Does the quality of performance measurement indicator affect employee self-view?

(1)

1

Amsterdam Business School

Does the quality of performance measurement indicator

affect employee self-view?

Name: Yuxin Yan

Student number: 11051450 Thesis supervisor: Qi Yang Date: 17 June, 2016

Word count: 12796

MSc Accountancy & Control, Accountancy

(2)

2

Statement of Originality

This document is written by student Yuxin Yan who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

This paper investigates on the effect of performance measurement indicator (PMI) quality on employee self-view, and how employee educational level influences this relationship. Prior researchers define PM quality as the extent to which employees find the performance measures sensitive to their actions, precise in measuring relevant aspects of their performance, and verifiable (Groen, 2012; Moers, 2006). So I test my research questions from these three aspects. I find that when precision and verifiability of PMI quality is high the correctness of employee self-view is low, while when the sensitivity of PMI quality is high the correctness of employee self-view does not have obvious change. What’s more, statistical analysis results show a significant negative relation between PMI sensitivity and employee self-view correctness.

(4)

4

1 Introduction

The focus of this paper is the effect of performance measurement indicator (PMI) quality on employee self-view. Recent years have witnessed an increase in articles about performance measures ( Gitelman et al, 2013; McCance et al, 2011; moers, 2006; Groen, 2012). The definition of performance measures is “everything used to measure the job performance of employees” (Groen, 2012). Prior researchers define PM quality as the extent to which employees find the performance measures sensitive to their actions, precise in measuring relevant aspects of their performance, and verifiable (Groen, 2012; Moers, 2006). Some of the recent papers focus on the method or process in developing PM in certain field. For example, Gitelman et al (2013) investigate road safety performance measures on trauma management. Some other papers focus on identifying PM which exist but have not been recognized. For example, McCance et al. (2011) researches on potential PMs in nursing and midwifery practice. It also investigates on the relationships between these PMs. However, few papers research on the effect of PM when it is used within an organization. In other words, there is a literature gap on the effect of PM (Can high-quality PM help organizations in their daily operation? To what extend can PM help?).

Researches in terms of self-view are mainly within the field of psychology. For example, Critcher and Dunning (2009) investigate on how self-views influence self-assessments of a task performance. They find that self-views influence performance evaluations by first shaping perceptions of bottom-up experiences with the task, which in turn inform performance evaluations. Emery et al (2011) discuss how self-view influence people’s performance on leadership emergence and how the process of leadership emergence influences an individual’s self-view as a leader over time. However, no research links performance measurement indicators (a corporate management concept) with employee self-view (a psychology concept).

(6)

6 This research also uses employee educational level as a control variable. As state by Westbrook and Veale (2001), the greater one's education level, the more likely he or she will receive additional training. In another words, employee has higher educational level is more willing to learn further for his work and can also learn better from his work. As a result, I predict that educational level is a possible influence factor on how employee reacts to PM with different qualities.

In a PM system, high quality indicator can transfer the aims and goals of organization and manager to employees successfully. This can be achieved by the communication role of PM. In another words, operating PM with high quality indicators efficiently can facilitate internal communication within organization, especially the communication between employees and their managers. For example, if a manager setting the PMIs for his employee wants to improve customer satisfaction, he will list customer satisfaction as one or several important items of PMIs. After an employee sees these PMIs, even though he did not realize what the manager wants in the past, he would know that good customer satisfaction is expected by his manager.

This establishment of a formal communication method will lead to several advantages. Recent researches in this field found a strong positive relationship between international communication and job satisfaction (Nikolic et al, 2013). Also, Nikolic (2012) recognized a strong positive relationship between internal communication and strategic and economic business effects. While this paper investigates on whether a well-established communication tool (PM) has impact on the correctness of employees self-view.

It is found not only in working practice but also in academic researches that, despite spending more time with themselves than with any other persons, people often have surprisingly poor insight into their skills and abilities (Critcher and Dunning, 2009). This means that employees usually do not have clear self-views about their skill levels or working abilities. According to Critcher and Dunning (2009), this incorrect self-view is caused by chronic self-views people have about their abilities and whether they are skilled

(7)

7 or unskilled at a task. While performing their daily tasks, employees would more or less rely on the views about themselves. However, although these views provide a modicum of validity to performance estimates, relying on them too much can lead to bias and error (Critcher and Dunning, 2009). For example, if Alex thinks he is good at one task and also much better than others, he may ignore others’ suggestions on this task, even when these suggestions are actually sound and reasonable. Even worse, he may also contradict his manager on this task because of his deep confidence, even it is reasonable for most people that manager has a better view about a task and employees should follow the requirement of manager. As a result, incorrect self-view of employee can lead to inefficient corporate operation.

The aim of this paper is to examine the relationship between quality of performance measures indicators (PM) and employee self-view correctness. In other words, this paper discusses about to what extent the quality of PMI, as an internal communication tool in organization, impact on the correctness of employees self-view. My research questions are: 1) does PMI quality affects the correctness of employee self-view? 2) If yes, to what extant does PMI quality affect the correctness of employee self-view? 3) How does employee education level impact on the relationship between PMI quality and the correctness of employee self-review?

This investigation has importance on organization practice because of three reasons. First, if employee knows his strong and weak points clearly and knows in which aspect his manager wants him to improve, the employee can improve himself according to his manager’s requirement and be more suitable to his job position. In turn, he can locate himself correctly in his department and perform more appropriate. Secondly for managers, employees perform better will make his job goes more smoothly. Third, as a PM system affects the whole organization but not one individual employee, if it has positive effect on employees self-views, the efficiency of the whole organization will be improved. If the result of this research turns out that PM system does affect employee

(8)

8 self-view, then it suggests manager a way to facilitate the correctness of employee self-view and adds a reason of the importance of PM. If the result of this research turns out that PMI system does affect employee self-view, then it suggests manager a way to facilitate the correctness of employee self-view and adds a reason of the importance of PM.

This research will also contribute to prior literature. Due to the fact that there are few researches on the effect of PM, this paper contributes literature by adding knowledge about the effect of PM. As to the psychology field of self-view, this research links it with corporate governance concept and exams how it can be influenced in a certain condition (organizations).

Data of this paper is from a survey held by professor Groen in University van Amsterdam. This database originally contains 86 pairs of employees and their managers who answered several questions. Every student joint in should find at least 6 respondent pairs so this database will be extended in the future. In addition, as the students joint in are from different countries the respondent pairs are not restricted in certain types of organization, certain country or certain culture background. As a result, the finding of this paper will be a general one. There are several items contained in professor Groen’s questionnaire, I picked some of the questions to investigate my research questions. The quality of PMI is concluded from the judgments from managers and employees to organization’s PM system, and the correctness of employee self-view is from the comparison between employee self-evaluation and manager’s evaluation to the employee. The rest of this paper contains 4 chapters. Chapter 2 provides a literature review and establishes the hypotheses of this paper. Chapter 3 provides a more detailed description of data and methodology. Chapter 4 is the findings and results of this research and Chapter 5 gives a conclusion.

