CHAPTER 3: METHODOLOGY
3.3 M EASURES
3.3.2 Four factors of Planned Behavior
Attitudes, Subjective norms, Perceived behavioral control, and Behavioral intention were measured based on the key items in a research of Chow and Chen (2009) with five-point Likert-type scale, ranging from one (strongly disagree) to five (strongly agree).
Attitude toward behavior
An individual's positive or negative evaluation of self-performance of the particular behavior. The concept was the degree to which performance of the behavior was positively or
22
negatively valued. It was determined by the total set of accessible behavioral beliefs linking the behavior to various outcomes and other attributes. Recycling attitude was measured by asking five questions of the form (item 1-5), (1) “For me, recycling is good”, (2) “For me, recycling is beneficial”, (3) “For me, recycling is pleasant one”, (4) “For me, recycling is valuable”, and (5) “For me, recycling is enjoyable” and taking responses on five-point Likert-type scale (Likert, 1932) anchored by “strongly disagree” and “strongly agree”. The scale used is modified from the researches of Ajzen (2006), and Chow and Chen (2009).
Subjective norms
Subjective norms refer to an individual's perception of social normative pressures, or relevant others' beliefs that he or she should or should not perform such behavior (Ajzen &
Fishbein, 1975). Individuals' elaborative thoughts on subjective norms are perceptions on whether they are expected by their friends, family and the society to perform the recommended behavior. Recycling subjective norm was measured by asking five questions of the form (item 6-10), (6) “Most people who are important to me think that I should recycle”, (7) “It is expected of me if I recycle”, (8) “Many people like me recycle”, (9) “Most people who are important”, and (10) “The people in my life whose opinion I value would approve of me recycle”. Five-point Likert-type scale and the items which used is modified from the researches of Ajzen (2006), and Chow and Chen (2009) were adopted to measure this variable.
Perceive behavioral control
Perceive behavioral control was modified as an individual's perceived ease or difficulty of performing the particular behavior (Ajzen, 2002a; Zhao, Sayeed, Cappella, Fishbein, Hornik, & Ahern, 2006). Recycling perceive behavioral control was measured by asking four questions of the form (item 11-14), (11) “I have totally controlled in recycling”, (12) “I can improve recycling in the forthcoming weeks”, (13) “I definitely could improve recycling in the forthcoming weeks”, and (14) “It is mostly up to me whether or not I improve recycling in the forthcoming weeks”. Five-point Likert-type scale was constructed to measure this variable and the key questions that adopted from the researches of Ajzen (2006), and Chow and Chen (2009).
Behavior intention
Behavior intention was an indication of an individual's readiness to perform a given behavior. It was assumed to be an immediate antecedent of behavior (Ajzen, 2002b). It was based on attitude toward the behavior, subjective norm, and perceived behavioral control, with each predictor weighted for its importance in relation to the behavior and population of interest. Recycling behavior intention was measured by asking three questions of the form
23
(item 15-17), (15) “I will intend to recycle”, (16) “I will try to recycle”, and (17) “I will plan to recycle”. Five-point Likert-type scale and the items used is modified from the research of Chow and Chen (2009) were offered to measure the variable.
Moreover, the study also checked an actual behavior, both of past behavior and present behavior of the survey participants to watch for if the participants did have or did not have an actual behavior toward recycling with two questions of the form (item 18-19), (18) “I often recycle”, and (19) “I often have recycled”. Five-point Likert scale and the items that modified from the research of Chow and Chen (2009) were offered to measure the variable. The questionnaires asked through the items, which are referred to the researches of Ajzen (2006), and Chow and Chen (2009).
The survey form is reproduced in Appendix 3 - Sample questionnaire in English for formal test and Table 10 summarizes the items measuring Attitudes, Subjective Norms, Perceived Behavioral Control, Behavioral Intentions and Actual Behavior.
