Developing and Testing a Framework for Continuous Improvement in Visual Inspection Processes: a Case Study

Developing and Testing a Framework for

Continuous Improvement in Visual Inspection

Processes: a Case Study

Master thesis, MSc Supply Chain Management

University of Groningen, Faculty of Economics and Business

ABSTRACT

To many companies visual inspections are inevitable to comply to the increasing quality demands of their products. However, visual inspections are not perfectly reliable as inspectors are known to be subjective and miss up to 20% of product defects. Numerous procedures have been investigated that aim to improve the performance of these processes. Whereas previous research studied the procedures mainly separately from each other, this study opts to develop an integrated framework for the Continuous Improvement of inspection processes, containing six key procedures: training, feedback, feedforward, decision standard, job enrichment and incentives. The core of the framework is formed by an in-literature-grounded questionnaire that aims to expose aspects for improvement regarding the selected procedures. By means of a single case study the framework was tested. Initial insights suggests that the framework offers a useful tool, however, further research is required to determine whether the framework actually leads to a better performance.

Keywords:

Table of Contents

ABSTRACT ...

1. INTRODUCTION ...1

1.1 Objective and contribution ...1

1.2 Outline of the study ...2

1.3 Framework ...2

1.4 Research design ...3

2. LITERATURE ...3

2.1 Performance indicators ...3

2.2 Procedures for Continuous Improvement of Visual Inspection Processes ...3

2.2.1 Continuous Improvement ...3

2.2.2 Megaw’s overview of procedures for inspection improvement ...3

2.2.3 Procedures for improving inspection performance ...4

2.2.4 Motivation ...5

3. METHODOLOGY ...7

3.1 Framework for Continuous Improvement ...7

3.2 Case description ...7

3.3 Data collection ...8

3.4 Questionnaire ...8

3.5 Data analysis ...8

4. RESULTS ...9

4.1 Framework for Continuous Improvement ...9

4.2 Procedures for Continuous Improvement ...9

4.2.1 Training ... 10

4.2.2 Feedback ... 10

4.2.3 Feedforward ... 11

4.2.4 Decision standard ... 11

4.2.5 Motivation ... 12

4.2.6 Job enrichment ... 12

4.2.7 Incentives ... 13

4.3 Selection of improvement activities ... 13

5. DISCUSSION & CONCLUSIONS ... 14

5.2 Practical implications ... 15

5.3 Limitations and future research suggestions ... 15

5.4 Acknowledgements ... 15

REFERENCES ... 16

LIST OF FIGURES AND TABLES ... 20

APPENDIX A: Questionnaire with annotations ... 21

1 1. INTRODUCTION

Within modern factories an increasing amount of products is manufactured at an increasing pace. A main aspect is the quality level of these products. An area wherein quality plays a major role is in factories with assembly lines, such as in the electronics and automotive industry. Within these industries, quality gained importance over time due to increasing competition, stricter environmental regulations and more demanding customers (Lou and Huang, 2003). These developments have a major influence on these industries and force companies to perform at a higher level.

Quality is a hot topic and is widely addressed in both the academic world as in practice (Ebrahimi and Sadeghi, 2013). Still there are a lot of different opinions about what quality precisely is (Reeves and Bednar, 1994). A reason for this is the angle and the context in which quality is approached. For instance within a hospital, quality has a different meaning than within a factory. Despite these differences, a common ground is to ensure and improve quality. Within production environments there are roughly three ways to improve quality. One can prevent quality defects from occurring, discard quality defects (Ng and Hui, 1996) or repair quality defects. Quality defect prevention is mainly carried out by improving production processes, discarding quality defects is often done through quality inspections. The focus of this research will be on quality inspections.

Although quality inspections can be seen as waste (Oakland, 2003) it is inevitable for many companies to ensure their quality levels. A main function of quality inspections is the detection of quality deviations of products. Detection can be done automatically, i.e., through sensors/machines/computers, by humans, i.e., visual inspection or a combination of these methods, i.e., a hybrid approach. Although automated inspection is gaining popularity (Jiang et al., 2003), human visual inspection will continue to be part of any foreseeable inspection strategy (Rao et al., 2006). This research will focus on human visual inspection (from now on: visual inspection) and will examine how this process can be improved. Nalanagula et al., (2006) defined a visual inspection as ‘‘the process of locating a defect within the area of interest and deciding whether the defect is acceptable or not.’’ The two main functions and also main determinants of visual inspections are the visual search for deviations and deciding whether or not the deviation is acceptable (Drury, 1975). Human inspectors are flexible and have good decision making abilities (Chabukswar et al., 2003; Sannen and Van Brussel, 2013), but are prone to human error and are therefore not perfectly reliable (Sadasivan and Gramopadhye, 2007). Namely, inspectors miss defects, are subjective and can judge differently from other inspectors (Wang and Drury, 1989; Debrosse et al., 2010). By addressing the above-mentioned weaknesses of visual quality inspections a higher quality level is likely to be obtained.

The necessity to improve visual inspections is evident, as the number of defects that is missed by inspectors is reported to be around 20% (Ishikawa, 1990; Juran and Blanton Godfrey 2000). Many empirical studies have been conducted to identify the factors that influence search and decision making within visual inspection (e.g., Megaw, 1979). These factors, vary from illumination to inspection time and from personality of inspectors to search strategy. The objectives of these studies were either to understand the effect of these factors on inspection performance as to identify procedures to reduce the ill effects due to these factors (Jiang et al., 2004). Examples of such procedures include: different types of training (e.g., Wang et al., 1997; Gramopadhye et al., 1997a; Nickles et al., 2003), feedback (e.g., Gramopadhye et al., 1997b; Chabukswar et al., 2003) and the use of aids (Yeow and Sen, 2004). Due to the fact that there are numerous procedures to improve inspection performance, that are generally studied separately from each other, it is useful to develop an integrated approach that helps to identify and select procedures that suits a specific inspection process. To do so, this research will use the classification of Megaw (1979) as a starting point.

1.1 Objective and Contribution

The objective of this research is to develop a framework that supports continuous improvement within a human visual inspection process in a production line, using procedures as included in the overview of Megaw (1979), in order to increase its performance. The framework will be tested in a case study in order to determine its effectiveness.

2 1.2 Outline of the Study

First, the framework that is to be developed will be further explained. Next, in the second chapter relevant literature will be discussed to create a solid basis for data collection and analysis. In chapter three the methodology of this research is outlined. Subsequently, chapter four presents the results of the conducted case study. The final chapter contains a discussion and presents the conclusions that can be drawn.

1.3 Framework

The framework for Continuous Improvement of visual inspection processes is shown in Figure 1. The three steps will be further elaborated below.

