observed 80% of wealth in Italy was owned by 20% of the population. Juran later came to realize that the concept was misnamed as Vilfredo Pareto and did not expand beyond wealth. The Pareto principle was originally said to separate the vital few from the trivial many (Juran, 1975). However, this changed over time to the vital few and the useful many. It is used to identify the few problems that result in the most costs so that these problems can be improved, delivering maximum impact with less effort (Gryna, 2001).

A Pareto chart can be created for more than just costs. For example, Sharma et al. created a Pareto chart for types of defects. The Pareto chart showed that streaking defects alone counted for over 60% of all failures (2018). Alternatively, a Pareto chart can be used as a source of quality problems, costumers that generate the most sales, or even reasons for employee absenteeism (Murugaiah et al., 2010).

A scatter plot can also be known as a scatter diagram, and it is used to show potential relationships between paired values. The values are plotted graphically and observed to see if a potential relationship exists. A relationship does not exist when it looks like the values are scattered randomly (Villarreal & Kleiner, 1997). Although useful for identifying potential relationships, a scatter plot only provides an indication that there is a possible relationship and no confirmation of the relationship (Blank, 2014). In a scatter plot, the x-axis lists process outputs and the y-axis lists process inputs (Snee, 2008).

Sleeper (2006) presents an example of a scatter plot with miles per gallon used in two-seat vehicles in the city on the x-axis and the size of the vehicle engine on the y-axis. The histogram shows a possible relationship between the amount of fuel and the size of the engine, with fuel consumption increasing as engine size increases, with two outliers that were sports cars that had small engines and high fuel consumption.

Snee (2008) recommends always assuming there is a difference when two different production lines are producing the same thing. In such situations, stratification should be used. Stratification is the separation of data using categories such as characteristics or causes to separate the data. This is done to provide clarity (Minzuno, 1989). According to Soković et al. (2009), stratification is referred to as a flow chart or a run chart by some authors.

During a Six Sigma project, N. C. Anderson and Jamison (2014) used stratification to graph the failure rate of welders over time. However, some welders had special training and some did not; therefore, the data were stratified to separate the two groups of welders, which showed a difference in performance that would not have been visible if both groups had been mixed together.

Data are plotted in time order in a run chart. The run chart is used to identify changes or trends over time and is much like a control chart; however, the run chart does not use control limits (Villarreal & Kleiner, 1997). In an example, Snee (2008) listed time on the x-axis and listed yield on the y-axis to illustrate the use of a run chart for a pharmaceutical yield problem. Patterns that may occur in a run chart are often the result of differences between shifts, material batches, and seasons, as well as the result of tool wear (Blank, 2014).

N. C. Anderson and Jamison (2014) illustrate the use of run charts with an example for tracking the performance of welders by plotting repair rates on the y-axis and months on the x-axis. In addition, target values were plotted on the run chart to monitor performance so that causes can be investigated if performance goes below a target value.

A quality tool cannot solve a problem by itself. However, the use of a quality tool can be helpful for information gathering and assessment. The information analyzed can be used to find the root cause (Blank, 2014). Quality tools serve different purposes and should be approached like a tool box, with the quality tool used selected based on the intended use of the quality tool (B. Anderson, 2007). Multiple quality tools may be needed, and this is illustrated by Bamford and Greatbanks (2005) who explain which combinations of quality tools can be used for various everyday problems.

Furthermore, although the quality tools have their specific typical uses, they can be used for multiple types of different objectives when performing an RCA. For example, check sheets can be used to collect data for analysis and again after an improvement has been implemented to determine if the improvement is effective and lasts. Both control charts and histograms can be used for assessing the initial performance of a process, analyzing a process to find a root cause, and again after an improvement has been implemented to monitor process performance (Stamatis, 2002). Although Ishikawa believed most problems could be solved with only the classic seven quality tools, studies have shown that few people using the classic seven quality tools believe most problems can be solved using the classic seven quality tools alone (Antony et al., 2021; Mathur et al., 2022; McDermott et al., 2022).

