Factors Influencing Household Behavior in Storing WEEE: A Case Study from Java

2.1 WEEE

Electrical and electronic equipment, based on EU Directive 2002/96/EC, that has expired and been disposed of in the discharge stream is known as WEEE [12]. It includes all of its parts, subassemblies, and materials [2, 12, 15, 25]. Electronic products that have reached the waste stream due to their discontinuation of use are referred to as WEEE, or e-waste [6, 26]. Used electronics that are reused, resale, recycled, or disposed of are also regarded as waste or WEEE [6, 26].

The reason why an electronic product is not used anymore can vary, such as it is broken or has out-of-date technology and design. Therefore, it is no longer functional [27]. According to the study [28] in actuality, electronic goods are typically no longer in use even though they are still functional and can be replaced with new ones because users desire new features, the old ones are insufficient for the operator's newest services, or they are simply outdated. As a result, electronic items that are already unused eventually end up as waste, also known as WEEE. WEEE has different characteristics from other wastes. This is because the definition of WEEE is very dependent on the perspective of each person.

WEEE is one of the wastes with the fastest growth of around 8% per year [29]. The amount of WEEE produced worldwide in 2016 was 44.7 million metric tons (Mt), equivalent to 6.1 kilograms per population (kg/inh), compared with 5.8 kg/inh produced in 2014. The amount of WEEE is expected to grow to 52.2 million metric tons, or 6.8 kg/year, in 2021. From 44.7 Mt, only 8.9 Mt or 20% of WEEE is processed properly, and the rest is unknown where to go [13, 30].

Throughout the previous ten years, there has been a notable growth in the amount of WEEE produced in emerging nations, the United States, and the European Union. According to the study [29], WEEE production per person in developing nations like China and India is still only about 1 kg annually, but this amount can rise quickly. Due to their vast populations, these two countries will soon create more WEEE overall than Western nations. In addition, the amount of WEEE in new and The number of developing industrial countries is rising as well due to the import of WEEE from developed countries.

According to certain research, between 50 and 80 percent of the WEEE produced in affluent nations is exported to underdeveloped nations for recycling or reuse [29]. In several European and American countries, the disposal of WEEE is by sending waste to several developing countries in Asia and Africa, such as China, Indonesia, Vietnam and others. About 80% of the total WEEE produced is disposed of or sent to countries in Asia and Africa [31]. With a high rate of EEE penetration, Australia stands as the most extensive and advanced country in the region. In 2014, EEE sales were 35 kilograms per capita [32].

There are six categories of WEEE [2, 13, 24, 30]:

1. Products like refrigerators, freezers, air conditioning units, and heating devices exemplify temperature exchange apparatus.
2. Display screens and monitors, including products like televisions, monitors, laptops, notebooks, and tablets.
3. Lamps, with examples of products such as bulb lights and LED lights.
4. Large equipment, with examples of products such as washing machines, dishwashers, electric stoves, and photocopiers.
5. Small equipment, with examples of products such as calculators, digital scales, microwaves, bakeries, and cameras.
6. Small IT and telecommunication equipment, with examples of products such as cellphones, home phones, GPS (Global Positioning System), printers, and routers.

2.2 Theory of Planned Behavior (TPB)

The Theory of Planned Behavior (TPB) was a hypothesis formulated by Ajzen [19], which is a refinement of the theory of reason action (TRA) proposed by Fishbein and Ajzen [33]. The main focus of this Theory of Planned Behavior is similar to the theory of rationale for action, specifically, people's intent to engage in particular actions. It is believed that intention may be the driving force behind actions. Intention serves as an indication of the level of effort an individual is willing to exert and how determined they are to perform a particular behavior.

TPB (Figure 1), developed by Ajzen from TRA, was developed to overcome measurement weaknesses that caused a weak relationship between attitude and behavior, the conduct articulated by Seymour Epstein in his two articles that drew the interest of psychologists in 1979 and 1980 [34, 35]. Fishbein and Ajzen [33] argued that the low correlation between attitude and behavior is caused by different measurement levels. Attitudes are measured at a very general level, while behavior is measured at a specific level.

1.png

Figure 1. Model of theory planned behavior [19]

In improving the predictive power of attitudes toward behavior, it is necessary to measure attitudes and measure behavior at the same level. Attitudes towards the use of contraceptives (specific) and the use of contraceptives (specific). Reviewing the idea or its forming aspects in TPB was the first step in understanding more about behavioral control, subjective standards, and attitude assessment.

2.3 Partial Least Square (PLS)

Due to its lack of reliance on numerous assumptions, PLS (Partial Least Square) is an extremely effective analytical technique [22, 23]. PLS can be used to explain whether there is a relationship between latent variables or not, as well as to support the theory [36]. The fundamental difference of PLS, which is a variant-based SEM using covariant-based LISREL or AMOS software, is the purpose of its use, whether testing theory or developing theory for prediction purposes. In contrast to covariance-based Structural Equation Modeling (as seen in AMOS, LISREL, and EQS software), the variant-based Partial Least Squares (PLS) method successfully circumvented two significant issues encountered by covariance-based SEM, specifically the challenges of inadmissible solutions and indeterminacy factors [22, 37].