Table 10: Measurement of variables Survey
items
Dependent variables Measurement References 5 items Attitude toward behavior
5 items Subjective norms
4 items Perceive behavioral control 3 items Behavior intention
2 items Actual Behavior
5-point Likert-type scale 3.4 Method of data analysis
Statistical Methods
The collected data in this study were analyzed using SPSS data analysis software. In the process of data analysis, the statistics method involves means of analysis of frequency, descriptive statistics and reliability analysis (Cronbach’s alpha), and regression. All analyses were made by SPSS software version 16 (SPSS Inc., 2008). The level of significance was set to p≤.05 (95% confidence interval) (See Table 11).
24
Table 11: Methods of data analysis
Hypothesis Methods H1: Attitude directly affect to behavioral intention toward recycling. Regression
H2: Subjective norm directly affect to behavioral intention toward recycling. Regression H3: Perceived behavioral control directly affect to behavioral intention
toward recycling. Regression
H4: Repetitive communicating of a message had a direct effect on attitude toward recycling so it had an indirect effect on behavioral intention toward recycling through attitude toward recycling.
t-test
H5: Repetitive communicating of a message had a direct effect on subjective norm toward recycling so it had an indirect effect on behavioral intention toward recycling through subjective norm toward recycling.
t-test
H6: Repetitive communicating of a message had a direct effect on perceived behavioral control toward recycling so it had an indirect effect on behavioral intention toward recycling through on perceived behavioral control toward recycling.
t-test
H7: Repetitive communicating of a message had a direct effect on behavioral intention toward recycling.
In statistics, regression analysis is used when the focus is on the relationship between a dependent variable and one or more independent variables. Regression analysis helps understand how the typical value of the dependent variable changes when any one of the independent variables is varied. Regression analysis is also helping understand and explore the forms of relationships which among the independent variables are related to the dependent variable (Aldrich, 2005; Yang, 2009). In this study, independent variable was repetitive message and dependent variables were attitude, subjective norm; perceive behavioral control, and behavioral intention toward recycling.
T-test analysis
The t-test analysis is used to test if the means of two groups are statistically different from each other. A t-test is any statistical hypothesis test in which the test statistic follows a Student's t distribution, if the null hypothesis is supported. This test is only used when both:
the two sample sizes are equal and small (Zimmerman, 2004; Fay & Proschan, 2010). This study used t-test analysis appropriate to compare the means of two groups: repetitive message effect to planned behavior’s variables at the first time and repetitive message effect to planned behavior’s variables at the third time.
Descriptive statistics
25
The study used Descriptive statistics to describe the basic features of the data.
Descriptive statistics are used to present quantitative descriptions. It means describing what is or what the data shows that provided simple summaries about the sample and the measures.
For example, the study used percentages to describe the: percentage of students in different genders and percentage of students in different ages. One of three major types of estimates of central tendency is mean also be used to compute average values of the responses.
Reliability test of constructs (Cronbach’s α)
The study accessed each decision variable by performing within-scale factor analysis.
This showed that all measurement items converged on the decision variables, with each factor loading having a value not less than 0.6. All decision variables demonstrated unidimensionality.
Cronbach’s alpha (α) was used to assess the internal consistency of the proposed decision variables. Table 12 presents a summary of loading ranges and α reliability estimates for each decision variable identified and used. All α values of the constructs greater than 0.7;
thus, they were considered reliable constructs. The reliability statistics showed, Cronbach's Alpha = .938 > 0.7; thus, five-point Likert-type scale was considered reliable constructs.
There were no items deleted because they were all greater than 0.7.
Table 12: The reliability of dependent variables
Variables Cronbach's
Attitude toward recycling .882 5
Subjective Norm toward recycling .867 5
Perceived Behavioral Control toward recycling .837 4
Behavior Intention toward recycling .874 3
Actual Behavior toward recycling .894 2
.938
To compare the responses of students for nineteen different items, the study involved a one factor (message repetition) x three levels of message repetition (round one, round two, round three) x one level of repetition (one time exposure). The five-point Likert-type scale (strongly disagree, disagree, fair, agree, strongly agree) was used for nineteen questions that measure attitude, subjective norm, perceived behavioral control, behavioral intention toward recycling, and actual behavior toward recycling through item one to item nineteen. Item twenty asked information about age, item twenty one was for gender information, and item
26
twenty two required the student identification number for checking their attendant in three rounds and their school year.
