To test the hypotheses, a series of regression analyses were conducted. The regression model in Table 6 examined various blocks of predictors and moderating variables toward green purchase intention and actual green purchase as the criterion variables.
Table 6
Predicting Purchase Intention & Actual Purchase (n = 602)
Criterion Variable
Studying Status (studying and working) .038 .004
Allowance and/or Income (with allowance/ income) -.042 -.008
Living Condition (not staying with parents) -.020 .042
Incremental R² (%) 2.0%✝ 2.3%*
Block 2
Ecological Knowledge .006 -.031
Ecological Affect .396*** .084✝
35
Total R² (%) 43.9% 29.1%
Notes:
(1) ✝p<.10; *p<.05; **p<.01; ***p<.001
(2) Entries are standardized regression coefficients
The first block that consisted of demographic controls such as gender, age, religion, studying status, allowance and/or income, and living condition were accounted for about 2 percent of the variance for purchase intention and 2.3 percent of the variance for actual purchase. However, out of the six control variables, only religious belief exerted a statistically significant effect (β = .074, p < .019). Respondents who have a religious belief were more likely to have higher purchase intention.
Based on Table 6, regression results showed that ecological knowledge was not associated with either green purchase intention (β = .006, p = ns) or actual green purchase (β
= -.031, p = ns). Thus, in response to RQ1, the relationship between ecological knowledge with green purchase intention and actual green purchase was not positively related. Green purchase intention and behavior is irrespective to respondents’ knowledge about environment.
On the other hand, lending support to H1, results revealed that ecological concern significantly impacted green purchase intention (β = .396, p < .001) and actual green purchase (β = .084, p < .087). Hence, the data supported H1 in that ecological concern was positively correlated to purchase intention and actual purchase, as presented in Table 6.
Respondents that displayed higher concern for the environment tended to exhibit higher green purchase intention and behavior.
Next, results in Table 6 indicated that man-nature orientation also strongly elicited green purchase intention and actual green purchase. Man-nature orientation was positively associated to green purchase intention (β = .228, p < .001) as well as positively impacted actual green purchase (β = .118, p < .019). Accordingly, these positive beta coefficient support H2 that the relationship between man-nature orientation with green purchase
36
intention and actual green purchase was positively related. A man that believed that one should live in peace with nature tends to engage in more green purchasing intention and behavior.
Ecological knowledge, ecological concern, and man-nature orientation that makes up the second block contributed a significant 40.3 percent of the variance for purchase intention and 16 percent variance for actual purchase.
Table 6 shows that price sensitivity significantly affected green purchase intention (β = .150, p < .001) and price sensitivity accounted for about 1.6 percent of the variance for purchase intention. The result implied that the level of price sensitivity of a consumer will influence his or her intention to purchase. However, price sensitivity was negatively associated with actual green purchase (β = .026, p = ns). Results implied that a person’s sensitivity to price did not influence his or her actual purchase. Nonetheless, purchase intention alone reflected a strong positive correlation with actual purchase in Table 6 (β = .402, p < .001) and this indicated that a respondent’s actual behavior is influenced by his or her intention. Both price sensitivity and purchase intention accounted for 9.6 percent of the variance for actual purchase.
The last block in Table 6 examined the four interactions for actual purchase, including (1) price sensitivity with purchase intention, (2) studying status with purchase intention, (3) allowance and/or income with purchase intention, and (4) living condition with purchase intention. These interactions explained 1.2 percent of the variance. These tests in Table 6 are in response to H3 to examine if price sensitivity will moderate the relationship between purchase intention and actual purchase. Results for price sensitivity as a moderating variable for purchase intention against actual purchase reflected positive relationship (β = .082, p = .062). Hence, supporting H3 that price sensitivity had moderating effect on purchase intention and actual purchase.
37
Findings from Table 6 reported a negative relationship between study status with purchase intention (β = .014, p = ns). Results indicated study state had no moderating effect on purchase intention and actual purchase. The interaction between allowance and/or income with purchase intention reported positive relationship (β = .066, p = .064), supporting H4 that purchase intention and actual purchase was moderated by allowance and/or income. On the other hand, living condition with purchase intention reported negative relationship (β = .055, p = ns), reporting that a respondent’s living condition did not moderate the relationship between intention and their actual purchase behaviors.
This study took into account marginal significant relationships; that is, accepting p values at the .10 level as statistically significant. This is mainly because this study is exploratory in nature and serves as a precursor to further studies in identifying relationships differences and statistical phenomena at a groundwork level.
To visualize the moderation effects, Andrew Hayes’ PROCESS macro plug-in was incorporated into SPSS. Model 1 from the macro was employed to test the hypotheses that price sensitivity and allowance and/or income would moderate the relationship between purchase intention and actual purchase. Sociodemographic variables were included as covariates to test the interactions between those conditions. The output provided values of the dependent variable when the independent variable and moderating variable are one standard deviation below the mean (low), at the mean (medium), and one standard deviation above the mean (high). It is noteworthy that this study only used the low and high values when producing interaction plots in Figure 2 and 3.
