國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
30
Chapter 4 Discussion and Research Finding
Sharing Economy is becoming an increasingly popular business phenomenon (e.g., UBer, Airbnb, Couch Surfing) to connect businesses and organizations with their target demographics. However, not all implementations are effectively engaging with their audience. Considering these challenges, I can infer multiple implications from the research results for the strategic management of sharing economy.
4-1: Testing the Correlation between Intention and Attitude
Based on the hypothesis settings mentioned previously, I took the first look at the relationship between the “Intention to participate bike-sharing” and the “Attitude toward sharing”. A Chi-Square test is conducted as follows.
Chi-Square Tests
Value df Asymptotic Significance (2-sided) Pearson Chi-Square 1071.483a 444 .000
Likelihood Ratio 566.937 444 .000
Linear-by-Linear Association
170.733 1 .000
N of Valid Cases 278
a 494 cells (100.0%) have expected count less than 5. The minimum expected count is .01.
<Table 7> Chi-Square Tests – Attitude & Intention
The P-Value is 0 while the Pearson Chi-Square value is 1071.483. I have enough evidence to say that the two variables are strongly correlated. Here I prove again the strong relationship between the attitude toward sharing and the intention to participate as the model provided by Hamari, Sjöklint, and Ukkonen (2015).
4-2: Hypothesis Testing
As it was mentioned in the hypotheses, I am curious about if there are significant
‧
differences among participants’ intentions, attitudes, motivations and preferences under respective sharing circumstances. A one-way ANOVA test was conducted and the result is shown below, in which the clustering factor is “Sharing Scenario”.
Intention to participate I * Scenario Crosstabulation Count
Scenario
Gift giving Reciprocity Market Exchange Total
Intention to participate I yes 73 70 72 215
no 20 24 19 63
Total 93 94 91 278
<Table 8> Cross Table - Scenarios
ANOVA
<Table 9> ANOVA Tests – Scenarios with Intention, Attitudes & Motivations
Descriptives
‧
<Table 10> Descriptive Statistics Table – Scenario with Intention, Attitudes &
Motivations
These three tables above show that there is not enough evidence for me to reject the null hypothesis. It means there is no significant difference among each variable caused by respective sharing scenarios. To see the picture clearer, a paired t test is conducted as below.
Test of Homogeneity of Variances
Levene Statistic df1 df2 Sig.
Intention to participate bike-sharing 1.361 2 275 .258
‧
<Table 11> Paired Tests – Homogeneity of Variances And it is followed by the ANOVA test.
ANOVA
<Table 12> Paired t-tests
All tables above show the same result. There is not enough evidence to support our speculation that participant’s intentions, attitudes, or motivations would differ under respective sharing scenarios. Our previous alternative hypotheses, H1a and H1b, therefore have no evidence to be supported. In another word, there is no significant difference caused by gift giving, reciprocity, or market-exchange bike-sharing.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
34
Moreover, we see not enough evidence to support the speculation that sharing scenario influence sharing attitudes.
4-3: The Role of The Attitudes
I took a look at the relationship among the motivations, the intention to participate sharing, the attitude toward sharing, toward the intention to participate in bike-sharing. Linear regression was utilized in this part that the variables had been summarized through Likert scale average method.
Variables in the Equation
B S.E. Wald df Sig. Exp(B)
Step 0 Constant -1.228 .143 73.414 1 .000 .293
<Table 13> Regression Variables
In the table above it shows the p-value is 0.000. I have enough evidence to say there is at least one variable influence significantly the intention.
Model Summary
Step -2 Log likelihood Cox & Snell R Square
Nagelkerke R Square
1 210.059a .270 .411
a. Estimation terminated at iteration number 6 because parameter estimates changed by less than .001.
<Table 14> Regression Summary
The Cox & Snell R2 that equals to 0.27 > 0.1, which means it is proper to conduct the following regression. And the overall percentage correct of the regression is 81.7%
that is shown in Table 15.
Classification Tablea
‧
<Table 15> Explanatory Power
Table 16 below shows the linear regression coefficient of variables to sharing intention. a Dependent Variable: Intention to participate bike-sharing
<Table 16> Linear Regression – Intention
By checking the P-Values, only the influence of the attitude toward sharing is strong and significant under a 95% significance level. Influence of the enjoyment motivation and the materialism (centrality) is significant but the powers are weak. The influence led by scenario is proven again not significant.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
36
In this session, I took another linear regression to exam the relationship among motivations and materialism toward the sharing attitude. Table 9 below shows the result.
Coefficientsa
Model Unstandardized
Coefficients
Standardized Coefficients
t Sig.
B Std.
Error
Beta
1 (Constant) .554 .310 1.785 .075
M_Sustainability .347 .059 .363 5.866 .000***
M_Exonomic .240 .055 .228 4.368 .000***
M_Image .033 .056 .040 .588 .557
M_Enjoyment .275 .065 .278 4.222 .000***
Mt_Success -.089 .073 -.074 -1.215 .225
Mt_Centrality .071 .076 .055 .935 .351
Mt_Happiness .011 .083 .008 .126 .900
Scenario .011 .065 .007 .170 .865
a. Dependent Variable: Attitude toward sharing
<Table 17> Regression – Attitude
As it is shown in table 9, the awareness of sustainability, the economic benefit, and the enjoyment motivations are significant in this regression model. There is not enough evidence to support that materialism would lead to different participants’
attitudes.
Summarized the hypothesis testing above, I thereby come up the following SEM analysis of research model.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
37
4-4 Further Exploratory Study
By far I have found the brief result of figuring out the factors that influence the attitudes and intention. The statistical analysis also denies all hypotheses about the impacts led by sharing scenarios. I am, however, still curious about if there exist any other influential factors. By checking factors backward and forward, one after another, I finally put my eyes on the difference of participants’ nationalities. I would like to figure out whether the culture difference among participants will lead to a significant difference in their intention, attitude and motivation factors.
ANOVA
Sum of Squares
df Mean
Square
F Sig.
Intention to participate Between Groups 45.401 6 7.567 2.922 .009**
‧
Attitude toward sharing Between Groups 30.535 6 5.089 2.784 .012*
Within Groups 495.409 271 1.828
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
39
Total 333.912 277
<Table 18> ANOVA analysis of different sampling countries
Surprisingly, I found a very significant impact caused by the difference between participants’ nationalities. The influence is so significant that almost every p-values in this ANOVA test are lower than 0.01. Since there is so far no previous research discussing the impact due to culture difference to participant’s attitude and behavior in sharing economy, this finding reveals a very inspirational direction to future studies.