You are on page 1of 62

OL22-WEEK 12

Chapter 11
Continue:
One-Way Anova

Business Research Methods |


One Way ANOVA (Analysis of Variance)
• Helps to examine the significant
mean differences among more
than two groups on an interval or
ratio-scaled dependent variable.

Business Research Methods | 2


Business Research Methods |
One-Way ANOVA

Business Research Methods | 4


One-Way ANOVA

Business Research Methods | 5


One-Way ANOVA

Business Research Methods | 6


One-Way ANOVA

Business Research Methods | 7


One-Way ANOVA

Business Research Methods | 8


One-Way ANOVA

Business Research Methods | 9


HO: Satisfaction not differs among countries
H1: Satisfaction differs among countries

1. Compare: P (0.000) < α (0.05)


2. Decide: Reject H0
3. Conclude: Satisfaction differs
significantly among countries
4. Post-Hoc: Significant pairs are
found between:
• Singapore & China (P=0.022)
• Singapore & India (P=0.037
• Korea & China (P=0.045)

Business Research Methods |


APA Style Interpretation

A one-way ANOVA between-groups analysis of variance was performed to investigate the


mean difference of satisfaction among countries. Based on the table, the result shows a
significant difference of customer satisfaction among countries (p=0.000 < α). The result
indicates that the level of satisfaction is differs among customers from Singapore, China, India
and Korea. This finding is consistent with previous studies (e.g., Hanson, 2013; Md Noor &
Rizuan, 1999; Aimi et al., 2014) that highlighted the role of countries in determining the
customer satisfaction. Based on Post-hoc Multiple Comparison analysis, a significant difference
are found between Singapore & China (P=0.022), Singapore & India (P=0.037) and between
Korea & China (P=0.045).

Business Research Methods |


OL23-WEEK 12

Chapter 11
Continue:
Correlation

Business Research Methods |


Correlation Analysis
• Correlation is a bivariate analysis that measures the strength of association between
two variables and the direction of the relationship.
• In terms of the strength of relationship, the value of the correlation coefficient varies
between +1 and -1.

Strength of Correlation - interpreting 0.1 to 0.29 Weak correlation


Pearson correlation coefficient (r) 0.3 to 0.49 Moderate correlation
0.5 to 0.99 Strong correlation
Cohen’s rule of thumb (Cohen, 1981)

Business Research Methods |


Business Research Methods |
Business Research Methods |
Business Research Methods |
Business Research Methods |
H0: Stress has no relationship with OC
H1: Stress has relationship with OC

1. Compare: P (0.008) < α (0.05)


2. Decide: Reject H0
3. Conclude: Stress has negative and
weak relationship with OC

APA Style Interpretation


Pearson correlation was performed to determine the association between stress and organization
commitment. Based on the Table , the result shows a negative and weak correlation between stress and OC
(p=0.008, r=-0.265). The result indicates that the higher the level of stress, the lower the level of
organizational commitment. This finding is consistent with previous studies (e.g., Hanson, 2013; Md Noor &
Rizuan, 1999; Aimi et al., 2014).

Business Research Methods |


OL24-WEEK 12

Chapter 11
Continue:
Regression

Business Research Methods |


Regression Analysis
• Simple Linear Regression analysis is
used in a situation where one
metric independent variable is
hypothesized to affect one metric
dependent variable.

• Multiple Linear Regression analysis


is used in a situation where more
than one metric independent
variable is hypothesized to affect
one metric dependent variable

Business Research Methods | 20


Multiple Regression Analysis
• Next to test the hypotheses developed for this study, we used a
linear multiple regression analysis. As the norm in multiple
regression analysis, there are 7 seven assumptions that need to be
tested before we can interpret the results (Hair et al., 2018).

• The 7 assumptions tested are normality, normality of the error


terms, constant variance, linearity, multicollinearity, autocorrelation
and finally outliers.

Business Research Methods | 21


Assumptions

1. Normality
2. Normality of error term
3. Constant Variance
4. Linearity
5. Multicollinearity
6. Outliers
7. Autocorrelation

Business Research Methods | 22


1. Normality
Multivariate Normality

Business Research Methods | 23


2. Normality of Error Term
• You need to check that the residuals
(errors) are approximately normally
distributed.

• This can be done (a) a histogram


(with a superimposed normal curve)
and (b) a Normal P-P Plot;

Business Research Methods | 24


3. Linearity
• Types of Relationships
• There needs to be a linear
relationship between (a)
the dependent variable and
each of your independent
variables, and (b) the
dependent variable and the
independent variables
collectively.