PLS has used several reasons study [21, 22, 38]:

1. PLS is an approach for data analysis that operates under the assumption that the sample size does not need to be substantial;
2. PLS is applicable for examining theories considered to be relatively weak, as it is effective in making predictions;
3. PLS enables algorithms to utilize a series of ordinary least squares (OLS) analyses, enhancing the efficiency of algorithmic calculations;
4. In the PLS method, there is an assumption that all measures of variance can be employed to elucidate each variable;
5. PLS also does not pay attention to data distribution (normal or abnormal).

Use SmartPLS software to analyze data and use structural equation modeling with the following steps [23, 38]:

1. A Structural Model Design (Inner Model).
2. A Measurement Model Designing (Outer Model).
3. Constructing the Path Diagram.
4. Path Chart to Equation System Conversion.
5. Estimation: Path coefficient, Loading, and Weight assessments.
6. Determine the Fit's Quality.
7. Testing of Hypotheses (Resampling Bootstrapping).

There are 17 different types of electronic equipment on the list that were employed for the research, and most of them are found in homes. The information gathered regarding each respondent's ownership of electronic devices at home is displayed in Table 2.

Table 2. Recapitulation of respondent electronic equipment ownership data

Electronic Equipment	Number of Ownership (pcs)							Average (pcs)
	0	1	2	3	4	5	>5
TV	11	261	98	31	2	0	0	1.38
AC	130	179	74	13	7	0	0	0.98
Washing Machine	46	349	8	0	0	0	0	0.91
Refrigerator	14	356	27	6	0	0	0	1.06
Handphone	0	215	159	29	0	0	0	1.54
Tablet	278	117	8	0	0	0	0	0.33
Laptop	93	234	76	0	0	0	0	0.96
Monitor	292	100	10	1	0	0	0	0.31
CPU	294	100	7	2	0	0	0	0.30
Printer	232	161	10	0	0	0	0	0.45
Tape/CD/VCD/DVD Player	192	203	7	1	0	0	0	0.55
Electric Iron	7	361	33	2	0	0	0	1.07
Camera/Video Recorder	219	171	10	3	0	0	0	0.50
Electric Oven	241	150	12	0	0	0	0	0.43
Cooker / Rice Warmer	28	325	50	0	0	0	0	1.05
Blender	62	314	25	2	0	0	0	0.92
Electric Fan	17	225	120	33	8	0	0	1.48

4.png

Figure 4. Average electronic respondents' ownership

As can be seen from Figure 4 above, the average electronic equipment that respondents most frequently own includes TVs, refrigerators, cell phones, electric irons, rice warmers, cookware, and fans. Not all respondents have access to the remaining electronic devices, which include air conditioners, washing machines, tablets, laptops, monitors, CPUs, printers, tape/CD/VCD/DVD players, cameras/video recorders, electric ovens, and blenders.

Following each respondent's electronic data collection at home, the authors also gathered information on the reasons behind consumer equipment replacements for both personal and household devices. The respondents' justifications for changing their household's electronic equipment are listed in Tables 3 and 4 below.

It is known from Tables 3 and 4 that damage is the most common cause given by respondents for replacing household and personal electronic equipment at home. When it comes to updating electronic equipment in the home, there is less diversity than when it comes to personal electronics, as no respondent indicates being bored or lost. Due to its compact size and ease of portability, personal electronic equipment can be lost. Furthermore, respondents have more motivation to update their personal electronic equipment due to the advancement of such complex technologies.

Next, by asking respondents if they had any electronic equipment stored at home, the author started gathering information about their practices regarding the storage of broken or used electronics. In the event that you have, please list the kind of electronic products you have saved, their quantity, duration of storage, and your motivations. 305 out of the 403 respondents that filled out the survey, or nearly 76% of the total, stated that they have used or broken electronic devices. We collected information from each respondent regarding who owned broken or reconditioned electronics at home, which is indicated in Table 5.

According to Figure 5, respondents store mobile devices more often than any other type of electronic equipment. This indicates that mobile devices are used or damaged. It makes perfect sense given that, according to Figure 5, respondents own the greatest number of electronic devices, with cellphones being the most common. Electronic equipment needs to be stored at home because the more persons who possess it, the more likely it is to get destroyed or become obsolete.

Determined the kinds and numbers of damaged or old electrical equipment. The second question concerns how long electronic equipment that has been used or damaged is kept in storage at home. The data regarding the duration of time respondents maintained the broken or used electronic equipment at home is summarized in Table 5.

5.png

Figure 5. Average ownership of used / damaged respondents

According to Figure 6, a monitor, a tape/CD/VCD/DVD player, and a CPU are the electrical devices that have been used or damaged and have been kept at home the longest. The author then asks why electronic equipment that has been broken or used is kept at home. Respondents gave reasons for keeping used/broken household electronic equipment, as shown in Table 6. There is a distinction made here between personal/personal and domestic electronic devices. The data are summarized in Tables 7 and 8 below.