Purchase intention exerted a positive effect on actual purchase for people with various levels of price sensitivity in Figure 2. Yet, the effect was stronger for individuals with higher price sensitivity. In other words, those with higher price sensitivity were more likely to buy green products than people with lower price sensitivity when purchase intention was high.
38
Figure 3 shows the positive relationship between purchase intention and actual purchase with the moderating effects of the different measure of allowance and/or income.
The effect was stronger for respondents with higher allowance and/or income. It was found that respondents with higher allowance and/or income displayed higher actual purchase than those with lower allowance and/or income when their purchase intention was high.
Figure 2. Moderated regression for price sensitivity.
Figure 3. Moderated regression for allowance and/or income.
7.89
39
In conclusion, as observed in Figure 4, ecological knowledge had no significant influence on both purchase intention and actual purchase. In response to RQ1, ecological knowledge was negatively related to purchase intention and actual purchase. On the contrary, ecological concern, and man-nature orientation were both positively correlated to purchase intention and actual purchase, supporting H1 and H2. In Figure 5, price sensitivity and allowance and/or income were statistically reported to be strong moderating variables between purchase intention and actual purchase, supporting H3 and H4.
Figure 4. Mapping the pathway to purchase intention and actual purchase.
Figure 5. Mapping the pathway of price sensitivity and allowance/ income as moderating variables.
Notes:
(1) The coefficients of control variables are not shown (2) ✝p<.10; *p<.05; **p<.01; ***p<.001
(3) All entries are standardized coefficients based on OLS regression path analyzes Ecological
40
Chapter Five – Discussion
5. 1 Discussion
The aim of this study was to examine if ecological knowledge, ecological concern, and man-nature orientation affect Malaysian university students’ purchase intention and actual purchase, and whether this effect was moderated by price sensitivity and allowance and/or income. Participants were required to answer five to ten minutes of self-administered questionnaires which tested upon the six key constructs of environment.
Since all respondents were in their tertiary education, it is not surprising that these individuals with higher levels of education had relatively good factual knowledge concerning the environmental impact with (M = 7.02, SD = 2.36), highest is ten point. Although there is wide acceptance of the assertion that an increase in knowledge will influence attitude, however, this study finds that ecological knowledge had no positive correlation with purchase intention and actual purchase. The first research question probed if a student’s green purchase intention and actual green purchase were influenced by his or her level of ecological knowledge; the result indicated that it is not positively related.
These results contradicted assertions by Bamberg and Moser (2007); Chan and Lau (2000); Cheng and Wu (2015); D’Souza et al. (2006); Mostafa (2007); Smith and Paladino (2010); Vicente-Molina, Fernandez-Sainz and Izagirre-Olaizola (2013) which claim that subjects that demonstrate higher ecological knowledge will be more likely to partake in pro-environmentally behavior. Nevertheless, this study presented statistically no difference for ecological knowledge in influencing green behavior. Knowledge about the environment was irrespective to green purchase intention or actual green purchase. Both variables were negatively related.
41
Kaiser et al. (1999) alleged that this is consistent to other studies that found either no relationship between factual environmental knowledge and ecological behavior (Maloney &
Ward, 1973; Maloney et al., 1975), or at best a moderate relationship (Oskamp et al., 1991;
Smythe & Brook, 1980). The relationship between ecological knowledge and ecological behavior only gets stronger in the instance when it is about what can and cannot be done and how to improve the ecological system rather than factual knowledge about environmental issues (Smith & Fortner, 1994). Tanner and Wölfing-Kast (2003) emphasize the significance of action-related knowledge since it is pivotal for consumers to comprehend what actions they could make to protect the environment. Perhaps other future research opting for action-related knowledge rather than factual knowledge that is adopted by this study might generate better correlation with purchase intention and actual purchase.
The first hypothesis that predicts purchase intention and actual purchase are influenced by ecological concern was supported. Respondents with higher level of ecological concern exhibited greater purchase intention and actual green purchase. This results strongly suggest that subjects that feel emotional for the environment or hold greater level of concern will demonstrate a stronger inclination to purchase green products, and they will display higher actual green purchasing. This study endorses previous research findings concerning the strength of relationship between environmental concern with green purchase intentions (Newton, Tsarenko, Ferraro, & Sands, 2015). The results suggest that a consumers’ degree of emotionality on ecological issues positively influences their green purchase response.
As explained by Bohner and Dickel (2011), consumers will probably be influenced by what is important to them, or by what they feel most close to heart. The results of this study were consistent to the claims expressed by Bohner and Dickel, because it was found that there is a positive relationship between environmental concern and green purchase.
42
Respondents for this study exhibited high ecological concern with (M = 17.31, SD = 3.42), with five, five-point items.