Table 13: Mean values of dependent variables
Mean Mean values
Attitude toward recycling (Q1-5) 3.90 Agree
Subjective Norm toward recycling (Q6-10) 3.45 Agree Perceived Behavioral Control toward recycling (Q11-14) 3.55 Agree Behavior Intention toward recycling (Q15-17) 3.85 Agree Actual behavior toward recycling (Q18-19) 3.82 Agree
Up to the results, the respondents’ attitude toward recycling, subjective norm toward recycling, perceived behavioral control toward recycling, behavior intention toward recycling, and actual behavior toward recycling were all agree.
Table 14 reports mean values of round one and round three and Table 15 shows what the mean values signify.
27
Table 14: Mean values of each item round one and round three
Items Mean
1 Values Mean
3 Values
Attitude toward recycling 3.86 Agree 3.99 Agree
[Q1] For me, recycling is good. 4.38 Strongly
agree 4.27 Strongly agree [Q2] For me, recycling is beneficial. 4.21 Strongly
agree 4.17 Agree [Q3] For me, recycling is pleasant one. 3.35 Fair 3.71 Agree [Q4] For me, recycling is valuable. 4.04 Agree 4.10 Agree [Q5] For me, recycling is enjoyable. 3.33 Fair 3.69 Agree Subjective Norm toward recycling 3.26 Fair 3.69 Agree [Q6] Most people who are important to me think that I
should recycle. 3.17 Fair 3.77 Agree
[Q7] It is expected of me if I recycle. 3.23 Fair 3.75 Agree [Q8] Many people like me recycle. 3.10 Fair 3.73 Agree [Q9] Most of people who are important to me recycle. 3.19 Fair 3.48 Agree [Q10] The people in my life whose opinion I value
would approve of me recycle. 3.62 Agree 3.71 Agree
Perceived Behavioral Control toward recycling 3.35 Fair 3.78 Agree [Q11] I have totally controlled in recycling 3.33 Fair 3.85 Agree [Q12] I can improve recycling in the forthcoming
weeks 3.33 Fair 3.77 Agree
[Q13] I definitely could improve recycling in the
forthcoming weeks. 3.29 Fair 3.69 Agree
[Q14] It is mostly up to me whether or not I improve
recycling in the forthcoming weeks. 3.46 Agree 3.79 Agree Behavior Intention toward recycling 3.67 Agree 4.10 Agree [Q15] I will intent to recycle. 3.62 Agree 4.19 Agree
[Q16] I will try to recycle. 3.81 Agree 4.10 Agree
[Q17] I will plan to recycle. 3.56 Agree 4.02 Agree
Table 15: Mean values signify
Mean values Mean values signify References 1.00 - 1.80 Strongly disagree
1.81 - 2.60 Disagree 2.61 - 3.40 Fair 3.41 - 4.20 Agree
4.21 - 5.00 Strongly agree
Likert scaling (1932); Wuensch, (2005); Dawes, (2008)
Likert (1932); Wuensch, (2005); Dawes, (2008) showed a way of generating a quantitative value (numerical) to a qualitative questionnaire (strongly disagree, disagree, fair, agree, strongly agree). Five point scale values are assigned to each category and a mean figure
28
for all the responses is calculated via the sum of the products of the categories assigned value times the number of respondents for that category, divided by the total number of respondents.
For example, mean of respondents’ attitude in round one was calculated as a following formula: Total number of respondents = 48, assigned values are; strongly disagree = 1, disagree = 2, fair = 3, agree = 4, strongly agree = 5; respondents selecting following categories were; disagree = 2, fair = 5, agree = 14, strongly agree = 27. The quantitative mean
= ((2*2) + (5*3) + (14*4) + (27*5))/48 = 4.38 (see Table 14)
29
Chapter 4: RESULTS AND DISCUSSIONS
4.1 Descriptive analyses
The conducted sample design was described in Table 16.
Table 16: Sample design
Sample Department School year Three-rounds attendant
Time interval Student Hospitality & Tourism
Management
Freshman and sophomore
48 students 2 weeks There were 60.4% female respondents and 39.6% male respondents in the sample with 10.4% aged from 21-25 and 89.6% aged from 15-20. As this was a convenience sample, there was no attempt to survey equal number of males and females. The study did not compare the difference between genders.
