Pearson Correlation Analysis And Multiple Regression Analysis

Published Date: 02 Nov 2017

4.0 Introduction

This chapter begins with result of pilot test that had been conducted, descriptive analysis which comprises of demographic profile of the respondent and central tendencies measurement of constructs, followed by scale measurement and inferential analysis, and lastly the conclusion for this chapter. SAS version 9.3 software, Pearson Correlation Analysis and Multiple Regression Analysis were used to perform the analysis.

4.1 Pilot Test

Table 4.1 : Result of Reliability Analysis for Pilot Test

Cronbach’s Alpha

Number of items

Conscientiousness

0.86

5

Agreeableness

0.88

5

Openness to Experience

0.82

5

Extraversion

0.81

5

Neuroticism

0.85

5

Students’ Performance

0.72

8

Source: Developed for the research

Based on the Table 4.1, the questionnaire tested is consistent and reliable to be analysed as all the variables have a Cronbach’s alpha value of more than 0.7. Achieving values of more than 0.7 were considered the pilot test had passed the reliability test and the items represent a measure of high internal consistency (Bahri Yusoff, Abdul Rahim & Yaacob, 2009).

4.2 Descriptive analysis

4.2.1 Demographic Profile of the Respondents

4.2.1.1 Gender

Table 4.2 : Frequency Table for Gender

1=male 2=female

Gender

Frequency

Percent

Cumulative Frequency

Cumulative Percent

1

2

95

126

42.99

57.01

95

221

42.99

100.00

Source: Developed for the research

Figure 4.1 : Percentage of Respondents Based on Gender

Source: Developed for the research

Based on Table 4.2, it can be seen that the gender of respondents consist of 95 males (42.99%) and 126 females (57.01%).

4.2.1.2 Age

Table 4.3 : Frequency Table for Age

1= 20 years old and below 2= 21-24 years old 3= 25 years old and above

Age

Frequency

Percent

Cumulative Frequency

Cumulative Percent

1

2

3

6

213

2

2.71

96.38

0.90

6

219

221

2.71

99.10

100.00

Source: Developed for the research

Figure 4.2 : Percentage of Respondents Based on Age

Source: Developed for the research

Table 4.3 indicates that the age of respondents in this study consists of 3 stages which are 20 years old and below, 21 to 24 years old, and 25 years old and above. The respondents who are 20 years old and below have 6 people and the percentage is 2.71%; the respondents who are 21-24 years old have 213 people and the percentage is 213%; and lastly the respondents who are 25 years old and above have 2 people and the percentage is 2%.

4.2.1.3 Race

Table 4.4 : Frequency Table for Race

1= Malay 2= Chinese 3= Indian

Race

Frequency

Percent

Cumulative Frequency

Cumulative Percent

1

2

3

1

209

11

0.45

94.57

4.98

1

210

221

0.45

95.02

100.00

Source: Developed for the research

Figure 4.3 : Percentage of Respondents Based on Race

Source: Developed for the research

Table 4.4 shows that the race of respondents in this study consists of Malay, Chinese, and Indian. Therefore, the respondents in this study consists of 1 Malay (0.45%); 209 Chinese (94.57); and 11 Indian (4.98%).

4.2.1.4 Origin

Table 4.5 : Frequency Table for Origin

1= West Malaysia 2= East Malaysia

Origin

Frequency

Percent

Cumulative Frequency

Cumulative Percent

1

2

195

26

88.24

11.76

195

221

88.24

100.00

Source: Developed for the research

Figure 4.4 : Percentage of Respondents Based on Origin

Source: Developed for the research

From Table 4.5, it can be seen that the origin of respondents in this study consists of 195 from West Malaysia (88.24%) and 26 from East Malaysia (11.76%).

4.2.1.5 Nature of bachelor degree

Table 4.6 : Frequency Table for Nature of Bachelor Degree

1=Accounting course 2= Non-accounting course

Origin

Frequency

Percent

Cumulative Frequency

Cumulative Percent

1

2

221

0

100

0

221

221

100.00

100.00

Source: Developed for the research

Figure 4.5 : Percentage of Respondents Based on Nature of Bachelor Degree

Source: Developed for the research

Based on the Table 4.6, there is a total of 221 responses from accounting students were analyzed in this study which comprised 100%. This had confirmed that the responses are qualified to be analyzed as our target respondents are accounting students.