CONTENT

Step 1. Determination of performance indicators

Step 2. Selection of procedures for Continuous Improvement

Step 3. Determination of set of improvement activities

Figure 1. Framework for Continuous Improvement of visual inspections

The first step of the framework involves the determination of performance indicators of the inspection process. Performance measurement is a prerequisite for improvement, as it gives an indication of how well organizations or processes are doing and where (the most urgent) improvements have to be made (Slack et al., 2006). There should be a clear link between the strategic objective and the set of indicators that are used to measure performance, so that it is ensured that the indicators reflect the actual performance (Anupindi et al., 2011). Next to identifying a need for improvement, performance indicators can be used to reflect on the effect of utilized improvement activities.

Step two of the framework is twofold. First, procedures for improving inspection performance are selected and incorporated in the framework. This selection will be derived from a demarcation of the overview of Megaw (1979), as shown in Figure 2, which will be done in the literature chapter. Second, an in-literature-grounded questionnaire will be developed to assess the functioning of an inspection process in relation to the selected procedures. The questionnaire aims to expose the strengths and weaknesses of an inspection process and it indicates the aspects that need to be improved. The questionnaire also forms the core of the framework, as it is the mean to diagnose a visual inspection process.

The third and final step of the framework is about the determination of a set of improvement activities. This step is closely tied to the previous step, because the determination builds further upon the outcome of the questionnaire. However, the questionnaire does not intend to provide a conclusive approach that straightly leads to the ‘best’ set of improvement activities. Rather it provides a direction for improvement opportunities.

The framework offers a tool that aids managers to assess a visual inspection process and to identify the most important areas for improvement. Due to its structured approach it can prevent managers from overseeing possibly important aspects, or focusing on details instead of main issues. The framework will be developed as such that it can be applied to any visual inspection process. The steps of the framework and subsequently the testing of the framework result in the following research questions:

1. What performance indicators can be used for visual inspection processes? 2. What procedures can be used to improve visual inspections processes?

3. How can a Continuous Improvement framework be developed that can be applied to visual inspection processes?

3 1.4 Research Design

The research design that has been chosen to perform this study is a case study. Within this case study it will be tested whether the framework is successful. Due to the explorative nature of this research, a case study seems the most appropriate research design. Namely, it allows the collection of rich data and offers the opportunity for an in-depth understanding of the phenomenon that is studied (Yin, 1993). Although the findings of a single case study has a more limited generalizability than a multiple case study, it does allow a more in-depth understanding (Karlsson, 2009).

2. LITERATURE

This chapter is structured in accordance with the sequence of steps of the framework. Hence, under 2.1 the research question: ’’What performance indicators can be used for visual inspection processes?’’ will be addressed and under 2.2 the research question: ‘’What procedures can be used to improve visual inspection processes?’’ is discussed.

2.1 Performance Indicators

Traditionally, inspection performance has been measured with speed, accuracy, flexibility, reliability and cost (Jiang et al., 2003). However, flexibility and reliability are hard to measure and are therefore not frequently used. The principal performance indicators for visual inspection performed by humans concern speed and accuracy (Nickles et al., 2003). Speed can be expressed in (mean) search time for locating faults, (mean) time to classify and decide upon fault, (mean) stopping time which is the time at which the inspector terminates the search and moves to the next item, and (mean) time required to perform an entire inspection. Accuracy can be measured through the number of faults detected (Gramopadhye et al., 1998; Chabukswar et al., 2003), the number of faults that is not detected (Gramopadhye et al., 1998; Drury et al., 2009) and the number of false alarms which is an anomaly that is unjustified classified as a fault (Jiang et al., 2003; Drury et al., 2009). A less common indicator is response failure which indicates that the wrong action is taken despite successful search and decision (Drury et al., 2009).

2.2 Procedures for Continuous Improvement of Visual Inspection Processes 2.2.1 Continuous Improvement

Continuous Improvement (CI) is an ongoing approach that aims to improve performance of processes, products and services (Slack et al., 2010). It is an approach that involves people from all levels in an organization to make improvements without necessarily making big capital investments (Bessant et al., 1997). This can be either incremental improvement occurring over time or a breakthrough improvement occurring all at once. CI is a broad concept and draws on features that are mentioned under different names, such as Lean, Six Sigma and Total Quality Management (Berger, 1997). CI is also known as kaizen which find its origins in the Japanese industry. In either form, improvement is achieved by using methods that are dedicated to search for the source of problems, waste and variation, and finding ways to minimize them (Bhuiyan and Baghel, 2005). Two of the most well-known methods of CI are the Plan Do Check Act (PDCA) and the Define Measure Analyze Improve Control (DMAIC) cycle. Both methods are cyclical of nature, ensuring a never-ending process of improving processes (Slack et al., 2006). As companies are forced to improve performance, the need to improve continuously on a large scale within organizations increased (Bhuiyan and Baghel, 2005). Also, CI principles have proven to be successful (Fryer et al., 2007). However, not all CI methods can be applied to any organization as they are not necessarily effective at solving all issues (Bhuiyan and Baghel, 2005). As a consequence, organizations recently started to develop their own CI methods that fit their specific needs (Bhuiyan and Baghel, 2005).

2.2.2 Megaw’s overview of procedures for inspection improvement

4 about the actual inspection task that is to be performed. Lastly, Organizational factors concern all conditions that are set by the organization.

Within this research the emphasis will be on procedures related to the behavior of inspectors. With this approach it is assumed that the physical environment and the poule of employees is set. Therefore, Subject- and Environmental factors fall out of the scope of this research. Task factors also fall out of the scope of this study, since these factors mainly concern the quality of the product which cannot be influenced within an inspection. Consequently, the procedures that will be included in the framework concern Organizational factors. This demarcation prevents the framework from becoming immense and ensures the conciseness of this research. Further, the focus can help to ensure that sufficient depth is obtained in this research.

Figure 2. Factors that might influence inspection performance, derived from Megaw (1979)

A number of Organizational factors are revealed in Figure 2. It is important to note the difference between factors influencing inspection performance, as mentioned in Figure 2, and procedures that aim to improve inspection performance. Namely, not all factors are procedures. For instance, the shift in which inspectors work can affect performance, but is not a procedure to improve performance. Training, on the other hand, is as well a factor as a procedure that affects performance. Further, there can be overlap between factors. For instance, feedforward concerns information that is given prior to the inspection task (Nalanagula et al., 2006) including, amongst others, instructions. However, feedforward and instructions are in the overview of Megaw (1979) considered as two different factors. Factors that show a significant amount of overlap are in this research covered under a single procedure. Below the procedures that are expected to be most relevant and that fall within the course of the framework are described.