The classic seven quality tools can be used as standalone tools. However, they can also be used as part of problem-solving and quality improvement processes such as PDCA (Plan Do Check Act), Design for Six Sigma, and Six Sigma (Soković et al., 2009). Furthermore, Tang has found that methodologies such as PDCA, Six Sigma, 8D, and A3 all have significant overlaps in the actions taken and the classic seven quality tools used (2023).

Six Sigma, also known as Lean Six Sigma, is a statistics-oriented improvement methodology based on DMAIC (Define Measure Analyze Improve Control). The first step is defining the problem, and the second step is measuring to collect data. The data are then analyzed and improvements are identified and implemented. Control actions are then taken to ensure the improved performance is maintained (Araman & Saleh, 2023). Marques and Matthé (2017) present a case study in using Six Sigma to improve the performance of a die casting operation. The Six Sigma project used classic seven quality tools consisting of a flow chart, Pareto chart, cause and effect diagram, and a graphical check sheet. Although quality tools may be used in Lean projects, a survey of lean tools and methods used in ISO 9001 certified organizations did not find any overlap between the classic seven quality tools and lean tools and methods (Sá et al., 2022).

Ishikawa (1985) describes the QC Story, which is an approach to problem-solving that uses the classic seven quality tools. There are nine steps in the QC Story, starting with selecting a theme and objectives and then explaining why the theme was selected. The current situation is then assessed, an analysis to identify causes is performed, and then corrective actions are identified and implemented. Finally, standardization is implemented, remaining problems are considered, and then plans are made for future actions. Other approaches to problem-solving and quality improvement include the 8D methodology and the A3 methodology. Although both methodologies are based on reporting documents, they also have associated process steps. However, the 8D methodology is more suitable for problem-solving when attempting to identify the cause of a failure during RCA and the A3 methodology is more suitable for quality improvement, such as scrap reduction (Barsalou, 2023).

The 8D methodology has eight steps, starting with forming a team, defining the problem, and taking actions to contain the problem. An RCA is then performed using quality tools such as the cause and effect diagram. A solution is then identified and implemented. Actions are taken to ensure the problem is prevented from reoccurring, and the team is congratulated at the end of the process (George et al., 2021).

The A3 methodology is based on PDCA and originated with Toyota Motor Company. The report and methodology are named after the size of paper used to create an A3 report. The A3 report has boxes based on the steps of the A3 methodology, which consists of explaining the history of the problem, describing the current situation, and presenting the objective or goal. There are also boxes for the analysis, planned improvements, a plan for implementing improvements, and follow-up actions. A cause and effect diagram is used for the RCA as part of the analysis step (Santos Filho & Simão, 2022).

Another approach to problem-solving and continuous improvement is Kata Improvement, which is based on understanding the direction of change, understanding current conditions, establishing a near-term objective, and experimenting to achieve the desired objective (Legentil et al., 2018). A case study in the application of the Kata Improvement methodology in a fast-food restaurant included the use of three of the classic seven quality tools, consisting of cause and effect diagram, Pareto chart, and histogram (Suárez-Barraza et al., 2021).

A study of European health-care facilities by McDermott et al. found that 62% of the respondents were trained in the use of the classic seven quality tools. The most commonly used quality tools were the cause and effect diagram, histogram, and check sheet. The least used quality tool was stratification (2022). A study of seven classic quality tools in educational institutions found the most commonly used quality tools to be the cause and effect diagram and the Pareto chart (Mathur et al., 2022).

3. Methodology

The literature was reviewed to determine which quality tools various authors considered to be part of the classic seven quality tools. The classic seven quality tools, as defined by eight authors, are shown in Table 1. The authors were identified during the literature review based on databases from journal publishers and the American Society for Quality. Specifically, eight authors were selected to convey the wide range for tools identified by various authors as the classic seven quality tools.