Table 17: Description of respondents’ age and gender of the sample Age (N=48)
Scale items Frequency Percent
"15-20" 43 89.6
"21-25" 5 10.4
Gender (N=48)
Frequency Percent
"Male" 19 39.6
"Female" 29 60.4
4.2 Attitude, Subjective norm, and Perceived behavioral control directly affects behavioral intention toward recycling.
4.2.1 Attitude directly affected to behavioral intention.
Attitude was a significant predictor of behavior intention (Ajzen, 1991; 2002b; Ajzen
& Fishbein, 2005; Sutton, 2002; Chow & Chen, 2009; Sniehotta, 2009). Attitude toward recycling referred to the respondents’ general feeling of favorableness or unfavorableness for recycling. Therefore, attitudes were said to arise from a combination of respondents’ beliefs about recycling and the evaluations of recycling (Ajzen & Fishbein, 1969, 1970, 1972).
30
4.2.2 Subjective norm directly affected to behavioral intention.
Subjective norm toward recycling was the social pressures on an individual respondent’s behaviors was performed or not performed. Subjective norm toward recycling illustrated that the individual respondent perceived the importance of doing recycling that other individuals or groups have. Thus, the individual respondent’s behaviors were up to the subject recycling where their belief power for the issues was strong (Ajzen, 1991).
4.2.3 Perceived behavioral control directly affected to behavioral intention.
Perceived behavioral control was indeed predictive of behavior (Armitage & Conner, 2001; Hagger, Chatzisarantis & Biddle, 2002; Sheeran, 2002). The perceived behavioral control toward recycling meant the respondents can control a recycling behavior that they adopted from the received ideas and influenced by outside factors such as message.
Table 18 showed the results of the hypothesis testing of the structuring relationships between Planned Behavior Theory’s variables. That is a hierarchical regression analysis with values of R2 and ∆R2. ∆R2 is the change in R2, which represents the values contributing to the final R2 when a corresponding decision variable was adopted.
The results show that hypotheses H1, H2, H3 were not supported in time one and in time three, H1 and H3 were supported but H2 was not supported. The corresponding values of the path coefficients in time one are .281, .220, .260 and ∆R2 is .379. The corresponding values of the path coefficients in time three are .363, .130, .429 and ∆R2 is .661. This confirms that the variables attitude toward recycling, subjective norm toward recycling, and perceived behavioral control toward recycling did not affect behavioral intention toward recycling in time one, but the variables attitude toward recycling, and perceived behavioral control toward recycling affected behavioral intention toward recycling in time three and subjective norm toward recycling did not affect behavioral intention toward recycling in time three. There was difference between time one and time three of repetitive communicating of a message. There were no relationships between attitude with intention behavior; and perceived behavioral control with intention behavior in time one but there were relationships between attitude with intention behavior; and perceived behavioral control with intention behavior in time three of repetitive communicating of a message. It means the repetitive communicating of a message three times was more effective than the repetitive communicating of a message one time.
31
Table 18: Summary of structural model with hierarchical regression analysis of time one and time three (n=48)
Time Path Hypothesis R2 Adjusted R2 Path
coefficient* t-value p Results
ATT-BI H1 .281 1.713** .094 Not Note *: Dependent Variable: Behavior Intention toward recycling
Note **: p<0.05
4.3 Repetitive message affected to attitude, subjective norm, perceived behavioral control, and behavioral intention
4.3.1 Repetitive communicating of a message had a direct effect on attitude toward recycling so it had an indirect effect on behavioral intention toward recycling
The respondent’s attitude toward recycling were 3.86 (agree) in round one, and 3.99 (agree) in round three. There was no change of their attitude toward recycling from the first rounds to the third round. So, repetitive communicating of a message was not result in higher attitude toward recycling.
There was no change of the respondents’ attitude toward recycling under the effect of repetitive message one time and three times. It meant the repetitive message had no effect on attitude toward recycling and so it had no effect on recycling intention though recycling attitude.