2.2.3 Procedures for improving inspection performance

5 means that the same location may be inspected multiple times (Nalanagula et al., 2006). On the opposite, a systematic search strategy is defined as a process with perfect memory (Williams, 1967) in which the same location is never inspected more than once. A human inspector’s search strategy falls somewhere in between random and systematic (Arani et al., 1984). Multiple researchers proved that a systematic search strategy is more effective (Tsao et al., 1979; Karwan et al., Gramopadhye et al., 1997b) and that it can be trained (Wang et al., 1997).

Feedback is a dominant method for the improvement of visual inspections. Feedback can be provided either at individual or group level. In order to be effective, feedback needs to be objective (Buelens et al., 2006). Further it is important that not only positive or only negative feedback is provided (Bannister, 1986). Types of feedback specific to visual inspections are performance feedback, also known as outcome feedback, and cognitive feedback, also known as process feedback. Performance feedback is provided to an inspector after the inspection has been performed and concerns knowledge of the outcome of an inspection in terms of search times, search errors and decision errors (Micalizzi and Goldberg, 1989). This type of feedback has shown to improve inspection performance (Gramopadhye et al., 1997b). Cognitive feedback provides information on the process by which the inspector achieved the results. Cognitive feedback has several times been researched in relation to search strategy and it was found that it supports the adoption of a systematic search strategy and is (e.g. Gramopadhye et al., 1997b; Chabukswar et al., 2003). Also they conclude that cognitive feedback is better than performance feedback for visual search.

Feedforward is a commonly used method in visual inspections and has shown to improve performance (e.g., Harris and Chaney, 1969; Nalanagula et al., 2006; Rao et al., 2006). Feedforward is information that is given to the inspector prior to performing the inspection task and aims to reduce the uncertainty of inspectors and, subsequently, improve inspection performance (Drury and Gramopadhye,1992 ; Nalanagula et al., 2006). This information can be instructions, rules or guidance, demonstration, or information gathered from colleagues about where to look and what to look for (Gramopadhye., et al 1997a), i.e., where on the product can what type of quality deviation be expected. With this information inspectors know better how to look and thus perform the inspection more effective and efficient (Nalanagula et al., 2006). Feedforward also gives insight in the product, the process and the nature of quality deviations. The provision of feedforward is especially helpful for novel inspectors, since they have limited knowledge of the type of quality deviations and the location where the quality deviation is likely to occur (Nalanagula et al., 2006). Reversely, experienced inspectors possess more of this type of knowledge and benefit less from feedforward.

The decision standard plays an important role in visual inspection, by supporting the decision regarding accepting or rejecting a product. This decision is considered as the most critical point of the inspection process as it depends on the subjective judgment of an inspector (Baudet et al., 2013). The purpose of decision standards is to harmonize the inspector’s decisions (Debrosse et al., 2010) The standard prescribes what zones to control, what elements to check, how to classify quality deviations and what decision to make, i.e. accept or reject (Debrosse et al., 2010). An important aspect according to Debrosse et al., (2013) is that the decision standard not only defines what to inspect, but also how to inspect. This means that it should indicate the distance at which the product needs to be inspected, the search strategy that is to be used and the environmental conditions, including illumination and reflection, under which products need to be inspected and the decision has to be made (Debrosse et al., 2010; Baudet et al., 2013). Furthermore it is of importance that the decision standard is known by the inspectors and that it is uniformly interpreted (Kleiner and Drury, 1993).

2.2.4 Motivation

6 fact that motivation has hardly been researched in relation to visual inspections, it is worthwhile to take a further look at it.

Motivation has, to the best of the author’s knowledge, only once been researched within the context of visual inspections (e.g. Stok, 1965). For that reason more general literature stemming from the operations- and behavioral field is used to gain a deeper understanding of motivation and how to affect it. Abundant research has been conducted on motivation that resulted in many different theories. A widely used distinction of motivational sources is the one between intrinsic and extrinsic motivation (Ryan and Decy, 2000). Intrinsic motivation is driven by an enjoyment or interest in the work itself (Hackham and Oldham, 1975). It exists within a person rather than that it depends on external pressures or a desire for some form of reward (Buelens et al., 2006). Extrinsic motivation origins from outside the person and refers to the performance of an activity in order to attain an outcome (Buelens et al., 2006). This research limits itself to two key procedures that have shown to increase intrinsic and extrinsic motivation. These will be discussed below.

As mentioned, intrinsic motivation finds it origins in the actual work. Hence, in order to increase intrinsic motivation one should focus on procedures that are related to the characteristics of the work. Job enrichment is the main example of such a procedure. An influential model narrowly related to job enrichment is the job characteristics model developed by Hackham and Oldham (1975) that affect the intrinsic motivation of employees. This model contains five core job characteristics that have a major influence on the intrinsic motivation employees derive from a job. These are: skill variety, task identity, task significance, autonomy and feedback. As feedback has already been examined above, it will not be discussed any further. The definitions of the four remaining characteristics, as shown in Figure 3, are derived from Buelens et al., (2006). Job enrichment is a procedure that aims to (re)design a job such that a higher degree of the job characteristics is obtained. This results in a higher intrinsic motivation of the employees performing the job and subsequently a better performance. Evidence exists that job enrichment results in improved performance (e.g. Kopelman, 1985) Furthermore, job enrichment convincingly demonstrated to improve satisfaction, absenteeism and turnover rate (Fried and Farris, 1987). However, it has to be kept in mind that preferences regarding a job differs per individual; not everyone desires a highly enriched job (Buelens et al., 2006).

Job characteristic Definition

Skill variety ‘‘the extent to which the job requires an individual to perform a variety of tasks that require different abilities and skills’’

Task identity ‘‘the extent to which a job involves doing a complete from-beginning-to-end piece of work that has a visible outcome’’

Task significance ‘‘the extent to which a job affects, or is important to the lives of other people within or without the organization’’

Autonomy ‘‘the extent to which a job enables an employee to experience freedom and discretion in completing his or her job’’

Figure 3. Definitions of job characteristics, derived from Buelens et al., (2006)

7 3. METHODOLOGY

In this chapter the methodology of this research will be described, by answering the research question: ‘‘How can a Continuous Improvement framework be developed that can be applied within a visual inspection process?’’.

3.1 Framework for Continuous Improvement

Continuous Improvement (CI) includes various methods to improve the competitiveness of companies. Well-known examples are the PDCA and the DMAIC cycle. However, these methods are not necessarily effective at solving all issues (Bhuiyan and Baghel, 2005). This increasingly results in the development of CI methods that fit the specific needs of companies and processes (Bhuiyan and Baghel, 2005). This research also opts the development of a specific CI method that fits the needs of inspection processes. By assessing an inspection process on procedures that are specifically relevant to inspections in a structured way, it is expected to provide an overview that contains the most urgent points for improvement. The PDCA and DMAIC cycle, on the other hand, are more ad-hoc methods that are more likely to oversee urgent aspects. Furthermore, the questionnaire can be used as a baseline measurement to which future assessments can be compared. This allows the comparison of how an inspection process performs on a certain procedure over a longer period, and the determination of the effect of utilized improvement activities. As this questionnaire can be used over and over it enables continuous improvement. Also, the questionnaire can be updated if needed. For instance, questions and procedures can be added or removed in order to retain the fit with an inspection process.