Table 1. The classic seven quality tools according to various authors

Barsalou (2017)	Borror (2009)	Ishikawa (1991)	Okes (2002)	J. Smith (2017)	Soković et al. (2009)	Sreedharan et al. (2018)	Stamatis (2002)
Cause and Effect diagram	Cause and Effect diagram	Cause and Effect diagram	Check sheet	Check sheet	Cause and Effect diagram	Cause and Effect diagram	Cause and Effect diagram
Check sheet	Check sheet	Check sheet	Control chart	Control chart	Check sheet	Check sheet	Check sheet
Flow chart	Control chart	Control chart	Flow chart	Flow chart	Control chart	Control chart	Control chart
Histogram	Flow chart	Histogram	Histogram	Ishikawa diagram	Histogram	Histogram	Flow chart
Pareto chart	Histogram	Pareto chart	Ishikawa diagram	Run chart	Pareto chart	Pareto chart	Histogram
Run chart	Pareto chart	Scatter plot	Pareto chart	Scatter plot	Scatter plot	Scatter plot	Pareto chart
Scatter plot	Scatter diagram	Stratification	Run chart	Pareto chart	Flow chart	Stratification	Scatter plot
			Scatter plot

There was much overlap in quality tools listed by the authors; however, there were some differences and the name of a specific tool sometimes varied between authors. In the cause of the cause and effect diagram, the study listed cause and effect diagram, fishbone diagram, and Ishikawa diagram as one tool in case respondents only knew of the tool under a specific name.

The tools then identified for the study were cause and effect diagram, check sheet, control chart, flow chart, histogram, Pareto chart, run chart, scatter plot, and stratification. Although more than seven quality tools were used, this was due to the differences between authors in which tools they listed. For example, some authors listed a run chart (Barsalou, 2017; Borror, 2009) and other others did not (Soković et al., 2009; Stamatis, 2002). Although not one of Ishikawa’s original seven quality tools (Ishikawa, 1991), the flow chart was added due to being included by other authors (Barsalou, 2017; Borror, 2009; Okes, 2022). This resulted in nine quality tools being listed and none of the above options was added in case respondents did not use any of the listed quality tools.

A quality distribution list in an international organization was used to share the survey. The organization was a tier 1 supplier of assemblies in the automotive industry. The organization was a large organization; however, the individual locations range in size from over 500 employees to under 1,500 employees. These locations produced two different types of assemblies for OEM (Original Equipment Manufacturers) in the automotive industry.

The survey was created in Microsoft forms and was completely anonymous. There were 17 responses with seven in Germany, five in Portugal, two each from China and America, and two and one India. Almost half of the respondents had over 10 years of experience in performing RCAs. Specifically, 8 had 10 or more years of experience, 5 had 4 to 6 years of experience, 3 had 6 to 8 years of experience, and one had 8 to 10 years of experience. There were eight who reported their position as engineers, five who were managers, two were senior managers, directors, or above, and one each for technician and none of the above.

Study participants were presented with a list of quality tools and asked to select the most frequently used quality tool when performing an RCA. They were also presented with lists and asked to select the second most frequently used quality tool and the third most frequently used quality tool. The results, together with the combined results, are in Table 2.

Table 2. Frequently used quality tools

Quality Tools	Most Frequent Used	Second Most Frequently Used	Third Most Frequently Used	Results Combined
Cause and effect diagram	12	4	2	18
Check sheet	0	1	1	2
Control chart	2	1	3	7
Flow chart	2	1	1	4
Histogram	0	1	2	3
Run chart	0	2	0	2
Scatter plot	0	1	2	3
Pareto chart	0	3	3	6
Stratification	0	0	0	0
None of the above	1	0	0	1

A chi-square goodness-of-fit test was used to assess the combined results to determine if any values occur more or less often than would be anticipated if the occurrence of all values were equal (Keller et al., 1994). Minitab Statistical Software version 21 was used, and the resulting p-value was below 0.05; therefore, the actual occurrences did not match what would be expected if all values occurred equally (see Table 3).