The second hypothesis that predicted man-nature orientation would positively influence environmentally friendly purchasing behavior was statistically supported. Results supported that subjects with a high man-nature relationship displayed a positive relationship with green purchase intention and actual purchase. A man that has a closer relationship with nature has a tendency to influence his or her eco-friendlier purchases; so this does not come as a surprise that this relationship is positively related for Malaysians, as Asians have been practicing life compromising with nature (Feigenbaum & Manning, 2012).This finding is consistent to Chan and Lau’s (2000) notion that the Chinese place great emphasis on living in harmony with nature and appreciating all that nature has provided by not abusing the environment. This belief transcended boundaries, and this same sentiment was reflected by the Chinese population in Malaysia with plenty of respondents displaying this orientation (M = 12.11, SD
= 2.6), with three items and five as the highest agreement scale.
This orientation dictates that those influenced by man-nature qualities and beliefs would behave accordingly to maintain and co-inhabit with nature. Therefore, the idea of coexistence is expressed to actual commitment to ecological issues. This finding provided further insight on the importance of this traditional cultural value in shaping Malaysian students’ eco-friendly behavior, namely green purchase.
The third hypothesis from the proposed framework suggested that price sensitivity moderates the connection between green purchase intention and actual green purchase was statistically supported. The result implied that the level of price sensitivity of a consumer will influence actual green purchase; a form of pro-environmental behavior. The finding supported claims by Mandese (1991) that a product’s price is highly weighed in by consumers when they purchase eco-friendly goods.
43
Generally, it is assumed that a consumer who displayed higher price sensitivity would most likely shy away from eco-friendly products that were usually more expensive, whereas a consumer with lower price sensitivity would most likely show stronger purchase intention and would engage in actual purchase. However, the pattern of the results revealed an interesting relationship. Findings demonstrated the opposite of assumption, respondents with higher price sensitivity reported higher actual purchase than those who were lesser sensitive in price when the purchase intention was great.
As pointed out by Bolton and Lemon (1999), customers appraise what is “fair,” “right,”
or “deserved” for the price of a product. Hence, even though a consumer might be highly price sensitive, they still display great purchase intention and engage in more green purchase because they perceived the cost of owning the green product versus the money sacrificed in exchange of the trade as fair and deserving. According to Har et al. (2011), this is supported by Grunert and Kristensen (1992) study that 60 percent of consumers were willing to pay premium price for green products.
The findings would help marketers to comprehend Malaysian students level of price sensitivity with many of them possessing price sensitive behaviors (M = 40.84, SD = 8.32), with eleven items on a five-point scale, and five is the highest agreement to being price sensitive. It might be because half of the respondents, 57.3 percent were not living with their parents and hence, they spent frugally and were being thrifty in their spending. Consequently, being price sensitive influenced their purchasing decision.
According to Numbeo.com (2017), a single person’s monthly living expenses in Kuala Lumpur, the capital city of Malaysia is approximately RM 2,011 or 465 USD or 14,000 NTD.
However, 40 percent of the students had no monthly allowance and/or income and almost 50 percent of them had below RM 1,000 per month or an equivalence of approximately 230 USD or 6,950 NTD, which is half of the average monthly living expenses estimated for a
44
single person. The sample of this study was considered as a lower allowance and/or income group of consumers. Consequently, respondents tended to be more sensitive to price, and price played an important role in their final buying decision. It is noteworthy that the findings might differ if other researchers sample their respondents from a higher allowance and/or income group of consumers.
Likewise, the last hypothesis was supported as well. Allowance and/or income of students had a significant influence in moderating the relationship between green purchase intention and actual green purchase. This finding validated the point that economic merits are crucial to determined consumers’ ability to buy (Wang et al., 2001). Expectedly, results revealed that higher allowance and/or income respondents with higher purchase intention generated higher actual purchase.
5.2 Limitations
Studies reported here should be interpreted in light of the following shortcomings.
Firstly, this study adopted purposive sampling. Even though the respondents were made up of Malaysian university students from diverse socio-demographic profiles, the sample was not drawn using random sampling method because a truly representative sample is hard to achieve. Therefore, the sample might not represent the target population as a whole, and the findings of this study are subjected to sampling bias.
Secondly, the hypotheses for actual purchase were assessed within the context of three types of green product categories obtained from pilot study (beauty, health, and grocery), which represented only a small fraction of the entire range of green products. Replicated and extended reviews on other green products may additionally help to enhance and expand the comprehension of green purchase intention and actual green purchase factors.
45
Thirdly, because the choice of variables included in this context may not be exhaustive, in the future, other possible factors that might affect green purchase intention and actual green purchase could be included in this theoretical framework.
Lastly, there is the likelihood that respondents may have professed to be ecologically concerned even when they were not. Respondents tend to answer in favor to the researcher and to create a positive response (Rice, Wongtada, & Leelakulthanit, 1996). This potential social desirability bias must be acknowledged and treated as a shortcoming of this present study. This bias may be additionally heightened when considering that the current sample is relatively more educated, and thus might be more alert to not create social undesirability responses or ideas.
It should also be remembered that this research only investigated university students in mostly Kuala Lumpur. It is likely that green purchase intention might differ between states in Malaysia. Consequently, outcomes and implications should be generalized cautiously until further research attest the findings.