3.2 Case Description

The case study will be performed at a factory that is located in the Netherlands. This factory is devoted to the painting of surface parts of final products of a certain brand in the automotive industry. These parts are subsequently transported to assembly factories of that brand. These assembly factories represent the customer of the case company. The case company has a fully automated painting process that is executed by robots according modern technology. Daily around 5.000 surface parts are painted. The number of painted parts per final product varies between 10 and 30.

All parts are placed on skids, i.e., a type of rackets that move automatically through all process steps. After the painting process has finished, the skids move through the Inspection and Repair (I&R) department. Here the parts are being visually inspected and, if necessary, repaired by inspectors. The I&R department has to ensure that all parts that flow to the assembly department are of sufficient quality. The assembly is the last process step within the case company and should not be confused with the assembly factories that the painting factory is supplying. This research will focus on the process within the I&R department. This process consists of four different lines that all contain the same three steps, as shown in Figure 4. At the first step, inspectors visually search the parts for anomalies. An anomaly is any aspect that is somehow deviating from what it is supposed to be. Then for each anomaly it is decided whether or not it is acceptable. In case an anomaly is acceptable no further action will be taken. When the anomaly is not acceptable it will be classified as a quality deviation and the inspector has to decide whether the quality deviation needs to be repaired or the part has to be discarded. In case repair is not possible or takes too much effort, the part will be discarded. The reparation is subsequently executed at the second step. At the third step a final inspection is conducted. Once more, the parts are being inspected and a final decision is made. The parts are either accepted, discarded or need additional repair. Each step is staffed by one inspectors. Further, each step needs to be executed within the fixed tact time of 255 seconds.

1. First Inspection 2. Reparation 3. Final Inspection Painting Process Assembly

Inspection & Repair Department

Discard Part

8 3.3 Data Collection

Different sources of data will be collected to test the framework. Naturally, the main data source is the questionnaire that was filled out by inspectors and managers during December 2013 and January 2014. The questionnaires were filled out by fourteen inspectors and two managers. The employees of the I&R department that are mainly involved with repair activities opposed to inspection, were excluded. Further primary data sources include numerous as well semi-structured as unstructured face-to-face interviews and informal talks with people that are involved in the inspection process between September 2013 and January 2014. These people include inspectors, managers, a process engineer and a quality expert. During or after these interviews and informal talks, notes were taken. Further, observations of the inspection process were made and as well a training session as a work meeting of inspectors was attended. Throughout the research, this data helped to provide direction and demarcation of the framework. In addition, secondary data was collected from an employee satisfaction report that was conducted in September 2013 by a research firm specialized in these reports. This report includes statements measured according a 5-point likert scale, ranging from strongly disagree to strongly agree. The use of different data sources is called triangulation (Yin, 1993) and remedies the specific shortcomings and biases of the instruments individually as they correct and complement each other in combination (Van Aken et al., 2012).

3.4 Questionnaire

The questionnaire consists of two parts: a part filled out by inspectors and a part filled out by managers of the inspection department. The part filled out by managers is directed at the policy regarding the procedures that are part of the framework. The part for inspectors aims to find out how inspectors experience that policy and to get insight in their opinion and experience regarding the inspection process. The questionnaires for the managers and inspectors were both filled out individually and in the presence of the researcher. Before the questionnaires were actually filled out, a pilot test was conducted to test the effectiveness of the questions. If necessary questions were revised, removed or new questions were included. The pilot test enhanced the quality of the questionnaire and avoided a possible loss of time due to an inadequate design (Karlsson et al., 2009). The researcher introduced all questions to the managers and inspectors and afterwards they were asked whether they understood the question. If not, further clarification was given. Consequently it was ensured that the questions were interpreted in the same way. Also, the researcher was able to ask further questions during the questionnaire in order to get a deeper understanding of the case. During and rights after the questionnaire was filled out, the researcher made notes of the additional comments.

The questionnaire for inspectors contains both open-ended questions and closed questions in the form of statements. The questionnaire for managers only contained open-ended questions. The statements, are measured according the 7-point Likert scale. A 7-point Likert scale is favored over a 5-point Likert scale because it provides a more granulated insights in the opinion of the respondents (Dawes, 2008). The statements are included because they obtain people’s position on certain issues (Karlsson et al., 2009) and reflect a relative quantity, i.e., a score (Hair et al., 1992). Moreover, they are more readily analyzed and interpreted than open-ended questions (Karlsson et al., 2009). Namely, the score resulting from the statements enables some form of statistical analysis which leads to a more objective ground to select certain improvement activities than open-ended questions do. To conclude, the statements are more suitable to be used as a baseline measurement to which future assessments can be compared. Overall, this approach combines the strengths from interviews, i.e., providing rich data (e.g., Karlsson et al., 2009), and surveys, i.e., which are known for its high reliability and accuracy (e.g., Miller, 1991).

The questionnaire, which can be found in Appendix A, is grounded in literature and the questions are either directly distracted from literature or in a less direct way. In case a question is directly distracted from literature, the reference of the article is given in the list of annotations that is also shown in Appendix A. If a question is not directly distracted from literature an explanation is given why the question is included and, if applicable, a supporting reference is given.

9 3.5 Data Analysis

All statements have a 7-point Likert, as shown in Figure 5. The options represent a value ranging from 1 to 7: strongly disagree equals 1; disagree equals 2; … and strongly agree equals 7. The values 1 to 7 represent respectively the values 0, 1.67, 3.33, 5, 6.67, 8.33 and 10. For each statement the score will be calculated by taking the average of all values that are filled out. This score will be expressed in a number on a 0-10 continuum, in which 0 equals strongly disagree and 10 equals strongly agree. In addition the standard deviation (SD) will be calculated. The SD shows how much variation from the mean exists. A low SD indicates that the answers tend to be close to the mean, whereas a high SD indicates the opposite. This can be helpful in understanding the results of the questionnaire, as it provides insights into the extent to which the respondents agree with each other. It must be noted that from statistical point of view the use of the SD for ordinal data is considered as inappropriate. Namely it is illegitimate to believe that the intensity between strongly disagree and disagree is equal to the intensity between the other consecutive answers (Cohen et al., 2000). However, the purpose of this research is not to provide a conclusive statistical approach, but rather use the SD as a tool to get a better understanding of the data.