(9)

9

2 Theory and Hypothesis

This paper discusses how the quality of PM affects the correctness of employee self-view and how employee educational level influences this effect. So in this part, I will illustrate related definitions and research about PM, self-view and educational level. As PMIs have strong communication role in organizations, I will also discuss theories on internal communication. Then following the theories, the four hypotheses of this research will be stated.

2.1 Performance measurement indicators

According to Groen (2012), performance measurement (PM) is traditionally used from a top-management perspective to align the interests of employees/managers with those of the shareholders. It is defined as “everything used to measure the job performance of employees” (Groen, 2012). This paper focuses on performance measurement indicators (PMI). According to professor Groen’s survey, the meaning of PMI is everything that is used to measure employee’s performance, which can be both individual and group indicators. Examples of PMIs include client satisfaction, efficiency, amount of work done in a certain amount of time, quality. As I mentioned before, most previous studies have focused on investigating the proper PMI in a certain field or a certain condition and how to establish these PMIs in practice (McCance et al, 2011; Gitelman et al, 2013). In his research in 2006, Moers defines PM quality as the extent to which employees find the performance measures sensitive to their actions, precise in measuring relevant aspects of their performance, and verifiable. The reason of using Moers’s definition is that it emphasizes on the sensitivity, precision and verifiability of PM, as performance measures only correctly reflect employee performance if they have good measurement properties (Groen, 2012).

When scoring the quality of PMI, managers look at the verifiability of the output of PMI and the precision in measuring relevant aspects of employees’ performance, while

(10)

10 employees usually also want the PMI sensitive to their efforts (Groen, 2012; Moers, 2006). As a result, the survey answer about PMI quality from employees and managers can be used to measure different aspects of PMI. If the grade of PI quality is high (closer to 7), the PI evaluated is in a high quality.

One question about the sensitivity of PM quality is asked to employees. Employees need to score the question from 1 to 7. A score 1 means a strong disagree of this sentence and a score 7 means a strong agree of this sentence. The question is about how sensitive PMI is to employee’s performance, and the mainly concern is whether PMI result is also influenced by indicators out of employee’s control. This measurement of PMI sensitivity is also used in Moers (2006).

Two questions about precision and verifiability of PMI quality are asked to managers. Same as the question about PMI sensitivity to employees, managers also need to score these two questions from 1 to 7. A score 1 means a strong disagree of this sentence and a score 7 means a strong agree of this sentence. One of these two questions is about the precision of PMI. It concerns about whether PMI result is also influenced by indicators out of employee’s control. If the PMI result is not influenced by other indicators, this means it can measure employee performance precisely (Moers, 2006). The other one asks directly if PMI is objective and verifiable.

These aspects of PM quality are related to the communication through PMI. If PMIs are sensitive and precise to employee performance the result PM can tell employee how good or bad their performance is and correct their self-view. The verifiability of PM can make the result of PMI more reliable to employees. Due to the communication role of PMI, I also want to introduce internal communication.

The internal communication role of PMI can explain why PMI quality has influences on employee self-view. Vercic et al define internal communication in their 2012 research as all forms of communication within the organization. PMI can be seen as an internal communication tool, as PMI has a role to transfer the aims and goals of the company (or

(11)

11 who has strong influence on indicators) to employees (who are evaluated by these indicators). As a result, if the PMIs are clear and in a high quality, employees can know better about the aims and goals of the company. As employees’ performances (which influence their bonuses, improvements and salaries) are measured by these indicators, it is reasonable that they will think of these indicators carefully and try to improve themselves according to these indicators. In this process, they compare themselves with managers’ expectations and generating a better self-view.

Internal communication is more and more discussed by researchers in these years. Johansen (2012) investigated on communication in crisis. Lies (2012) recognized the importance of internal communication in change process. This communication takes place between leaders, managers and employees. Internal communication within organization influences the efficiency of this organization greatly. Smooth and effective internal communication facilitates the transfer of information in organization and in turn, assists organization to achieve its goal. For example of PMI, high quality of PMI transfers goals of organization to lower level employees so the employees improve their performance relating to these goals.

The methods of internal communication can be divided to formal and informal internal communication methods. Informal communication method includes short massages, tale bearing and other forms. Formal communication method includes meetings, reports, performance measures, training and others. PMI as a type of performance measures is a kind of formal communication method.

Nikolic et al (2013) investigate on the role of internal communication on job satisfaction. They find that employees’ satisfaction with payment, promotion, supervision, contingent rewards, co-workers and nature of work are related to communication with supervisors about stuffs like organizational perspectives. The result shows that communication has influence on employees’ attitude towards their work. PMI as an internal communication tool can communicate manager’s goals and standards to employees directly. From this

(12)

12 version, it is sound to predict that PMI quality has influence on employee’s attitude. Ruck and Welch (2012) investigate the relationship between communication within organizations and performance. The result shows a positive impact of internal communication on performance. Based on the empirical data, 25 per cent of employees indicate that their manager rarely or never makes them feel their work counts. Besides, they recognized the lack of researches on employee communication. This research indicates not only the possibility of PMI’s influence on employee performance, but also the importance of good quality PMI, which can remedy the lack of communication between employees and managers.

If an organization uses low quality PMI, which does not match to its goals, to measure employee performance, there are two possible results. First, employees get wrong direction on their work. When they are trying to get higher score on performance measurement, they actually push the organization away from its core goal. Even they improve their performances according to the PMI, get a higher score and match the PMI better, their self-view are not correct as the low quality PMI gives a wrong direction to them. Second, they realize the PMI is in low quality and ignore it. The organization losses the chance to communicate to employees or wastes time and money on a useless test. In this situation, PMI has nothing to do with employee self-view as no one sees it important.

2.3 Self-view

Previous researches defined self-view (self-esteem) as how people evaluate their performance on specific tasks (Critcher and Dunning, 2009). In the past, researchers already found that self-view is not something stay the same forever. On the contrary, it can be influenced by many factors in the environment (Redeker et al, 2011). Redeker et al (2011) found that mimicking others can change ones self-view and that mimickers always define themselves more in relation to others than non-mimickers. Considering the

(13)

13 characteristic of self-view that easy to be influenced, this study investigates on how PM quality influences employee self-view.

Researchers also find that self-view is a key influence factor of people performance. For example, one’s self-view as a leader can influence his possibility of getting authority and becoming leader (Emery et al, 2011). Critcher and Dunning (2009) find that one’s self-view can influence or mislead his/her self-assessment of task performance. These researches give meaning to this study that correcting employee self-view can improve employee performance and in turn have good influence on company as a whole.