4.2.2 Central Tendencies Measurement of Constructs

Table 4.7 : Central Tendencies Measurement of Constructs: Conscientiousness

Items

Mean

Standard Deviation

Minimum

Maximum

N

CS1

3.5385

0.6358

2.00

5.00

221

CS2

3.4887

0.6001

2.00

5.00

221

CS3

3.3937

0.7347

1.00

5.00

221

CS4

3.3982

0.7415

2.00

5.00

221

CS5

3.4887

0.7840

2.00

5.00

221

Source: Developed for the research

Table 4.7 presents the mean and standard deviation for each of the item for Conscientiousness.

The item with the highest mean value is CS1 with 3.5385. The item ranked second is CS2 and CS5 with 3.4887, followed by CS4 with a mean of 3.3982. Furthermore, CS3 has the lowest mean which is 3.3937.

The item with the highest standard deviation is CS5 which is 0.7840. CS4 has the second highest standard deviation with 0.7415, followed by CS3 which has a standard deviation of 0.7347. The next item is CS1 with standard deviation of 0.6358. The item with the lowest standard deviation is CS2, with 0.6001.

Table 4.8 : Central Tendencies Measurement of Constructs: Agreeableness

Items

Mean

Standard Deviation

Minimum

Maximum

N

AG1

3.5385

0.7228

2.00

5.00

221

AG2

3.5973

0.6781

1.00

5.00

221

AG3

3.4525

0.7826

1.00

5.00

221

AG4

3.5656

0.8850

2.00

5.00

221

AG5

3.5566

0.6959

2.00

5.00

221

Source: Developed for the research

Table 4.8 presents the mean and standard deviation for each of the item for Agreeableness.

AG2 has the highest mean value with 3.5973. The item ranked second is AG4 which is 3.5656, followed by AG5 with a mean value of 3.5566. The next item is AG1 with 3.5385. AG3 has the lowest mean which is 3.4525.

The item with the highest standard deviation is AG4 which is 0.8850. AG3 has the second highest standard deviation with value of 0.7826, followed by AG1 which is 0.7228. Next is AG5 with a standard deviation of 0.6959. The item with the lowest standard deviation value is AG2, which is 0.6781.

Table 4.9 : Central Tendencies Measurement of Constructs: Openness to Experience

Items

Mean

Standard Deviation

Minimum

Maximum

N

OE1

3.4118

0.6520

2.00

5.00

221

OE2

3.4977

0.6440

2.00

5.00

221

OE3

3.4299

0.7513

2.00

5.00

221

OE4

3.6154

0.7577

1.00

5.00

221

OE5

3.4615

0.6840

1.00

5.00

221

Source: Developed for the research

Table 4.9 shows the mean and standard deviation for each of the item for Openness to Experience.

OE4 has the highest value of mean which is 3.6154. The second highest is OE2 which is 3.4977, followed by OE5 with a mean value of 3.4615. Then, OE3 is followed by, with 3.4299. Lastly, OE1 has the lowest mean which is 3.4118.

For standard deviation, OE4 has the highest value of standard deviation which is 0.7577. The second highest value of standard deviation is OE3 with 0.7513, followed by OE5, 0.6840. Then, OE1 is followed by, with a value of 0.6520. Lastly, OE2 has the lowest standard deviation, which is 0.6440.

Table 4.10 : Central Tendencies Measurement of Constructs: Extraversion

Items

Mean

Standard Deviation

Minimum

Maximum

N

EX1

3.2353

0.7500

1.00

5.00

221

EX2

3.2579

0.8589

1.00

5.00

221

EX3

3.1991

0.8822

1.00

5.00

221

EX4

3.2489

0.8978

1.00

5.00

221

EX5

3.3846

0.8536

1.00

5.00

221

Source: Developed for the research

Table 4.10 presents the mean and standard deviation for each of the item for Extraversion.

EX5 has the highest mean value which is 3.3846. The second highest is EX2 which is 3.2579. Next item is EX4 which has the value of mean of 3.2489, followed by EX1 with 3.2353. EX3 has the lowest mean value, with 3.1991.

EX4 has the highest standard deviation value which is 0.8978. The second highest standard deviation is EX3 which is 0.8822 followed by EX2, with standard deviation of 0.8589. The next item is EX5, with a value of 0.8536. EX1 has the lowest standard deviation value which is 0.7500.