4. RESULTS

The central question of this chapter is: ‘‘Is the developed framework effective to determine a set of improvement activities in a case study?’’. The chapter is structured according the sequence of steps of the framework. Hence, first the performance indicators are discussed, then the procedures that are included in the framework are described and finally a possible set of improvement activities is given.

4.1 Performance Indicators

Within the Inspection & Repair department (I&R) of the case company there are two main performance indicators. Both indicators are related to the accuracy of the inspection; they measure the number of quality deviations that are not detected. Thus in order to improve the performance of the inspection process, the number of non-detected quality deviations needs to decrease.

The first indicator is named the Quality Verification (QV) and measures the product quality of finished parts. The score of the QV reflects the number of quality deviations that are ascertained per final product. For instance, a QV score of 2 indicates that there are on average 2 quality deviations per final product. The QV score results from a daily sample that is taken by a quality expert within the case company. This sample is taken after all process steps have been completed and before the parts are transported to the assembly factory. Due to a sample size of one final product per day a reliable confidence interval of 95% is obtained. Also, the QV score represents a specific goal for the I&R department. The aim is to have a score below 1.25 quality deviations per final product. The I&R department has a direct effect on the QV score, as their job is to filter all parts with quality deviations and, thus, to avoid parts of insufficient quality reaching the customer. Each day the QV score is announced within the I&R department and is shortly reflected upon.

The second indicator is called the E-quality and origins from one of the assembly factories that the case company is supplying. When a quality deviation is detected on a part that is supplied by the case company, an E-quality report is created by the assembly factory and sent to the case company. This report describes the quality deviation that is found and include a picture of it. As well as with the QV score, the E-qualities are announced and reflected upon within the I&R department. Furthermore, the E-qualities are assigned to the inspectors that inspected the corresponding part. The managers keep track of the number of E-qualities per inspector and when the managers think an inspector is performing insufficient, they are appealed on it. There can be disagreement between the case company and one of the assembly factories on whether or not the E-quality is justified. Namely, it appears that the universal quality standard of the concerning automotive brand is not always maintained in the same way by all factories, which can lead to different interpretations. For that reason the E-quality is considered as a less objective performance indicator.

4.2 Procedures 4.2.1Training

10 the nature and cause of quality deviations. The on-the-job training consists of a period in which new inspectors work together with a trainer. This trainer is an experienced inspectors that teaches how to perform the inspection. A standard checklist is used containing all tasks that are required within the inspection, to ensure that all aspects are covered during the training. The first time the trainer demonstrates how to perform a task and the next time the trainee has to perform the task under supervision of the trainer. Overall, the trainees receive a step by step learn-by-doing training that takes as long as each individual needs to master all aspects of the inspection. As soon as the checklist is finished and the trainer thinks the trainee performs at a sufficient level, the trainee will start to do inspections individually and function as a ‘full inspector’.

In Table 1 the outcomes of the two statements answered by the inspectors are shown. With a rate of 7.97, statement 1 demonstrates that according to the inspectors the training program supports them sufficiently in acquiring knowledge and skills that are required to perform the inspection. Further, with a score of 8.21, the inspectors think they currently possess the knowledge and skills they need to perform the inspection. However, in the open-ended questions two inspectors mentioned that some form of re-training would be useful, as this is currently lacking. In addition, three inspectors indicated that insufficient time was taken for their on-the-job training, as they were ‘released’ too early.

Statement nr. TRAINING

Score

SD

1. The training program sufficiently supports the acquisition of knowledge and skills required to perform the inspection

7.97 1.43

2. You currently possess the knowledge and skills required to perform the inspection

8.21 0.43

Table 1. Outcome statements about training 4.2.2 Feedback

The feedback that is provided to inspectors consists of as well performance as cognitive feedback. Performance feedback has different forms. First, every day the number and type of e-qualities and Quality Verification scores are shortly discussed at group level and are hung on a bulletin. Second, inspectors receive individual feedback when too many e-qualities origin from their inspection line. This feedback makes clear that better performance is expected from the inspector. Third, inspectors give each other feedback. This occurs for instance when an inspector misses a quality deviation at the first inspection that is subsequently detected during the final inspection. In this case, fellow inspectors inform each-other about it. Cognitive feedback is normally only given by the managers and does not take place on a structural basis. It occurs when a manager walks by and sees that the inspection is executed insufficiently. Next to top-down feedback, a form of upward feedback exists. Namely, managers promote inspectors to think about ways to improve the process and to share it with them. However, this form of feedback hardly takes place.

From the open-ended question with regards to the opinion of inspectors about the feedback they receive, it is known that the inspectors are in general fine with it. However, four of the respondents mentioned that more positive feedback should be given by the managers. Currently, positive feedback is only incidentally provided. The employee satisfaction report also indicates that the presence of positive feedback is limited. Namely the statement ‘If I do it right, my manager will tell me’, as depicted in Table 6, was rated with 5.1. Moreover, this is significantly lower than the score of the other six departments of the case company. Statement 3, as presented in Table 2, shows it is common ground for the respondents to provide feedback to their fellow inspectors, as it is rated with 7.97. However, they are not strongly encouraged to provide each other feedback, i.e., a score of 5.83 with a SD of 1.97 on statement 4. The outcome of the fifth statement demonstrates that inspectors are to some extent encouraged to think along for process improvement. However, the SD of 1.81 suggests the opinions are somewhat diverging on that point.

Statement nr. FEEDBACK

Score

SD

3. You provide feedback to your fellow inspectors 7.97 1.12

4. You are encouraged to provide feedback to your fellow inspectors 5.83 1.97 5. You are encouraged to think along to improve the inspection process 6.55 1.87 Table 2. Outcome statements about feedback

11 Feedforward is given in a number of ways. First, during the classroom training, as discussed above, inspectors are informed with background knowledge regarding the product and knowledge about the nature and cause of quality deviations. Second, during the inspection inspectors make each other aware of specific quality deviations that occur a lot. Further, inexperienced inspectors gain knowledge by asking colleagues when they are curious about certain aspects regarding the product or quality deviations. Lastly, the I&R department developed a work standard that prescribes inspectors how to do the inspection. This standard tells for specific parts the locations where extra attention is required due to a higher likelihood of specific quality deviations.

Table 3 depicts the outcomes of the statements. First, the statement that sufficient feedforward is being provided is, with a rate of 7.14, considered as predominantly positive. However, the SD of 1.93 reveals that there is a considerable amount of disagreement about whether sufficient feedforward is provided among inspectors. Second, the statement that knowledge on the occurrence of quality deviations is sufficiently shared among inspectors could be answered confirmatory, as it scores 8.09. The eighth statement calls, with a score of 5.95, for a higher degree of attention. Moreover, the high SD of 2.50 proves that there is a large difference in the level of knowledge between the inspectors, regarding the product, the process and the cause and nature of quality deviations.