In my study, employee self-view is defined as to what extent employee’s self-evaluation corresponds to manager’s evaluation to this employee. In other words, for points an employee think is his strong point, does his manager has the same feeling. This is tested by asking the same questions about employee performance to managers and employees and comparing their answers. A higher similarity between the two answers means a higher level of employee self-view correctness. This method is based on the assumption that employees work to fulfill their managers’ requirements and that managers as standing in the higher level in the organization can see their employee clearly.

Two questions are asked to employees and managers about employees’ performance. The questions asked to employee and their pair managers are in nearly the same way. One of the two pairs of questions is about whether the employee do what he is required to do and the other questions is about whether the employee do his job in required quality(Zhang and Bartol, 2010). These two questions together provide information about whether employee does what he is expected to do in an expected quality. By asking the same questions to employees and their pair managers, I get employee self-view and manager’s evaluation to employee on the two aspects.

As stated above, I evaluate the employee self-view correctness from two aspects. The difference between these two aspects of questions is that the first pair of questions

(14)

14 evaluates how correct employees understand their working scopes, while the second pair of questions asks whether employees do the work in required quality.

The reason of my prediction that the quality of PMIs has positive impact on employee self-view is that PI has a role to transfer the aims and goals of the company (or who has strong influence on indicators) to employees (who are evaluated by these indicators). As a result, if the indicators are clear and in a high quality, employees can know better about the aims and goals of the company. And as the PMIs are used to determine employee bonus, improvement and salary, employees are expected to think of the PMIs carefully, compare their current performance with the PMIs and adjust themselves according to the PMIs. The correctness of employee self-view will be improved during this process.

2.4 educational levels

In this study, I choose employee educational level as a moderating variable for the relationship between PMI quality and employee self-view correctness. According to Westbrook and Veale (2011) research, employees with higher educational level are more likely to accept current and preferred work-related education. Their results tell us that employees having higher educational level are more willing to learn and are more able to learn. So I predict that employee within higher educational level are also more sensitive to the implied requirement in PMI and are able to adjust them in line with these requirements better.

2.5 The relationship between PMI quality and employee self-view correctness

As stated above, I predict PMI quality will influence the correctness of employee self-view. And the reason is that a good PMI system can tell employees what they are expected to achieve and what behave they are expected to perform. The information expressed by PMI will give employee a clearer target of their work and daily performance. A good PMI system also includes a high quality measuring process. If employee can be viewed fairly and correctively by the high quality PMIs, they will know how much they

(15)

15 can match with the targets and expected performances. This process will finally improve the correctness of employee self-view.

I also predict that employee educational level has positive effect on the relation between PMI quality and employee self-view correctness. The reason is that employees with a higher educational level are predicted to be more willing to and able to learn. This is also stated in Westbrook and Veale (2011) that, comparing with others, employees with higher educational level are more likely to accept current and preferred work-related education. In the case of PMI system, knowing manager’s expectations from PMIs, considering at which point he needs to improve and trying to adjust behavior to match the expectations is similar to the process of learning. As a result, I make the prediction that employee with higher educational level can understand the hidden information of PMI better and knowing their shortcomings better, and in turn, have a more correct self-view.

In my research, I will test if the relation between PMI quality and employee self-view correctness and the effect of employee educational level by scoring the two variables, dividing data into different groups and comparing the scores in each group to see if the relation exists. To find whether the relation is in a liner basis, I will also do liner regression analysis.

2.6 Hypotheses

This research investigates on two aspects. One is whether and how PMI quality influences employee self-view correctness. As stated, I separate PIM quality into three parts, sensitivity, precision and verifiability. Sensitivity of PMI quality is tested by 2 questions asked to employees while precision and verifiability of PMI quality are tested by questions asked to managers. I also separate employee self-view correctness into two parts. One part concerns the correctness of employee self-view, which is tested by 2

(16)

16 questions. The other part concerns the correctness of employee self-view comparing to his manager, which is also tested by 2 questions.

As a result of the reparation, I made three hypotheses for the relationship between PMI quality and employee self-view correctness.

H1: Employee has a higher level of self-view correctness when the sensitivity of PMI in organization he/she is working for has a higher quality.

H2: Employee has a higher level of self-view correctness when the precision of PMI in organization he/she is working for has a higher quality.

H3: Employee has a higher level of self-view correctness when the verifiability of PMI in organization he/she is working for has a higher quality.

The second aspect investigated in this study is whether and how employee educational level influences the relationship between PMI quality and employee self-view correctness. One hypothesis is made test this possible effect.

H4: This relationship between PMI quality and employee self-view correctness is more visible when employee has a higher educational level.

Figure2.1 Hypotheses structure

Sensitivity Precision Verifiability

PMI quality

Employee self-view correctness

(17)

17

3 Methodology

Data of this study is from a survey held by prof. Bianca A.C. Groen in University van Amsterdam. Professor Groen provides the survey to students willing to join this program. And students who find at least 6 respondent pairs can share the final database. We originally found 117 pairs of participants. Each pair contains one employee and his/her direct manager. Totally 106 employees and 103 managers answered the questionnaire. The response rate is 90.6% for employees and 88% for managers. As the data useable must be in the form of pairs (i.e. one employee and his/her manager), the finally database contain 99 pairs of participants. The response rate for pairs is 84.6%. The participants of this survey are mainly from Netherlands (53.03%) and China (30.81%). The other participants are from Turkey (5.56%), Ukraine (0.51%), France (0.51%), Spain (0.51%), south Africa (1.01%), Belg (1.01%), Britain (2.02%), Russia (0.51%), Cyprus (1.52%) and Azerbaijan (1.01%). 4 participants do not fulfill their nationality, which is 2.02% of the total participants. These samples are selected through following students’ network. The working industries of participants include economic, hotel, IT, law service, etc.

The questionnaire of this survey is designed by professor Groen. It tests many items/scales like influence on PMIs, Norm to perform well, Leader-member exchange, Self-reported job performance and etc. It also asked some basic information of the participants. This questionnaire is well designed. For some key questions, to get reliable answer the same question is asked twice in different ways as different place. All the items have a 7-point fully anchored Likert scale: (1) Totally disagree, (2) Disagree, (3) Moderately disagree, (4) Neutral, (5) Moderately agree, (6) Agree, (7) Totally agree. This design also enables me to make a regression analysis to make the relation more clear. There are five Chinese students in this survey program. Professor Groen and we made the Chinese version of the survey. In the past, professor Groen has translated the survey from Dutch to English. We firstly arranged two students to translate the English version

(18)

18 to Chinese and then two others translated the Chinese version back to English. After comparing the difference of these two English versions, we discuss the unclear part together and for the questions we cannot handle we ask professor Groen for the aim of this survey question and the meaning of the original Dutch question. In this way, we made a Chinese version quickly and know better of the aim of this survey. After that the one another student translates the invitation email to be sent to the participants and upload the Chinese version survey to the survey system. Then we test if the Chinese version can runs smoothly in the system and find out dealing methods for some little problems.