Business Research Methods | 25


Linearity

Business Research Methods | 26


4. Constant Variance
• Homoscedasticity

Business Research Methods | 27


5. Multicollinearity

Business Research Methods | 28


Collinearity Check

Business Research Methods | 29


6. Outliers
• There should be no significant • Detect outliers using (a)
outliers, high leverage points or "casewise diagnostics" and
highly influential points. Outliers, "studentized deleted residuals",
leverage and influential points are (b) check for leverage points and
different terms used to represent (c) check for influential points
observations in your data set that using a measure of influence
are in some way unusual when known as Cook's Distance,
you wish to perform a multiple
regression analysis.

Business Research Methods | 30


Outliers
• • Outliers in regression are
observations that fall far from the
“cloud” of points. These points are
especially important because they
can have a strong influence on the
least squares line. Normally outside
± 3 can be considered as outliers

Business Research Methods | 31


Outliers

Business Research Methods | 32


7. Autocorrelation

Business Research Methods | 33


Setting up the Analysis

Business Research Methods | 34


Setting up the Analysis

Business Research Methods | 35


Setting up the Analysis

Business Research Methods | 36


Setting up the Analysis

Business Research Methods | 37


Outputs

Business Research Methods | 38


Outputs

Business Research Methods | 39


1. Normality

Business Research Methods | 40


2. Normality of the Error Term

Business Research Methods | 41


3. Constant Variance (Homoscedasticity)

Business Research Methods | 42


4. Linearity

Business Research Methods | 43


4. Linearity

Business Research Methods | 44


4. Linearity

Business Research Methods | 45


5. Multicollinearity

Business Research Methods | 46


6. Autocorrelations

Business Research Methods | 47


7. Outliers

Business Research Methods | 48


Removing Outliers

Business Research Methods | 49


Removing Outliers

Business Research Methods | 50


After removing Outliers

Business Research Methods | 51


Run the Analysis again – New Results

Business Research Methods | 52


New Results

Business Research Methods | 53


Reporting your Results

Business Research Methods | 54


Multiple Regression Analysis
• Table 4.5 presents the results of multiple regression analysis to test
the hypotheses generated. All the 3 independent variables together
explained an R2 of 0.788 indicating that 78.8% of the variance in
intention to share can be explained by the 3 independent variables.

• Attitude (β = 0.757, p < 0.01), Perceived behavioral control (β =


0.120, p < 0.01) were positively related to intention to share
knowledge while Subjective norm (β = 0.065, p > 0.05) was not
significant. Thus, H1 and H2 of this study are supported while
H3 is not supported.
Business Research Methods | 55
Summary of Findings

Business Research Methods | 56


Business Research Methods |
Business Research Methods |
Business Research Methods | 59
H0: Sales not influenced by price, promotion, quality, and brand
H1: Sales influenced by price, promotion, quality, and brand
1. Compare: P (0.000) < α (0.05)
2. Decide: Reject H0
3. Conclude: Sales significantly influenced by
price, promotion, quality, and brand
4. Model Summary: The model is good since
48.4% (R2=0.484) variation in Sales is explained
by predictors (price, promotion, quality, brand)
5. Individual result:
• Price influence negatively on Sales
(P=0.000, t=-4.318)
• Promotion influence positively on Sales
(P=0.009, t=2.721)
• Quality does not influence significantly on
Sales (P=1.99, > 0.05)
• Brand influence positively on Sales
(P=0.019, t=2.012)
6. Contribution: Highest = Price (Beta=-8.42)
Lowest = Brand (Beta=0.371)
7. Equation: Sales = 42.690 – 5.234 price + 4.561
promotion + 1.358 brand + e
Business Research Methods |
Price DV

Promotion
Sales
Quality

Brand

Business Research Methods |


APA Style Interpretation
• Multiple Linear Regression was performed to determine the best set of predictor variable in predicting sales. The
ANOVA table revealed that the F-statistics (17.066) is large and the corresponding p-value is highly significant
(0.000) or lower than the alpha value of 0.05. This indicates that the slope of the estimated linear regression
model line is not equal to zero confirming that there is linear relationship between sales and the four predictor
variables (price, promotion, quality, brand). This finding is consistent with previous studies (e.g., Hanson, 2013, Md
Noor & Rizuan, 1999; Aimi et al., 2014).

• Based on Model Summary table, the R-squared of 0.484 implies that the four predictor variables explain about
48.4% of the variation in the sales. This is a quite a good and respectable model.

• Based on Coefficient table, three predictor variables which are price (p=0.000 < alpha), promotion (p=0.009) and
brand (p= 0.019 < alpha) were found to be of significance in explaining sales. Meanwhile, quality is not able to
show any significant result (p=0.199 > alpha). The largest beta coefficient is found in price (beta=-4.318) and
followed by promotion (beta=0.448) and brand (beta=0.371. This means that price makes the strongest
contribution to explain the sales.

Business Research Methods |

You might also like