Statement nr. FEEDFORWARD

Score

SD

6. Sufficient feedforward is being provided on where on the product quality deviations can be expected, such that you can perform your task well

7.14 1.93

7. Knowledge on the occurrence of quality deviations is sufficiently shared among inspectors, such that you can perform your task well

8.09 1.06

8. You possess sufficient knowledge about the product, the process and the nature and cause of quality deviations, such that you can perform your task well

5.95 2.50

Table 3. Outcome statements about feedforward 4.2.4 Decision standard

The standard plays a very important role within the I&R department. The standard incorporates the global quality standard that is used throughout the automotive brand. The global quality standard is developed at the headquarters and is based on the demands that final customers have regarding the quality. It prescribes in detail the quality level each part has to comply to. Hence, it forms the ground for deciding either to accept, repair or discard a part. The global quality standard is more or less imposed by the headquarter to the case company and is updated only once in a couple of years. The standard that is used within the I&R department is based on the global quality standard, but additionally tells in detail how to perform the inspection. It includes the sequence of all acts, the distance from which has to be searched, the search strategy and the possible use of reflection and illumination. This additional part of the standard is developed by the managers of the I&R department. The managers are also responsible for updating the standard, which occurs in an ad-hoc way. Based on changed circumstances, for instance the introduction of new parts, and improvement suggestions of inspectors the standard gets updated. All questions and statements are about the standard as used within the I&R department, which includes the global quality standard and the part that prescribes how to perform the inspection.

The 9th statement is rated with 7.86, which means that the standard is a workable and adequate protocol according to the respondents. Furthermore, the score of 7.14 on the 10th statement reflects that, be it with a SD of 1.83, it is largely possible to make an unambiguous accept/repair/discard decision. Still, one of the managers stated: ‘‘there sometimes remains a grey area regarding the decision as all parts are in some sense unique.’’ Further, in the open-ended questions 5 out of 15 respondents stated that inspectors need to be more aligned regarding the accept/repair/discard decision. Two respondents even said: ‘‘all inspectors have their own standard’. Also, the rate of 3.57 on statement 11 confirms that inspectors do not comply to the decision standard in the same way. In contrast, statement 12 shows that according to the inspectors themselves they comply to the work standard, as it is rated with 7.86. Further, during the conversations with inspectors and managers it appeared that inspectors over time develop their own habits, which means they deviate from the standard on some points.

12 9. The decision standard provides an adequate and workable protocol to

perform the inspection

7.86 0.98

10. Based on the decision standard you can make an unambiguous accept/repair/discard decision

7.14 1.83

11. All inspectors comply to the decision standard in the same way 3.57 1.87

12. You comply to the decision standard 7.26 1.73

Table 4. Outcome statements about the decision standard 4.2.5 Motivation

Before continuing to discuss the results around the procedures to increase the motivation, i.e., job enrichment and incentives, motivation itself will be discussed. Statement 13, regarding the motivation of inspectors, was rated 8.50 which means that they are highly motivated. However, from conversations with inspectors and managers it appeared that there are differences between inspectors in the will to do their job well and the effort that is related to it. A way to observe this is the extent to which the time available to do an inspection, i.e., the tact time of 255 seconds is utilized. Five inspectors mentioned that some of their colleagues do the inspection in such a short time that makes them wonder whether they did their inspection in a good way. One respondent mentioned: ‘‘I am a quick inspector and I perform quite good. However, if I would use the entire time available to do the inspection, i.e., the tact time, I will probably detect more quality deviations and there would be less e-qualities.’’ Further, from the employee satisfaction report, as depicted in Figure 9, it is known that the I&R department has, with a rate of 6.3, a significantly lower employee satisfaction than the remaining six departments in the case company. It can be safely assumed that this is somehow also related to motivation.

Statement nr. MOTIVATION

Score

SD

13. You are motivated to do your work well 8.45 0.99

14. You want to have more variety in your work 7.38 2.33

15. You want to have more responsibility in your work 7.73 1.73

Table 5. Outcome statements about motivation and job enrichment 4.2.6 Job enrichment

Roughly three different positions exist within the I&R department. As shown in Figure 4, first the parts are inspected, then it flows through to a repair station, and lastly a final inspection is carried out. Not all inspectors are able to work at all of the positions and not all of them perform well at the three different positions either. The policy of the department is to teach as many as different positions as possible to the inspectors, which makes it easier to make a schedule for the different positions. In addition, this offers more skill variety to the inspectors. Further, the department aims to switch inspectors from position after each break, i.e., three times a day, such that the variety increases and to make it easier for inspectors to maintain concentration throughout the day. As the work within the I&R department is cut into three different positions, the inspectors perform just a tiny task of the whole production process. Further, a visible outcome of the work that is done is lacking, except for the reparation position. As a result, the task identity of the work within the I&R department can be considered as low.

With a score of 7.38 and 7.73, statements 14 and 15 reveal that the respondents prefer to have respectively more variety and responsibility in their work. However, the SD’s of respectively 2.33 and 1.73, show, that there is no consensus on these points. The employee satisfaction report confirms the desire for personal development, as the 3th statement was rated with 8.1. However, the report demonstrates that the inspectors are, with a score of 5.2 significantly less satisfied about the possibilities to develop themselves than their fellow colleagues of the other departments within the case company. The same holds for the possibility to acquire sufficient new knowledge. Further, the statement that a challenging work environment is provided within the case company was rated with 6.4, as can be seen in Table 6.

4.2.7 Incentives

13 of this department are reluctant regarding goal setting and incentives, especially at individual level. First, the inspection performance lies partly outside the influence of inspector. This deserves some explanation. To start, the performance indicators of the inspection process focus on the number of non-detected quality deviations. However, the number of quality deviations depends on the quality of the painting process which lies entirely outside the influence of inspectors. Yet, it does have an influence on the inspection performance. Namely, it can be safely assumed that the higher the number of quality deviations located on a part, the higher the chance quality deviations remain non-detected and, hence, the worse the inspection performance. Second, incentives based on the number of non-detected quality deviations can negatively affect the behavior of inspectors. Namely, inspectors can prevent non-detected quality deviations from occurring by simply discarding the parts about which they doubt to accept or reject it. The ill effect of this is additional costs and possible delivery problems. Due to the difficulties with the use of incentives any further practice at this point seems infeasible for the case company.

Statement

Nr.

I&R dep.

score

Company

score

1.

If I do it right, my manager will tell me

2.

I am, overlooking all, satisfied about working at the case

company

6.3* 7.2

3.

I want to develop myself within my discipline

4.

I am satisfied about the possibilities to develop myself within

the case company

5.2* 6.3

5.