I use some of items in this survey to investigate my research question. The items I find useful for my paper are the questions evaluating the quality of PMIs and questions evaluating employee performance answered by both employees and their managers. These questions are listed below.

3.1 Questions about PM quality

The evaluation of PMI from employees and their managers is used to see the quality of PMI in the organizations. According to the definition of PM quality, the questionnaire mainly tests the sensitivity, precision and verifiability of PM (Groen, 2012). According to previous researches (Groen, 2012; Moers, 2006), employees usually want their efforts to be displayed through performance measurement and whether the input of an employee influences their score is important for them, which means employees see the sensitivity of PMI more important. And managers mainly look at output when determining employee job performance, so managers see the precision and verifiability more important. So I use survey result from employees to evaluate the sensitivity of PMI quality and ask questions to managers to evaluate the precision and verifiability of PMI quality.

(19)

19 Question 1: My performance expressed in the performance indicators is strongly affected by changes in the behavior of parties outside the organization, such as customers, suppliers or competitors.

This type of question testing PMI sensitivity is also used in Moers (2006). The reason of this question is that PMI is sensitive to employee’s performance if it is only influenced by employee performance. As a result, if PMI is affected by changes in the behavior of parties outside the organization it is less sensitivity to employee’s performance.

Employees need to score these two questions from 1 to 7. A score 1 means a strong disagree of this sentence and a score 7 means a strong agree of this sentence. This question concerns about whether PMI result is also influenced by indicators out of employee’s control. From the expression of the question, we can see that if employee scores 7 to this question, which means a strong agree, the performances evaluated by PMI are affected by changes outside the organization. The PMI is not sensitive to employee performance. So for this question, a score 7 means PMI quality is bad and a score 1 means PMI quality is good.

Two questions about precision and verifiability of PMI quality are asked to managers. Question 2: My employee’s performance expressed in the performance indicators is strongly affected by decisions made in other parts of the organization.

Question 3: The performance indicators are objective and verifiable.

Same as the questions about PMI sensitivity to employees, managers also need to score these two questions from 1 to 7. A score 1 means a strong disagree of this sentence and a score 7 means a strong agree of this sentence. Question 2 concerns about whether PMI result is also influenced by indicators out of employee’s control. If the PMI result is not influenced by other indicators, this means it can measure employee performance precisely. Question 3 asks directly if PMI is objective and verifiable.

(20)

20 Reading question 2 and 3, you can find that a score 7 for question 2 means precision of PMI bad while a score 7 for question 3 means PMI is good at verifiability. To make the 3 question unified and easier to understand, I change the original survey result of question 1 and 2 to |8-x| (x is the original answer of question 1 and 2). After this transformation, for all these three questions, a score 1 means bad PMI quality and a score 7 means good PMI quality.

3.2 Questions about employee performance

The evaluation of employees from themselves and their managers is used to see the correctness of employee self-view. There are two questions about employee performance to employees and managers. I calculate the score of employee self-view correctness by comparing the difference between the answers from employees and their managers. Participants need to score their answer from 1 to 7. The difference between the answer from employee and his manager will be calculated to score the correctness of employee self-view. That is, if there is a big difference between scores from employee and his manager, the correctness is low, which means that the employee does not have an appropriate view on himself. This method is based on the assumption that employees work to fulfill their managers’ requirements and that managers as standing in the higher level in the organization can see their employee clearly.

Two questions are asked to employees and managers about employees’ performance. The questions asked to employees are:

Please imagine that there is someone who is always with you in all working situations. This person would therefore know you thoroughly in all kinds of situations.

To what extent do you agree with the following statements? This imaginary person would say about me:

EM1. He/she always performs all tasks that are expected of him/her.

(21)

21 The questions asked to managers are:

The following statements are about this employee. To what extent do you agree with the following statements?

MA1. He/she always performs all tasks that are expected of him/her. MA2. He/she always meets all formal performance requirements of the job.

Observing the above questions, EM1 and MA1 are asked in nearly the same way about whether the employee do what he is required to do and EM2 and MA2 are asked in nearly the same way about whether the employee do his job in required quality. By asking the same questions to employees and their pair managers, I get employee self-view and manager’s evaluation to employee. The core concerns of these two pairs of questions are different. Question EM1 and MA1 concerns about whether employee does all he is expected to do, and question EM2 and MA2 concerns about whether employee does his work in an expected quality.

Seeing these two pairs of questions separately we can observe the relationship between PMI quality and different aspect of employee performance. Putting them together, we can get a whole picture about how PMI quality influences the correctness of employee self-view. As a result, I also calculate the average these two results to know PMI quality’s influence on the employee self-view as a whole.

After calculate the scores of the quality of PMI and the correctness of employee self-view, I will divide the data gathered into different groups by its score and observe scores of the correctness of employee self-view in different group. And then do a liner regression analysis to get more specific results.

(22)

22

3.3 employee educational level

Employee educational level is tested as a moderating variable to research on its effect on the relationship between PMI quality and employee self-view correctness. The question about employee educational level is:

What is your highest completed education? (If your education is not listed, please choose the most similar one)

1 Primary education

2 Lower vocational education 3 Intermediate general education 4 Intermediate vocational education 5 Higher general education

6 Higher vocational education 7 Scientific education

Employees need to choose one from options 1 to 7. Option 1 is the lowest educational level and option 7 is the highest educational level. The result of this question about employee educational level is represented as EDU in later discussions. The higher the EDU is, the higher educational level employee has. I predict that the relationship between PMI quality and employee educational level is more visible when employee has a higher educational level.

3.4 Process of results

As discussed above, when scoring the quality of PMI, managers look at the verifiability of the output of PMI and the precision in measuring relevant aspects of employees’ performance, while employees usually also want the PMI sensitive to their efforts (Groen, 2012; Moers, 2006). This divides PMI quality into three parts. Two of the three parts are

(23)

23 represented by the answers given by managers, which reflect the verifiability and precision of PMI. The third part is represented by the scores given by employees, which reflects the sensitivity of PMI.

As stated above, one question about PMI sensitivity is asked to employees. Result of this question will be transferred by formula |8-x|. The transferred result of this question is represented as SENSI. One question about PMI precision is asked to managers. Result of this question will also be transferred by formula |8-x|. The transferred result of this question is represented as PRECI. One question about PMI verifiability is asked to managers, which is represented as VERIF. SENSI, PRECI and VERIF are all ranges from 1 to 7. The higher the score is, the better the sensitivity/precision/verifiability of PMI is.