Table 4.11 : Central Tendencies Measurement of Constructs: Neuroticism

Items

Mean

Standard Deviation

Minimum

Maximum

N

NE1

3.4389

0.8269

1.00

5.00

221

NE2

3.3258

0.8162

1.00

5.00

221

NE3

3.3032

0.8493

2.00

5.00

221

NE4

3.3122

0.8724

2.00

5.00

221

NE5

3.2579

0.9870

1.00

5.00

221

Source: Developed for the research

Table 4.11 presents the mean and standard deviation for each of the item for Neuroticism.

For mean, NE1 has the highest mean value which is 3.4389. The second highest is NE2 which has 3.3258, followed by NE4 with a mean value of 3.3122. The next item is NE3, with 3.3032. NE5 has the lowest mean value which is 3.2579.

Meanwhile, NE5 has the highest standard deviation with 0.9870. NE4 is the second highest standard deviation, with 0.8724 followed by NE3 which is 0.8493. Then, it is NE1 with a standard deviation of 0.8269. Lastly, the lowest standard deviation value is NE2 which is 0.8162.

Table 4.12 : Central Tendencies Measurement of Constructs: Students’ Performance

Items

Mean

Standard Deviation

Minimum

Maximum

N

SP1

2.8326

0.9066

1.00

5.00

221

SP2

3.3665

0.8348

1.00

5.00

221

SP3

3.6833

0.7318

2.00

5.00

221

SP4

3.8326

0.7532

2.00

5.00

221

SP5

3.8462

0.7348

2.00

5.00

221

SP6

3.6154

0.7456

2.00

5.00

221

SP7

3.6787

0.7391

2.00

5.00

221

SP8

3.8462

0.7885

2.00

5.00

221

Source: Developed for the research

Table 4.12 presents the mean and standard deviation for each of the items of students’ performance.

For mean calculation, SP5 and SP8 have the highest mean value with 3.8462. The second highest mean value is SP4 which has 3.8326. Next item is SP3 which has the mean value of 3.6833 followed by SP7 with 3.6787. After that, is the SP6 with a mean value of 3.6154 followed by SP2, with a mean value of 3.3665. SP1 has the lowest mean value which is 2.8326.

For standard deviation calculation, SP1 has the highest standard deviation value with 0.9066. The second highest standard deviation value is SP2 which is 0.8348. Next item is SP8 which has the standard deviation value of 0.7885 followed by SP4 with 0.7532. After that, is the SP6 with a value of 0.7456 followed by SP7 with 0.7391. Followed by, is the SP5 with standard deviation value of 0.7348 and lastly, SP3 has the lowest standard deviation value which is 0.7318.

4.3 Scale Measurement

4.3.1 Reliability Test

To determine the consistency and stability of data collected for all the variables tested in the research, Cronbach’s alpha value is used (McCrae, Kurtz, Yamagata & Terracciano, 2011). Below are the results of reliability analyses carried out on every item.

Table 4.13 : Result of Reliability Test

Variables

Cronbach’s Alpha

No. of Items

Conscientiousness

0.6885

5

Agreeableness

0.7440

5

Openness to Experience

0.7708

5

Extraversion

0.8017

5

Neuroticism

0.6098

5

Students’ Performance

0.7951

8

Source: Developed for the research

Based on Table 4.13, generally all the variables tested are consistent and reliable as majority of the variables tested have alpha value of more than 0.7. Thus, the items were considered to represent a measure of high internal consistency (Bahri Yusoff, Abdul Rahim & Yaacob, 2009). However, conscientiousness and neuroticism have lower alpha score with Cronbach’s alpha 0.6885 and 0.6098 respectively. In this case, it will still be acceptable because any measure having an intra-class correlation coefficient of at least 0.6 is still at the minimum acceptable level of reliability and considered useful (Abdul-Halim & Che-Ha, 2009; Bruton, Conway & Holgate, 2000).

4.3.2 Normality Test

Table 4.14 : Result of Normality Test

Test

Statistics

p-Value

Shapiro-Wilk

0.9776

0.0014

Kolmogorov-Smirnov

0.0562

0.0875

Source: Developed for the research

Due to the larger data samples of this research (more than 200), Kolmogorov-Smirnov test is used to determine the normality distribution of the results obtained instead of the Shapiro-Wilk test whereby it is more recommended for smaller data testing (sample size less than 50) (Mohd Razali & Yap, 2011). Table 4.14 above shows that the p-value is 0.0875 (p > 0.05), therefore the results of data is normally distributed.