In my work I can acquire sufficient new knowledge

6.

The case company offers me a challenging work

environment

6.4 6.9

* Average score of the Inspection & Repair department is significantly lower than the average score of the company as a whole.

Table 6. Results derived from the employee satisfaction report 4.3 Selection of Improvement Activities

From the assessment of the inspection process of the case company that was performed under heading 4.2, various aspects occur that could be improved. Rather than describing all possible improvement activities it is opted to select a top-three of activities, resulting in a more concise overview. These three activities are selected based on the fact that they seem to offer the most potential for improvement.

The first improvement activity is related to job enrichment. It is clear that in general inspectors have the ambition to further develop themselves. However, the possibilities for personal development do not fully match their ambitions. Job enrichment offers possibilities for personal development of inspectors and improves the intrinsic motivation. Job enrichment consists of different aspects: feedback, skill variety, task identity, task significance and autonomy. The work needs to be redesigned in a way that these aspects are positively affected. An option is to integrate the first two steps of the inspection process, i.e., first inspection and reparation, into a single step that is performed by one inspector. In this way as well the skill variety as the task identity increases. However, there can be many other ways to apply job enrichment. When applying job enrichment it is important to take into account the wishes of the individual inspectors and the constraints of the system.

The second improvement activity is about the decision standard. It appeared that there are considerable differences between inspectors in the way they use the standard. By reducing the differences among inspectors, a more consistent quality level is expected to be obtained. First, it needs to be determined how the standard is currently interpreted and used by the different inspectors. A way to do this is by letting all inspectors inspect the same parts independent of each other to see whether there are differences. In this way it gets clear which inspectors deviate from the standard. Also, it should be determined why inspectors deviate from the standard. If this is caused by a lack of knowledge, the inspectors in case can be assigned to get re-training regarding the standard. In case it is due to different insights, or personal habits that evolved, inspectors can be pointed to the standard. This activity should be repeated on a regular basis in order to keep the different inspectors, as much as possible, on the same line.

14 that there is room for improvement regarding the level of knowledge on this field. Yet, the high SD on the ninth statement shows the need for this knowledge differs among the inspectors. Therefore, the inspectors that need to obtain more knowledge, need to be identified. Thereafter, the level of knowledge of these inspectors can be improved by, for instance, teaching from an expert. With this knowledge inspectors know better how to look for defects and, thus, to perform the inspections more effective.

5. DISCUSSION & CONCLUSION

In this research a framework for the improvement of inspection processes was developed and tested. Based on a questionnaire that is grounded in literature the functioning of the visual inspection process of the case company was assessed. The questionnaire provided insights in six procedures that appeared to be of importance for visual inspections: training, feedback, feedforward, decision standard, job enrichment and incentives. Strengths and weaknesses of the visual inspection of the case company regarding these procedures were exposed and avenues for improvement popped up. Up to this point, the framework turned out to be a useful and working tool for inspection improvement. However, due to time restrictions the improvement activities are still to be implemented. Hence, an improved performance is not achieved yet and therefore it is not possible to conclude whether or not the framework is effective. In that sense, this research serves as a first attempt to develop and test a framework, on which can be continued in future research. Despite the fact that the framework is not entirely tested, some notes and initial insights can be drawn with regards to theory, practice and the case company. Also future research suggestions are made and the limitations of this study are discussed.

5.1 Theoretical Implications

It can be concluded that the framework is successful in providing insight in the current functioning of an inspection process regarding each of the selected procedures and in identifying improvement possibilities. Also, the integration of multiple procedures into a single framework turned out well and resulted in a concise overview of recommended improvement activities. For the conciseness of this research, the framework was demarked such that it resulted in a limited number of procedures. Due to the fact that the framework is not completely applied and in just one case study, it is premature to judge whether the most important procedures are included. Yet, the framework can be easily adapted. For instance, other procedures can be added either resulting from the overview of Megaw (1979) or from other sources.

In previous research, considerable difference in performance between inspectors have been reported (e.g. studies Wiener, 1975; Gallwey, 1982; Wang and Drury, 1989). Despite the fact that many procedures tried to encounter this, these differences persist (Jiang et al., 2004). In this research a first attempt was made to link different levels of performance to different levels of motivation among inspectors. The case study indicates different levels of motivation between inspectors, which is amongst others expressed in the extent to what inspectors utilize the available tact time to perform the inspection. However, it was not measured whether the low-motivated inspectors actually perform worse than the high-motivated inspectors. Therefore one should be cautious in drawing conclusions on the relation between motivation and inspection performance. In this research, the application of job enrichment is recommended in order to increase the intrinsic motivation of inspectors. Job enrichment is a response to the wish for ‘richer’ work that appeared within the case company. It was found that inspectors want to have more variety, responsibility and possibilities for personal development in their work. This is in line with literature, where inspections are described as monotonous jobs that can result in boredom (O’Hanlon, 1981). Again, one should be cautious in drawing conclusions regarding the benefits of job enrichment for inspection processes, as it remains to be investigated. Still, based on the findings of this research there seems to be sufficient ground to expect that job enrichment positively affects inspection performance.

15 and, subsequently, to set goals limits managers in using incentives to increase the motivation of inspectors.

5.2 Practical Implications

The application of the framework to the inspection process of the case company gave an overview of the functioning of the process. Regarding each of the procedures several aspects became apparent that could be improved. A top-three of improvement activities is described that are expected to have the most impact. The first activity is the application of job enrichment. Job enrichment makes the work more interesting which results in a higher intrinsic motivation and is subsequently expected to lead to improved performance. The second activity is related to the decision standard. By means of an experiment it can be determined how the standard is interpreted and used by the different inspectors. Based on that, inspectors can be assigned to be re-trained in order to reduce the differences between inspectors. The third activity aims to increase the level of knowledge of inspectors regarding the product, the process and the nature and cause of quality deviations by means of teaching from an expert. This knowledge enables the inspectors to perform the inspections more effective.

The framework is designed such that it can be applied to any inspection process, in order to enhance the generalizability. Whereas the improvement activities are specified to the case company, the questionnaire can be applied to any inspection process. It might be that some procedures or questions are not relevant in some cases or that additional procedures or questions are required. This is inevitable as each inspection process has its specific characteristics which cannot be completely covered by the questionnaire. In these cases the questionnaire can be adapted such that it fits the specific needs of an inspection process. Overall, it has to be kept in mind that the questionnaire does not offer a conclusive approach that results in an undisputable set of improvement activities. Rather it provides a direction for the selection of improvement activities.