The score of employee self-view is calculated by comparing the answers of pairs of questions between an employee and his manager. There are two pairs of questions for employees and managers about employee performance. According to the survey questions listed above, these two pairs of questions about employee self-view correctness are about different aspects. One is about how correct employee understands his working scope, and the other type is about how correct employee thinks his working quality. The final score used to evaluate employee self-view correctness is the difference between employee’s and manager’s answer. The smaller the difference is, the more correct employee self-view is. The results of the three pairs of questions are represented as SELF_VIEW1 and SELF_VIEW2.To get a whole image of employee self-view correctness, I also calculate the average of SELF_VIEW1 and SELF_VIEW2 and name this score SELF_VIEW.

Employee educational level is gathered directly in single scores from 1 to 7. Level 1 refers to the lowest educational level and level 7 refers to the highest educational level. When analyzing the result scores, I use EDU to represent the score of employee educational level. The next part discusses the findings of this research.

(24)

24

4 Findings

This part represents the findings of this research. I investigate on the relationship between PMI quality and the correctness of employee self-view, and how employee educational level influences on this relationship. As PMI quality consists of 3 elements sensitivity, precision and verifiability, I define three dependent variables, SENSI, PRECI and VERIF, which refers to the sensitivity, precision and verifiability of PMI quality respectively.

I made four hypotheses for this research. Three of them are about the relationship between employee self-view correctness and PMI quality. I predict that employee has a higher level of self-view correctness when the performance measurement indicators the organization he/she is working for has a higher quality. The other hypothesis is about the influence of employee educational level on the relationship between PMI quality and employee self-view correctness. For this hypothesis, I predict that the relationship between performance measurement indicators quality and employee self-view is more visible when employee has a higher educational level.

By dividing the concept of PMI quality into three elements (sensitivity, precision and verifiability), I can test the relationship between PMI quality and employee self-view correctness from three aspects. I expect that with the increase of the quality of PMI on sensitivity, precision and verifiability employee would have a better self-view correctness. Hypothesis 4 investigates on if employee educational level is an influence factor to the relationship between PMI quality and employee self-view correctness. As hypothesis 1 to hypothesis 3 discuss three situations of the relation between PMI quality and employee self-view correctness, Hypothesis 4 should also be tested from these three aspects. The statistical analysis tests how employee educational level affects the relationship between employee self-view correctness and the three elements of PMI quality (sensitivity, precision and verifiability). I expect that, when employee has a higher educational level,

(25)

25 the positive relation between employee self-view correctness and the three elements of PMI quality (sensitivity, precision and verifiability) will be more visible.

The analysis of survey results can be divided to description analysis and statistical analysis. I use group analysis method for description analysis and liner regression analysis method for statistical analysis. The conclusion and detail of the analysis results are shown below.

4.1 description analysis

I first analyze in a lower accurate level, i.e. group level. To observe if relationships between SENSI and SELF_VIEW, SENSI and SELF_VIEW2, SENSI and SELF_VIEW, PRECI and SELF_VIEW1, PRECI and SELF_VIEW2, PRECI and SELF_VIEW, VERIF and SELF_VIEW1, VERIF and SELF_VIEW2 and VERIF and SELF_VIEW exist in group level, I divide the 99 survey results into groups according to SENSI, PRECI and VERIF. Group analysis is a less accurate analysis method comparing to liner regression analysis. So even though there is no relation between employee self-view correctness and the three elements of PMI quality (sensitivity, precision and verifiability) appears in liner regression analysis, the relationship may be observed in group analysis.

The group deviations are shown in Figure 4.1, Figure 4.2 and Figure 4.3. Group analysis using groups divided according to SENSI will illustrate the distribution of employee self-view correctness in different PMI sensitivity levels. Group analysis using groups divided according to PRECI will illustrate the distribution of employee self-view correctness in different PMI precision levels. And group analysis using groups divided according to VERIF will illustrate the distribution of employee self-view correctness in different PMI precision levels. So group analysis will be tested in these three aspects.

(26)

26 score Number of pairs

1 11 2 32 3 21 4 9 5 8 6 13 7 5

Total number of pairs 99

Figure 4.1 Group according to SENSI

score Number of pairs

1 7 2 16 3 31 4 9 5 13 6 17 7 6

(27)

27 score Number of pairs

1 1 2 3 3 3 4 9 5 33 6 35 7 15

Figure 4.3 Group according to VERIF

Figure 4.1 shows the groups according to score SENSI. The total 99 pairs of survey results are divided into 7 groups. The number of survey results located in group 2 is obvious bigger than the number of survey results located in other groups. And only 5 survey results located in group SENSI 7. So when analyzing in a group basis, the average number of employee self-view correctness may be less reliable than the average in other groups. Similarly, group VERIF 1 has only 1 item, group VERIF 2 has 3 items and group VERIF 3 has 3 items. The same problem may appear in these three groups also that the calculated average score of employee self-view correctness numbers are less reliable than the averages in other groups.

I firstly analyze the difference of SELF_VIEF in the different groups. As stated above, a higher SELF_VIEW means employee has a less correct self-view. In another word, employee’s score to himself has bigger difference with manager’s score to him. And a lower SELF_VIEW means employee has a self-view with higher correctness. The results of group analyses about the relationship between PMI quality and SELF_VIEW are shown in Figure 4.4, Figure 4.5 and Figure 4.6.

(28)

28

Figure 4.4 the average scores of SELF_VIEW in groups divided according to SENSI

Figure 4.4 shows the correctness of employee self-view not comparing to his manager in different sensitivity of PMI quality levels. And SELF_VIEW1 only exams how correct employee realize the scope of their work, SELF_VIEW2 exams whether employee do his work in required quality and SELF_VIEW tests a combination of SELF_VIEW1 and SELF_VIEW2.

According to hypothesis 1, I expect a positive relationship between the correctness of employee self-view and the sensibility of PMI quality. Because SELF_VIEW is lower when employee has a more correct self-view, SENSI is expected to have a negative relationship with SELF_VIEW1, SELF_VIEW2 and SELF_VIEW. If it is shown in line graph, the lines in Figure 4.4 should all have in downward trend with or without fluctuations.

However, we can observe upward trends in Figure 4.4. From the line representing SELF_VIEW1 we can see that the highest average score is 1.800 in group 7 and the lowest score if 0.667 in group 4. Even though a downward trend can be observed from group SENSI 1 to 4, the average SELF_VIEW1 goes up in group SENSI 5 to 1.5 and

0.909 0.813 _0.762 0.667 1.500 1.154 1.800 0.727 0.906 1.333 1.111 1.125 1.154 1.400 0.818 0.859 1.048 0.889 1.313 1.154 1.600 0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400 1.600 1.800 2.000 1 2 3 4 5 6 7 SENSI Average:SELF_VIEW1 Average:SELF_VIEW2 Average:SELF_VIEW

(29)

29 goes down to 1.154 in group SENSI 6. The highest average SELF_VIEW1 in group SENSI 7 of 1.800 can be ignored due to the small amount of results in group SENSI 7, while the other 6 SENSI groups do not show a clear trend.