4.4 Inferential Analysis

4.4.1 Pearson Correlation Coefficient

4.4.1 Pearson Correlation Coefficient

Table 4.15 : Pearson Correlation Coefficient

Variable

CS

AG

OE

EX

NE

SP

CS

1

AG

0.2885

1

OE

0.3905

0.1229

1

EX

-0.2280

-0.0507

-0.0096

1

NE

-0.0856

-0.0183

0.0609

0.0731

1

SP

0.6645

0.3276

0.3896

-0.1371

-0.1452

1

Note : CS = Conscientiousness, AG = Agreeableness, OE = Openness to experience, EX = Extraversion, NE = Neuroticism, SP = Students’ performance

Source: Developed for the research

Table 4.15 shows the correlation coefficient between independent variables and dependent variable, to determine whether there exists a multicollinearity problem among the independent variables. The highest correlation between the variables is 0.6645 whereas the lowest correlation is -0.1371. All of the independent variables tested have a coefficient of lesser than 0.9. Therefore, this indicates that there is no multicollinearity problem in this research (Saunders et. al., 2009).

There is a moderate strength of association between the variable conscientiousness and student performance with a coefficient value of 0.6645, which is within the range of 0.41 to 0.70 at significant value of lesser than 0.0001. Thus, the result indicates that there is a significant relationship between conscientiousness and student performance.

There is a small but definite relationship between the variable agreeableness and student performance with a coefficient value of 0.3276, which is within the range of 0.20 to 0.40 at significant value of lesser than 0.0001. Thus, the result indicates that there is a significant relationship between agreeableness and student performance.

There is a small but definite relationship between the variable openness to experience and student performance with a coefficient value of 0.3896, which is within the range of 0.20 to 0.40 at significant value of lesser than 0.0001. Thus, the result indicates that there is a significant relationship between openness to experience and student performance.

There is a slight and almost negligible association between the variable extraversion and student performance with a coefficient value of -0.1371, which is within the range of 0.00 to 0.20 at significant value of 0.0418. Thus, the result indicates that there is a significant relationship between extraversion and student performance.

There is a slight and almost negligible association between the variable neuroticism and student performance with a coefficient value of -0.1452, which is within the range of 0.00 to 0.20 at significant value of 0.031. Thus, the result indicates that there is a significant relationship between neuroticism and student performance.

4.4.2 Multiple Linear Regression

Table 4.16 : Model Summary

Root MSE

0.3611

R-Square

0.4922

Dependent Mean

3.5877

Adjusted R-Square

0.4804

Coefficient Variance

10.0639

Source: Developed for the research

By conducting multiple regression analysis, relationships among a set of variables and a set of responses will be determined while accounting for correlations among the responses (Hashim, Murphy, Purchase & O’Connor, 2009). The table above shows that the adjusted value of coefficient of determination (adjusted R-square) of 0.4804 shows that 48.04% of variance in students’ performance is affected by conscientiousness, agreeableness, openness to experience, extraversion and neuroticism. Meanwhile, 51.96% of variance in students’ performance is caused by other factors that are not being considered by the research.

Table 4.17 : ANOVA

Source

Df

Sum of Squares

Mean Square

F-value

Pr > F

Model

5

27.1643

5.4329

41.68

S<.0001

Error

215

28.0278

0.1304

Corrected Total

220

55.1920

Source: Developed for the research

According to the table shown above, the p-value of lesser than 0.0001 indicates that the research model is fit as it is lesser than the 0.05 level of significance (Hasan, Al-Niami & Al-Shammari, 2010). On the other hand, there is a significant relationship between all the independent variables in this research and the students’ performance based on the F-value of 41.68 which is considered as large.