4.3.2 Repetitive communicating of a message had a direct effect on subjective norm toward recycling so it had an indirect effect on behavioral intention toward recycling
The respondent’s subjective norm toward recycling were 3.26 (fair) in round one, and 3.74 (agree) in round three. The respondents changed their subjective norm toward recycling from “fair” in the first round into “agree” in the third round. Thus, repetitive communicating
32
of a message resulted in higher subjective norm toward recycling.
There was change of the respondents’ subjective norm toward recycling under the effect of repeated message one time and three times. In other words, repetitive message effect resulted in higher subjective norm toward recycling from a comparison of the results in round one and round three. Therefore, it had an indirect effect on behavioral intention toward recycling through subjective norm toward recycling.
4.3.3 Repetitive communicating of a message had a direct effect on perceived behavioral control toward recycling so it had an indirect effect on behavioral intention toward recycling
The respondent’s perceived behavioral control toward recycling for each round were 3.35 (fair) in round one, and 3.78 (agree) in round three. There was a change of the respondent’s perceived behavioral control toward recycling from “fair” in the first round into agree in the third rounds. So, repetitive communicating of a message resulted in higher perceived behavioral control toward recycling.
The repetitive message had a direct effect on perceived behavioral control toward recycling. It resulted in higher perceived behavioral control toward recycling in round three compare to round one. So, there was significant difference between repeated message one time and three times. Besides, the repetitive message had an indirect effect on behavioral intention toward recycling through perceived behavioral control toward recycling.
4.3.4 Repetitive communicating of a message had a direct effect on behavioral intention toward recycling
The responses for behavioral intention toward recycling of the participant were
“agree” in both round one and round three. So, repetitive communicating of a message was not result in higher behavioral intention toward recycling but it had a direct effect on behavioral intention toward recycling. There was significant different between repeated message one time and three times. The participants’ behavioral intention toward recycling was directly affected by repetitive message effect but it was not directly result in higher behavioral intention toward recycling
Compare the results of attitude toward recycling, subjective norm toward recycling, perceived behavioral control toward recycling, and behavior intention toward recycling of round one and round three as following Table 19. In the t-test table, attitude toward recycling’
p = .488 > .05; thus, do not reject H0; there was no significant difference between round one and round three. Subjective norm toward recycling’ p = .012 < .05, perceived behavioral
33
control toward recycling’ p = .009 < .05, and behavior intention toward recycling’ p = .012
< .05; thus, reject H0, there were significant difference between round one and round three or one time and three times.
Table 19: t-test of difference of variables between round one and round three repeated message Subjective Norm toward recycling (Q6-10)
Time 3 48 3.69 .752 Time 1 48 3.35 .729 .009 Perceived Behavioral Control toward recycling
(Q11-14) Time 3 48 3.78 .819
Time 1 48 3.67 .851 .012 Behavior Intention toward recycling (Q15-17)
Time 3 48 4.10 .831 Note*: p<0.05
In sum, repetitive communicating of a message had a direct effect on subjective norm (RE-SN), perceived behavioral control (RE-PBC) and behavioral intention toward recycling (RE-BI) but had not a direct effect on attitude toward recycling (RE-ATT).
Table 20: Summary of the structural model repetitive message affect to attitude, subjective norm, and perceived behavioral control
Path Hypothesis p* Results
RE-ATT H4 .488 Not supported
RE-SN H5 .012 Supported
RE-PBC H6 .009 Supported
RE-BI H7 .012 Supported
Note *: p<0.05
There was no significant change on the respondents’ attitude toward recycling.
However, there were significant change on subjective norm, perceived behavioral control, and behavioral intention toward recycling after the message was repeated three times. Besides, repetitive message had not an indirect effect on behavioral intention through the path repetitive message-attitude toward recycling-behavioral intention toward recycling (RE-ATT-BI). Repetitive message had a direct effect on behavioral intention through the path repetitive message-subjective norm-behavioral intention toward recycling (RE-SN-BI); and repetitive message had a direct effect on behavioral intention through the path repetitive message-perceived behavioral control-behavioral intention toward recycling (RE-PBC-BI).
34
Nevertheless, there was no change in higher attitude and behavior intention for audiences. In
Nevertheless, there was no change in higher attitude and behavior intention for audiences. In