As a result, even though we can observe a trend that in the low range of SENSI level groups with the increase of PMI sensibility the average SELF_VIEW1 is lower. The trend is not observed in the whole PMI sensibility level. This cannot prove that employee understands working scope better under PMI system with better sensibility.

The other two lines representing SELF_VIEW2 and SELF_VIEW both show a fluctuated upward trend from group SENSI 1 to 7. The average SELF_VIEW2 increases from 0.727 in group SENSI 1 to 1.400 in group SENSI 7. The average SELF_VIEW increases from 0.818 in group SENSI 1 to 1.600 in group SENSI 7.

When PMI sensitivity is not high (between score 1 to 4), a positive relation between PMI sensitivity and employee’s understanding of working scope can be observed. However this cannot prove the positive relationship between PMI sensitivity and employee self-view correctness expected in hypothesis 1.

(30)

30

Figure 4.5 the average scores of SELF_VIEW in groups divided according to PRECI

Figure 4.5 shows the correctness of employee self-view in different precision of PMI quality levels. As stated in hypothesis 2, I expect that a positive relationship may exist between the correctness of employee self-view not comparing to manager and the precision of PMI quality. So PRECI is expected to have a negative relationship with SELF_VIEW1. In other words, if the lines shown in Figure 4.5 show downward trends, the positive relationship between the correctness of employee view about working scope and the precision of PMI quality can be proven.

The line representing average scores of SELF_VIEW1 shows a clear trend from group PRECI 1 to 4 that SELF_VIEW1 decreases with the increase of PRECI. However, same as the line representing average scores of SELF_VIEW1 Figure 4.4, the average number of SELF_VIEW1 increases sharply in group PRECI 5 and decreases again in group PRECI 6 and 7. While different as Figure 4.4, the average SELFF_VIEW1 in group PRECI 7 is in the lowest point at 0.667 which is same as the average number is group PRECI 4. So we can observe a fluctuated negative trend between PRECI and SELF_VIEW1. According to this group analysis result, the correctness of employee

1.286 1.188 0.903 0.667 1.231 0.706 0.667 1.000 1.125 1.226 1.111 0.769 0.765 1.667 1.143 1.156 1.065 0.889 1.000 0.735 1.167 0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400 1.600 1.800 1 2 3 4 5 6 7 PRECI avreage:SELF_VIEW1 average:SELF_VIEW2 average:SELF_VIEW

(31)

31 self-view has a positive relationship with the precision of PMI to employee’s performance.

The line representing the average scores of SELF_VIEW2 goes up from 1.000 in group PRECI 1 to 1.226 in group PRECI 3. It then has a downward trend from group PRECI 3to group PRECI 6 (0.765), while grow sharply in group PRECI 7 to 1.667. The average scores of SELF_VIEW2 in group PRECI 7 of 1.667 is also the highest score. This line cannot provide evidence of the negative relationship between PRECI and SELF_VIEW2.

As a combination of SELF_VIEW1 and SELF_VIEW2, the line representing average scores of SELF_VIEW shows a fluctuated downward trend from group PRECI 1 of 1.143 to group PRECI 6 of 0.735. However, due to the influence of SELF_VIEW2, the average score of SELF_VIEW goes to the highest point at group PRECI 7 to 1.167. So this line provides weak evidence to the positive relationship between employee self-view correctness and PMI precision.

As a result, the three lines in Figure 4.5 together can only show weak evidence about the relationship between employee self-view correctness and PMI precision. We can observe a possible existing trend but cannot prove the trend through the ling graph.

(32)

32

Figure 4.6 the average scores of SELF_VIEW1 in groups divided according to VERIF

Figure 4.6 shows the correctness of employee self-view not comparing to his manager in different verifiability of PMI quality levels. Same as Figure 4.4 and Figure 4.5, SELF_VIEW1 exams how correct employee realize the scope of their work, SELF_VIEW2 exams whether employee do his work in required quality and SELF_VIEW tests a combination of SELF_VIEW1 and SELF_VIEW2.

The trend of the three lines in Figure 4.6 is same. They go up from group VERIF 1 to group VERIF 2 and goes down largely in group VERIF 3. The average scores of SELF_VIEW1, SELF_VIEW2 and SELF_VIEW in group VERIF from 3 to 7 are in a same level. And observe the lines in detail, we can find that the three lines in Figure 4.6 are in lightly downward trend from group VERIF 4 (0.889, 1.222 and 1.056 for the average scores of SELF_VIEW1, SELF_VIEW2 and SELF_VIEW respectively) to group VERIF 4 (0.800 for all the three lines). As stated above, group VERIF 1 has only 1 item, group VERIF 2 has 3 items and group VERIF 3 has 3 items. So it is reasonable to assume that the average scores in these three groups are not as reliable as the average scores in other groups. If ignore the results shown in group VERIF 1, 2 and 3, Figure 4.6 provides evidence about a positive relationship between employee self-view correctness and PMI verifiability.

1.000 2.333 1.000 0.889 0.939 0.914 0.800 2.000 3.000 0.667 1.222 0.970 1.086 0.800 1.500 2.667 0.833 1.056 0.955 1.000 0.800 0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 1 2 3 4 5 6 7 VERIF Average:SELF_VIEW1 Average:SELF_VIEW2 Average:SELF_VIEW

(33)

33 According to Figure 4.4, 4.5 and 4.6, the line graphs show positive relationship exist between employee self-view correctness not comparing to his manager and PMI precision and verifiability. No evidence of positive relationship between employee self-view correctness not comparing to his manager and PMI sensitivity are found in group analysis.

As mentioned above, employee educational level is answered from 1 to 7 directly in the survey answer. The scores 1 to 7 refers to primary education, lower vocational education, lower vocational education, intermediate vocational education, higher general education, higher vocational education, scientific education. In the survey in Chinese version, we change the name of the each option, and delete the score 2 from the 7 options according to Chinese education system. However we try to find the best match of each educational level in Netherland system and make sure that, from score 1 to score 7, the educational level of the related employee grows in the similar of the Dutch version.

Using the group division in Figure 4.1, 4.2 and 4.3, I make Figure 4.7 to 4.15 considering the influence of employee educational level on the relation between PMI quality and employee self-view correctness. As SELF_VIEW1 reflects how correct employee understand this working scope, SELF_VIEW1 reflects how correct employee evaluates his working quality and SELF_VIEW is the average of SELF_VIEW1 and SELF_VIEW2, providing the whole image of employee self-view correctness. Figure 4.7, 4.8 and 4.9 show how employee educational level influences the relationship between PMI sensitivity and employee self-view correctness. Figure 4.10, 4.11 and 4.12 show how employee educational level influences the relationship between PMI precision and employee self-view correctness. And Figure 4.13, 4.14 and 4.15 show how employee educational level influences the relationship between PMI verifiability and employee self-view correctness.