Table 3. Chi-square goodness-of-fit test observed and expected counts

Category	Observed	Test Proportion	Expected	Contribution to Chi-Square
Cause and effect diagram	18	0.1	5.1	32.6294
Pareto chart	8	0.1	5.1	1.6490
Control chart	7	0.1	5.1	0.7078
Flow chart	4	0.1	5.1	0.2373
Scatter plot	3	0.1	5.1	0.8647
Histogram	3	0.1	5.1	0.8647
Run chart	3	0.1	5.1	0.8647
None of the above	3	0.1	5.1	0.8647
Check sheet	2	0.1	5.1	1.8843
Stratification	2	0.1	5.1	5.1000
Chi-Square Test
N DF Chi-Sq P-Value
51 9 45.6667 0.000

Figure 1 depicts both the expected values of all values occurred equally and the actual occurrence of the values. Here, the cause and effect diagram, Pareto chart, and control chart occur more often than would be expected if all values were equal. All other quality tools occur less than what would be anticipated if all values were equal.

Figure 1. Chart of observed and expected values from the chi-square test.

Figure 2 shows the contribution to the chi-square value by category. Here, the cause and effect diagram had the greatest impact on the chi-square value. The second greatest deviation was for stratification. However, stratification’s impact was due to occurring less than would be expected if all values were equal. In this case, none of the respondents selected stratification.

Figure 2. Chart of contribution to the chi-square value by category.

The study participants were again given a list of the quality tools; however, multiple quality tools could be selected this time. They were then told to select the quality tools that they think everybody should know for the RCA. The results are in Figure 3. The cause and effect diagram was selected more than other quality tools, with only one respondent that did not select the cause and effect diagram. Four other responders selected only the cause and effect diagram and the remaining respondents all selected at least two or more quality tools.

Figure 3. Classic Seven Quality Tools everybody should know for RCA.

4. Discussion

The cause and effect was by far the most used of the classic seven quality tools. This makes sense as an Ishikawa diagram is useful for listing explanatory hypotheses in many types of investigations, ranging from a service failure to a product failure.

The cause and effect diagram is also consistently listed as one of the classic seven quality tools by the author in Table 1, although the name may vary, and this may also provide a small degree of contribution to the Ishikawa diagram being selected more often than other quality tools. The other quality tools were not selected as often as an Ishikawa diagram. This makes sense as the Ishikawa is versatile and can be applied to many types of problems, many of the other quality tools have more specific uses.

The cause and effect diagram is useful for listing explanatory hypothesis that can explain the cause of the failure under investigation and is applicable to a wide range of uses, which may explain its popularity. Although the cause and effect diagram itself does not point directly towards the root cause, having all hypotheses in one graphical depiction can be advantageous and once listed, the hypotheses can be prioritized. A cause and effect diagram is not a static quality tool and it should be updated as needed, such as when a hypothesis is confirmed to be rejected or new data results in a new hypothesis.

The control chart was the second most commonly identified quality tool. In contrast, the run chart was not often selected. The control chart is one of the original classic seven quality tools, but some authors (J. Smith, 2017; Okes, 2002) list the run chart in place of the control chart. Both control charts and run charts require data in time series and cannot be used correctly if the order in which the values were created is unknown. Control charts require the use of statistical software, or mathematical calculations, which can be a disadvantage if an organization lacks software and the required knowledge to implement control charts.

Alternatively, organizations without statistical software or the knowledge required to calculate control limits can use run charts for viewing data collected in the order in which the values were produced. A run chart cannot asses the stability of a process, but a run can give an indication that something out of the ordinary has occurred at a specific time or date and this can be investigated further.

Flow charts came in fourth place, even though the flow chart is not one of the original classic seven quality tools. Flow charts are useful for gaining a deeper understanding of the operations of a process. Simply observing a process to create a flow chart can lead to discovering that the process is not as it should be, such as when operators deviate from a procedure.

Check sheets are useful for data collection and can be used prior to hypothesis creation to collect data, or as part of an investigation action such as for determining if a specific product variant is experiencing the problem. Alternatively, a check sheet can be used for process monitoring after implementation of a corrective action.