Table 4.18 : Multiple Linear Regression

Variables

Df

Parameter

Estimate

Standard Error

t-Value

Pr > |t|

Tolerance

Variance Inflation

Intercept

1

0.8079

0.3229

2.5

0.0131

-

-

CS

1

0.5882

0.0605

9.72

<.0001

0.7359

1.3589

AG

1

0.1385

0.0478

2.9

0.0042

0.9164

1.0912

OE

1

0.1616

0.0528

3.06

0.0025

0.8326

1.2010

EX

1

0.0041

0.0396

0.1

0.9177

0.9384

1.0657

NE

1

-0.0971

0.0451

-2.15

0.0325

0.9800

1.0204

Note: CS = Conscientiousness, AG = Agreeableness, OE = Openness to Experience, EX = Extraversion, NE = Neuroticism

Source: Developed for the research

4.4.2.1 Parameter Estimate

The regression equation developed for this research is based on the parameter estimate:

SP = 0.8079 + 0.5882 CS + 0.1385 AG + 0.1616 OE + 0.0041 EX – 0.0971 NE

SP = Students’ Performance

CS = Conscientiousness

AG = Agreeableness

OE = Openness to Experience

EX = Extraversion

NE = Neuroticism

The equation formed above indicates that students’ performance is 0.8079 without any other factors affecting it and all of the variables except for one, are having a positive relationship with the students’ performance. Therefore, students’ performance is expected to increase by 0.5882 when conscientiousness increase by 1; increase by 0.1385 when agreeableness increase by 1; increase by 0.1616 when openness increase by 1; increase by 0.0041 when extraversion increase by 1; and lastly, decrease by 0.0971 when neuroticism increase by 1. To sum up, conscientiousness has the biggest influence on student’s performance among the other variables and it is followed by openness to experience, agreeableness, neuroticism and lastly extraversion.

4.4.2.2 Multicollinearity

According to Saunders et. al. (2009), tolerance value and variance inflation factor (VIF) are other common measures for multicollinearity problem other than correlation coefficients. The rule of thumb indicating a multicollinearity problem is when the tolerance value is 0.10 or below and VIF value is 10 or above (Saunders et. al., 2009). Based on table 4.18, there is no indication of multicollinearity problem as the tolerance and VIF values for all independent variables were greater than 0.10 and lesser than 10 respectively.

4.4.2.3 Test of Significance

Table 4.19 : Multiple Linear Regression

Independent Variables

Parameter

Estimate

Pr > |t|

Hypothesis

Supported/Not Supported

Conscientiousness

0.5882

<.0001

H1

Supported

Agreeableness

0.1385

0.0042

H2

Supported

Openness to Experience

0.1616

0.0025

H3

Supported

Extraversion

0.0041

0.9177

H4

Not Supported

Neuroticism

-0.0971

0.0325

H5

Supported

Source: Developed for the research

H1: There is a positive relationship between conscientiousness and students’ performance.

Based on the analysis above, the hypothesis (H1) was supported. The p-value of conscientiousness which is less than 0.05 and parameter estimate of 0.5882 indicates the correlation between conscientiousness and students’ performance is statistically significant and positively related.

H2: There is a positive relationship between agreeableness and students’ performance.

Based on the analysis above, the hypothesis (H2) was supported. The p-value of agreeableness is 0.0042 which is less than 0.05. This indicates the correlation between agreeableness and students’ performance is statistically significant. Whereas parameter estimate of 0.1385 reflects a significant positive relationship between two variables.

H3: There is a positive relationship between openness to experience and students’ performance.

Based on the analysis above, the hypothesis (H3) was supported. The p-value of openness to experience is 0.0025 which is less than 0.05 indicates the correlation between openness to experience and students’ performance is statistically significant. Whereas parameter estimate of 0.1616 reflects a significant positive relationship between two variables.

H4: There is a negative relationship between extraversion and students’ performance.

Based on the analysis above, the hypothesis (H4) was not supported. The p-value of extraversion is 0.9177 which is greater than 0.05. This indicates the absence of significant relationship between extraversion and students’ performance.

H5: There is a negative relationship between neuroticism and students’ performance.

Based on the analysis above, the hypothesis (H5) was supported. The p-value of neuroticism is 0.0325 which is less than 0.05 indicates the correlation between neuroticism and students’ performance is statistically significant. Whereas parameter estimate of -0.0971 reflects a negative relationship between two variables.

4.5 Conclusion

Results of data collected from questionnaire survey were generated by using SAS version 9.3 and have been analyzed and interpreted. The analysis included descriptive analysis, scale measurement, and inferential analysis. Pearson Correlation Analysis and Multiple Regression Analysis were also conducted to examine the relationship between independent variables and dependent variable. In the last chapter, discussion, conclusion, and implications of the study will be presented accordingly.