These 9 graphs show the average SELF_VIEW1, SELF_VIEW2 and SELF_VIEW first divided by employees’ different educational level, then divided by groups shown in

(34)

34 Figure 4.1, 4.2 and 4.3. As stated in hypothesis 2, I expect the relationship between PMI quality and employee self-view correctness to be more visible when employee has a higher educational level. If this shows up in the figures, the downward trend would be bigger when employee educational level is higher. Even a positive relation does not appear in group analysis results shown early, it may appear at the range with higher employee educational level.

Figure 4.7 the average SELF_VIEW1 in groups according to SENSI

Figure 4.8 the average SELF_VIEW2 in groups according to SENSI 0.000 0.500 1.000 1.500 2.000 2.500 3.000 4 2 3 4 5 1 2 3 4 6 7 1 2 3 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 0.000 0.500 1.000 1.500 2.000 2.500 3.000 3.500 4 2 3 4 5 1 2 3 4 6 7 1 2 3 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7

(35)

35

Figure 4.9 the average SELF_VIEW in groups according to SENSI

Figure 4.7 to Figure 4.9 show the average SELF_VIEW1, SELF_VIEW2 and SELF_VIEW for employees with same SENSI group and same educational level. According to the result in Figure 4.3, little relationship exists between SENSI and SELF_VIEW. The analysis taken in Figure 4.7, 4.8 and 4.9 are to see if the relationship between SENSI and SELF_VIEW will change with the change of employee educational level. With the assumption that employee with higher educational level can understand the hidden beaning of PMI better, I expect the downward trend of SELF_VIEW in more visible at the place employee educational level in higher.

We cannot observe a downward trend of the average score of SELF_VIEW1, SELF_VIEW2 or SELF_VIEW in any educational level group in Figure 4.7 to 4.9, while we can observe a clear upward trend in the three graphs when employee educational level is 7. So Figure 4.7 to 4.9 do not provide evidence in proving hypothesis 4.

0.000 0.500 1.000 1.500 2.000 2.500 4 2 3 4 5 1 2 3 4 6 7 1 2 3 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7

(36)

36

Figure 4.10 the average SELF_VIEW1 in groups according to PRECI

Figure 4.11 the average SELF_VIEW2 in groups according to PRECI 0.000 0.500 1.000 1.500 2.000 2.500 7 2 3 4 5 2 3 4 6 7 1 2 3 4 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7 0 0.5 1 1.5 2 2.5 3 3.5 7 2 3 4 5 2 3 4 6 7 1 2 3 4 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7

(37)

37

Figure 4.12 the average SELF_VIEW in groups according to PRECI

Figure 4.10, 4.11 and 4.12 bases on Figure 4.5 testing how employee educational level influence the relationship between employee self-view correctness and the precision of PMI quality. When considering the effect of employee educational level, I expect the downward from low PRECI to high PRECI becomes more and more visible with the increase of employee educational level.

Observing the three graphs, we cannot find clear trend from any educational level in Figure 4.10, while light downward trend can be found in Figure 4.11 in group EDU 5 and 6, and clear down trend can be found in Figure 4.12 in group EDU 5. While these findings do not provide enough evidence to a influence of employee educational level to the relationship between PMI precision and employee self-view correctness.

0 0.5 1 1.5 2 2.5 7 2 3 4 5 2 3 4 6 7 1 2 3 4 6 1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7

(38)

38

Figure 4.13 the average SELF_VIEW1 in groups according to VERIF

Figure 4.14 the average SELF_VIEW2 in groups according to VERIF 0 0.5 1 1.5 2 2.5 3 3.5 1 6 6 4 5 6 7 2 3 5 6 7 2 3 4 5 6 7 2 4 5 6 7 1 2 3 4 5 6 7 0 1 2 3 4 5 6 1 6 6 4 5 6 7 2 3 5 6 7 2 3 4 5 6 7 2 4 5 6 7 1 2 3 4 5 6 7

(39)

39

Figure 4.15 the average SELF_VIEW in groups according to VERIF

Figure 4.13, 4.14 and 4.15 bases on Figure 4.6 testing how employee educational level influence the relationship between employee self-view correctness and the verifiability of PMI quality. When considering the effect of employee educational level, I expect the downward from low VERIF to high VERIF becomes more and more visible with the increase of employee educational level.

Observing the three graphs, we can find fluctuated downward trends of SELF_VIEW1, SELF_VIEW2 and SELF_VIEW in group EDU 6 and 7 in all the three graphs, and a downward trend can be found in group EDU 4 in Figure 4.14 and 4.15. No trends can be found in other EDUCA groups. As the downward trend in EDU 6 and 7, when employee educational level is high, is more visible than in the other groups, Figure 4.13, 4.14 and 4.15 provide evidence to my assumption that the relationship between PMI verifiability and employee self-view correctness not comparing to the manager is more visible when employee has a higher educational level.

In conclusion to the 9 graphs from Figure 4.7 to Figure 4.15, my findings can support the assumption that the relationship between PMI verifiability and employee self-view correctness is more visible when employee has a higher educational level. The influence

0 0.5 1 1.5 2 2.5 3 3.5 4 1 6 6 4 5 6 7 2 3 5 6 7 2 3 4 5 6 7 2 4 5 6 7 1 2 3 4 5 6 7

(40)

40 of employee educational level to the relationship between employee self-view correctness and the sensitivity and precision of PMI quality cannot be proven.

4.2 Statistical analysis

The results of liner regression analysis are shown in Figure 4.16 to Figure 4.25. The 10 figures represent the statistical analysis result of four hypotheses. As PMI quality is higher when SENSI, PRECI and VERIF is higher, and employee self-view correctness is better when SELF_VIEW1, SELF_VIEW2 and SELF_VIEW is lower, a significant

relevant relationship between SENSI/PRECI/VERIF and

SELF_VIEW1/SELF_VIEW2/SELF_VIEW can be the evidence for hypotheses. Figure 4.16, 4.17 and 4.18 represents the relationship between PMI sensitivity and employee self-view correctness. As stated in hypothesis 1, I expect that employee has a more correct self-view when the sensitivity of PMI quality is better. Because SELF_VIEW1, SELF_VIEW2 and SELF_VIEW are lower when employee has a more correct self-view, SENSI is expected to have a negative relationship with SELF_VIEW1, SELF_VIEW2 and SELF_VIEW.

Liner regression analysis

Multiple R 0.217

R Square 0.047

Adjusted R Square 0.037 standard deviation 0.955 observed value 99.000

Coefficients Standard error t Stat P-value

Lower 95% Upper 95% Intercept 0.554 0.204 2.709 0.008 0.148 0.960 X Variable 1 0.120 0.055 2.191 0.031 0.011 0.228

Figure 4.16 liner regression analysis–SENSI and SELF_VIEW1

Figure 4.16 represents the relationship between SENSI and SELF_VIEW1. SENSI refers to the sensitivity of PMI quality and SELF_VIEW1 reflects the correctness of employee understanding of working scope.