A histogram can be used to quickly assess the spread of data and to see if the process is skewed. Outliers can also be observed in a histogram when a value is located far removed from the other values. Histograms can be quickly created in statistical software. Alternatively, they can be created using a pen and paper with a little bit of math, or in a spreadsheet program.

The Pareto chart is not useful for finding a root cause. However, a Pareto chart is useful for prioritization, such as deciding which problem to address first when confronted by multiple problems. The Pareto chart is another quality tool that can be created in statistical software, with pen and paper, or in a spreadsheet.

Scatter plots can also be created with pen and paper, spreadsheets, or statistical software. Scatter plots are useful for identifying potential relationships between paired data points. However, a scatter plot cannot test the potential relationship for statistical significance and when possible, a regression analysis should be performed. However, even without a regression analysis, potential relationships identified by a scatter plot may offer a basis for further investigation.

Stratification is often listed as one of the classic seven quality tools and it is even one of the original tools identified by Ishikawa (1991); however, it is more of a concept to apply when using other methods to assess data. For example, separating data by production machine when using a histogram.

The study results are comparable to a study in health-care organizations by McDermott et al. (2022) in regard to the cause and effect diagram being one of the most commonly used quality tools, with stratification being one of the least commonly used quality tools. The differences between the studies may potentially be due to differences between a health-care setting and the automotive industry. In addition, this study was limited to the application of the classic seven quality tools for RCA, while the health-care organization may have included quality use in general, such as quality improvement. The study results are also comparable to a study of educational institutions by Mathur et al. (2022), who also found the cause and effect diagram to be the most commonly used of the classic seven quality tools.

5. Conclusion

A survey was conducted to identify which of the classic seven quality tools are used by quality practitioners during an RCA. Respondents overwhelmingly selected the cause and effect diagram, both as the most used quality tool and as the most recommended quality tool. However, extensive use of one quality tool should not exclude the use of other quality tools as needed. The classic seven quality tools should be seen as complementary and not in conflict with each other.

The classic seven quality tools should be approached as a versatile and easy to use tool box to select from based upon need. For example, histograms are great for graphically viewing data. However, a histogram would not be needed for an RCA when there are no numerical data, such as the failure of a unique prototype part or a one-time transaction as part of a service. The quality tools can also build upon each other. For example, data may be collected in a check sheet, viewed in a run chart, and a hypothesis formed based on the data could be listed in a cause and effect diagram.

The author recommends a basic RCA tool kit consisting of cause and effect diagrams, check sheets, and control charts if the organization has the ability to use control charts and run charts for organizations that lack the means to use control charts, flow charts, histograms, and scatter plots. Stratification should be a consideration when assessing data with graphical quality tools.

Although the classic seven quality tools serve as a good starting point for an RCA tool kit, additional quality tools should be considered. For example, an is/is-not analysis for both writing problem statements and identifying critical differences to investigate and five reasons for digging down the underlying causes.

The study has limitations. The sample size was low and the study was world-wide but only involved one organization. A consideration for future research would be to conduct a comparable study across multiple organizations. Such a study could also ask respondents if there is one specific problem-solving they use, such as the eight steps of an 8D report, or Six Sigma’s DMAIC approach. This would be useful for determining if specific quality tools are strongly associated with a specific problem-solving methodology.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Matthew Barsalou

Matthew Barsalou has a master’s degree in business administration and engineering (industrial engineering) from Wilhelm Büchner Hochschule, and a master’s degree in liberal studies from Fort Hays State University. He is an Academician in the International Academy for Quality, an ASQ fellow, and past chair of ASQ’s Statistics Division. He is certified as a lean Six Sigma Master Black Belt and an ASQ-certified Six Sigma Black Belt, manager of quality/organizational excellence, quality technician, and quality engineer and holds TÜV certifications as quality manager, quality auditor, ISO/TS 16949 auditor, and quality management representative and is also an INTACS certified Automotive SPICE® Provisional Assessor. His main research interest is on root cause analysis. This paper is a small part of an overall theme aimed at developing an optimized and easy to apply approach to root cause analysis with a supporting tool set.