5.3 Limitations and Future Research Suggestions

A main limitation of this research is the fact that the framework has only been tested partially and therefore its effectiveness cannot be completely determined. Further, the framework is only applied in one case which results in a low generalizability of the findings. Another limitation is that after the assessment of an inspection process the choice to select certain improvement activities remains somewhat subjective. In future research the first two limitations can be overcome, by, respectively, fully testing the framework and studying its effect on the performance indicators and applying the framework in other cases and evaluate whether it actually improves the performance. Next to that, the relation between motivation and inspection performance and the effect of job enrichment on inspection performance can be further investigated.

5.4 Acknowledgments

16 REFERENCES

Anupindi, R., Chopra, S., Deshmukh, S. D., Van Mieghem, J. A., & Zemel, E. (2011). Managing business process flows. Pearson/Prentice Hall.

Arani, T., Karwan, M. H., Drury, C. G. (1984). A variable memory model of visual search. Human Factors, 26(6), 631-639.

Bannister, B. D. (1986). Performance outcome feedback and attributional feedback: Interactive effects on recipient responses.

Baudet, N., Maire, J. L., Pillet, M. (2013). The visual inspection of product surfaces. Food Quality and Preferences. 27, 153-160.

Berger, A. (1997). Continuous improvement and : standardization and organizational designs, Integrated Manufacturing Systems, 8(2), 10 – 117.

Bernardin, H. J., Dahmus, S. A., & Redmon, G. (1993). Attitudes of first‐line supervisors toward subordinate appraisals. Human Resource Management, 32(2‐3), 315-324.

Bessant, J., Caffyn, S., Gilbert, J., Harding, R., & Webb, S. (1994). Rediscovering continuous improvement. Technovation, 14(1), 17-29.

Bhuiyan, N., & Baghel, A. (2005). An overview of continuous improvement: from the past to the present. Management Decision, 43(5), 761-771.

Buelens, M., Sinding, K., Waldstrøm, C., Kreitner, R., & Kinecki, A. (2006).Organisational behaviour. McGraw-Hill Higher Education.

Burkert, M., Fischer, F. M., & Schäffer, U. (2011). Application of the controllability principle and managerial performance: The role of role perceptions. Management Accounting Research, 22(3), 143-159.

Chabukswar, S., Gramopadhye, A. K., Melloy, B. J., Grimes, L. W. (2003). Use of aiding and feedback in improving visual search performance for an inspection task. Human Factors and Ergonomics in Manufacturing & Service Industries, 13(2), 115-136.

Cohen, L., & Manion, L. (2000). Research methods in education. K. Morrison (Ed.). Taylor & Francis. Dawes, J. (2008). Do data characteristics change according to the number of scale points used? An experiment using 5 point, 7 point and 10 point scales. International Journal of Market Research, 51(1). Debrosse, T., Pillet, M., Maire, J. L., Baudet, N. (2010). Sensory perception of surfaces

quality-Industrial practices and prospects. THE PROCEEDINGS OF THE KANSEI ENGINEERING AND EMOTION RESEARCH.

Deci, E. L., & Ryan, R. M. (2000). The" what" and" why" of goal pursuits: Human needs and the self-determination of behavior. Psychological inquiry,11(4), 227-268.

Drury, C. G., (1975). Inspection of sheet materials—model and data. Human Factors, 17, 257–265. Drury, C. G. and Gramopadhye, A. K. (1992). Training for Visual Inspection: Controlled Studies and Field Implication, Proceedings of the Seventh Federal Aviation Administration Meeting on Human Factor Issues in Aircraft Maintenance and Inspection: Science Technology, and Management: A Program Review, 135-146.

17 Ebrahimi, M., Sadeghi, M. (2013). Quality management and performance: An annotated

review. International Journal of Production Research, 51(18), 5625-5643.

Fried, Y., & Ferris, G. R. (1987). The validity of the job characteristics model: A review and meta‐analysis. Personnel Psychology, 40(2), 287-322.

Fryer, K. J., Antony, J., & Douglas, A. (2007). Critical success factors of continuous improvement in the public sector: a literature review and some key findings. The TQM Magazine, 19(5), 497-517. Gallwey, T. J. (1982). Selection test for visual inspection on a multiple fault type task,

Ergonomics, 25, 1077–1092.

Gramopadhye, A., Bhagwat, S., Kimbler, D., Greenstein, J. (1998). The use of advanced technology for visual inspection training. Applied Ergonomics, 29(5), 361-375.

Gramopadhye, A. K., Bowling, S. R., Khasawneh, M. T., Kaewkuekool, S., Melloy, B. J., Duchowski, A., ... & Cournia, N. (2002, April). Using Training to Improve Performance of Inspectors on the Hangar Floor. In Proceedings of the 16th Human Factors in Aviation Maintenance Symposium, April (pp. 2-4). Gramopadhye, A.K., & Drury, C.G. (1991). Speed and accuracy in aircraft inspection. Galaxy Scientific Corporation, 1993. Human Factors in Aircraft Maintenance—Phase Two Program Report

(DOT/FAA/AM-93/5). National Technical Information Service, Springfield, VA.

Gramopadhye, A. K., Drury, C. G., Jiang, X., & Sreenivasan, R. (2002). Visual search and visual lobe size: can training on one affect the other?. International Journal of Industrial Ergonomics, 30(3), 181-195.

Gramopadhye, A. K., Drury, C. G., Prabhu P. V. (1997a). Training strategies for visual inspection. Human Factors and Ergonomics in Manufacturing 7(3), 171–196.

Gramopadhye, A. K., Drury, C. G., Sharit, J. (1997b). Feedback strategies for visual search in airframe structural inspection. International Journal of Industrial Ergonomics, 19, 333–344.

Hackman, J. R., Oldham, G., Janson, R., & Purdy, K. (1975). A new strategy for job enrichment. California Management Review, 17(4), 57-71.

Hair, Jr J. F., Anderson, R. E., Tatham, R. L. and Black, W. C. (1992). Multivariate data analysis, New York: Maxwell MacMillan.

Harris, D. H., and Chancy, F.B. (1969). Human Factors in Quality Assurance. New York: Wiley. Ishikawa, K. (1990). Introduction to quality control. Productivity Press.

Jiang, X., Gramopadhye*, A. K., & Melloy, B. J. (2004). Theoretical issues in the design of visual inspection systems. Theoretical Issues in Ergonomics Science, 5(3), 232-247.

Jiang, X., Gramopadhye, A. K., Melloy, B. J., & Grimes, L. W. (2003). Evaluation of best system performance: human, automated, and hybrid inspection systems. Human Factors and Ergonomics in Manufacturing & Service Industries, 13(2), 137-152.

Johnston, D.M., 1965. Search performance as a function of peripheral acuity. Human Factors 7, 528– 535.

Juran, J. M., Blanton Godfrey, A. (2000). Juran’s quality handbook. McGraw-Hill Professional. Karlsson, C. (Ed.). (2008). Researching operations management. Routledge.