(41)

41 The liner regression analysis result shows that the Multiple R is 0.217, the R square is 0.047, the Adjusted R Square is 0.037 and the Standard Deviation is 0.955. Multiple R refers to the correlation between actual and predicted values of the dependent variable. In this liner regression analysis, the Multiple R of 0.217 is little means that SENSI has little relation with SELF_VIEW1.

R Square describes to what extent SENSI can explain the changes of SELF_VIEW1. It is illustrates how close the data are to the fitted regression line. R Square is a figure between 0 and 1. In general, the higher the R square, the better the model fits the data. Here the R Square between SELF_VIEW1 and SENSI is 0.047 means that 4.7% of SELF_VIEW’s change can be explained by changes of SENSI. The low R Square indicates low cross correlation between PMI precision and verifiability and employee self-view correctness. Adjusted R Square is 0.037, which also indicts that SELF_VIEW1 has little relation with SENSI. Standard deviation is used to quantify the amount of variation or dispersion of a set of data values. A low standard deviation indicates that the data points tend to be close to the mean of the set, while a high standard deviation indicates that the data points are spread out over a wider range of values. In this liner regression, the standard deviation is 0.955, which means that the amount of Scores tested are close to their mean.

Even though P value of the liner regression analysis is 0.031 which represents a significant relation level, coefficients between SENSI and SELF_VIEW1 is positive. So this liner regression analysis result represents a significant positive relation between SENSI and SELF_VIEW1. This result cannot prove hypothesis 1.

(42)

42 Liner regression Multiple R 0.125 R Square 0.016 Adjusted R Square 0.005 standard deviation 1.106 observed value 99.000 Coefficients Standard

error t Stat P-value

Lower 95% Upper 95% Intercept 0.812 0.237 3.428 0.001 0.342 1.282 X Variable 1 0.078 0.063 1.237 0.219 -0.047 0.204

Figure 4.17 liner regression analysis–SENSI and SELF_VIEW2

Figure 4.17 represents the relationship between SENSI and SELF_VIEW2. SELF_VIEW2 reflects the correctness of employee evaluating their working quality and SENSI refers to the sensitivity of PMI quality. SENSI is expected to have a negative relationship with SELF_VIEW2.

The liner regression analysis result shows that the P value is 0.219, Multiple R is 0.125, the R square is 0.016, the Adjusted R Square is 0.005 and the Standard Deviation is 1.106. The Multiple R of 0.125 is close to 0 means that SENSI has little relation with SELF_VIEW2. R Square between SELF_VIEW2 and SENSI is nearly 0, means that SELF_VIEW2’s change cannot be explained by changes of SENSI. Coefficients of 0.078 shows positive relation between the two variables. P value of 0.219 does not show a significant relationship between SENSI and SELF_VIEW2.

Figure 4.17, the result of liner regression analysis, does not show a relationship between PMI sensitivity and the correctness of employee self-view. This liner regression analysis cannot prove hypothesis 1.

(43)

43 Liner regression Multiple R 0.206 R Square 0.042 Adjusted R Square 0.033 standard deviation 0.834 observed value 99.000 Coefficients Standard

Lower 95% Upper 95% Intercept 0.683 0.179 3.822 0.000 0.328 1.038 X Variable 1 0.099 0.048 2.073 0.041 0.004 0.194

Figure 4.18 liner regression analysis–SENSI and SELF_VIEW

Figure 4.18 is the result of liner regression analysis between SENSI and SELF_VIEW. SENSI refers to the sensitivity element of PMI quality and SELF_VIEW is the combination of SELF_VIEW1 and SELF_VIEW2, which shows a whole image of employee self-view correctness. I expect that SELF_VIEW has a negative relation with SENSI.

The Multiple R of 0.206 shows little relationship between SENSI and SELF_VIEW. R Square between SELF_VIEW and SENSI is 0.042, means that 4.2% of SELF_VIEW’s change can be explained by changes of SENSI. The P-value is 0.041, which means that the relation between SENSI and SELF_VIEW is significant. While a positive coefficients of 0.099 represents that this significant relation is positive.

As a result, Figure 4.18 shows a significant positive relationship between the sensitivity element of PMI quality and the correctness of employee self-view. This liner regression analysis result does not provides evidence to hypothesis 1.

To test hypothesis 2, the relationship between PMI precision and employee self-view correctness, I make liner regression between PRECI and SELF_VIEW1, PRECI and SELF_VIEW2 and PRECI and SELF_VIEW. The results of these three liner regression

(44)

44 analysis are shown in Figure 4.19, Figure 4.20 and Figure 4.21. If my assumption in hypothesis 2 is proven correct, significant negative relationship would be observed in these three figures.

liner regression Multiple R 0.140 R Square 0.020 Adjusted R Square 0.010 standard deviation 0.968 observed value 99.000 Coefficients Standard

Lower 95% Upper 95% Intercept 1.252 0.238 5.268 0.000 0.780 1.723 X Variable 1 -0.079 0.057 -1.394 0.166 -0.192 0.034

Figure 4.19 liner regression analysis–PRECI and SELF_VIEW1

Figure 4.19 shows how PMI precision influences the correctness of employee understanding working scope. The Multiple R of 0.140 shows little relationship between PRECI and SELF_VIEW1. R Square between SELF_VIEW1 and PRECI is 0.020, means that 2% of SELF_VIEW1’s change can be explained by changes of SENSI. These figures show little relation between PRECI and SELF_VIEW1. Coefficients of -0.079 shows a negative relation between PRECI and SLEF_VIEW1. However, the P-value is 0.166, which is bigger than 0.05. So it means that the relation between PRECI and SELF_VIEW1 is not significant. Figure 4.19 cannot provide evidence to hypothesis 2.

(45)

45 liner regression Multiple R 0.046 R Square 0.002 Adjusted R Square -0.008 standard deviation 1.113 observed value 99.000 Coefficients Standard

Figure 4.20 liner regression analysis–PRECI and SELF_VIEW2

Figure 4.20 shows how PMI precision influences the correctness of employee evaluating his working quality. The Multiple R of 0.046 shows little relationship between PRECI and SELF_VIEW2. R Square between SELF_VIEW2 and PRECI is 0.002, means that 0.2% of SELF_VIEW2’s change can be explained by changes of PRECI. These figures show little relation between PRECI and SELF_VIEW2. Again, we can observe a negative relation from negative coefficients, while the P-value of 0.649 shows this relation is not significant. So it means that the negative relation between PRECI and SELF_VIEW2 is not significant. Figure 4.20 cannot provide evidence to hypothesis 2.

liner regression Multiple R 0.111 R Square 0.012 Adjusted R Square 0.002 standard deviation 0.847 observed value 99.000 Coefficients Standard

Does the quality of performance measurement indicator affect employee self-view?