6.png

Figure 6. Average time to store used / damaged electronic equipment

Table 6. Reasons for used / damaged household electronic equipment stored

Table 7. Reasons for used / damaged personal / stored electronic equipment stored

Tables 6 and 7 show that, when it comes to home electronics, the majority of respondents (32.5%) prefer to fix it later, with the second most common reason being that they are unsure of where to dispose of it (24.6%). With regard to personal electronic equipment, the majority of respondents (38.7%) indicated that they would use it as a backup or replacement down the road, while 13.77% said they wanted it repaired later.

With an average of 0.76 units saved per respondent and 58 percent of respondents owning them at home, the study's findings show that mobile phones are the most often kept electronic equipment in terms of quantity. Subsequently, regarding the duration of Time saved, based on 17 electronic devices under investigation, it was discovered that the CPU emerged as the most commonly used and damaged device at home, with an average save time of 2,783 years. Additionally, as much as 32.5% of the respondents kept it for electronic distribution within their homes because they hoped to enhance later, and as much as 24.6% said they were unsure of where to go. Up to 38.7% of people who save their personal electronic equipment do so with the intention of using it as a backup or replacement in the future, and 13.77% do so in order to make improvements.

The PLS approach was used in the data analysis that followed. Version 3.0 of the SmartPLS program was used to evaluate the structural models in PLS. Partial Least Square (PLS) involves the following steps:

4.1 Designing a structural model

Models are built using predefined variables and indicators in conjunction with conceptual models. A structural model and measurement that are derived from the formulation of behavior storage concerns are shown in Figure 7.

7.png

Figure 7. Structural models of the behavior of storing WEEE

Description:

DEM: Demographics

ATT: Attitude

SN: Subjective Norm

PBC: Perceived Behavioral Control

PB: Past Behavior

SF: Situational Factors

CON: Consequences/Outcomes of Storing

INT: Storing Intention

SB: Storing Behavior

4.2 Perform an outer model test

Analysis of the outer model was designed to make sure the measurements were practical (valid and reliable). The examination of the outer model is evident from several indicators:

1. Convergent validity is represented by the loading factors on latent variables along with their indicators, with an anticipated value exceeding 0.5.
2. Discriminant validity is assessed through cross-loading values, serving as a valuable factor to establish whether a construct has sufficient discriminant validity. This involves comparing loading values with the intended construct, ensuring they exceed the loading value associated with another construct.
3. Composite reliability refers to data with a composite reliability exceeding 0.7, indicating a high level of reliability.
4. For Average Variance Extracted (AVE), the anticipated AVE value should be greater than 0.5.
5. Cronbach's Alpha reliability testing was strengthened by Cronbach's.

Based on these preliminary results, a Convergent Validity and Discriminant Validity test will be conducted. This test is done by evaluating the AVE and cross-loading values obtained from the initial data processing. The results of the Convergent Validity and Discriminant Validity tests for the storage behavior of WEEE are displayed in Table 8 below.

Table 8. Test results of convergent validity and early discriminant validity behaviors of storing WEEE

For storing behavior, because several variables or indicators are not valid, then the indicator must be deleted, namely PBC_4 and SF_1, except for the DEM variable (one indicator chosen has the smallest Cross Loading value, DEM_5).

Figure 8 below is a model obtained for the behavior of storing WEEE after the Outer Model test.

8.png

Figure 8. Evaluation results of the storing behavior model

4.3 Perform an inner model test

In guaranteeing the accuracy and robustness of structural models, inner model testing was conducted. Various indicators indicate the Inner Model Test, such as:

1. The determination coefficient (R²) for endogenous variables
2. Predictive Relevance (Q²)
3. Goodness of Fit Index (GoF)

Table 9 below is the R² value of the measurement model for storing WEEE.

Table 9. R² values of the measurement model for the behavior of storing WEEE

From the results of data processing, it is known that for the measurement model of the behavior of storing WEEE, the value of R² from the intention construct variable (INT) with the path scheme is 0.382. it means that the variation of intentions that can be explained by the construct variables DEM, ATT, CON, PBC, PB, SN, and SF is 38.2%, while other variables outside the model explain the other 61.8%. Likewise, for the construct behavior (SB) construct variable, the variation of storing behavior that can be explained by the construct and INT construct variables is 58.4%. According to Ghozali [37], the value of R² is 0.67 (strong), 0.33 (moderate) and 0.19 (weak). After that, Inner Model testing can be done by looking at the Q² value (predictive relevance).

$Q^2=1-\left(1-R_1^2\right)\left(1-R_2^2\right) \ldots .\left(1-R_p^2\right)$           (1)

$Q^2=1-(1-0.382)(1-0.584)$       (2)

$Q^2=0.743$     (3)

From the results of data processing, the Q² value for the behavior of storing WEEE is 0.743. The last is by looking for the value of Goodness of Fit (GoF).

$G o F=\sqrt{\overline{A V E} \times \overline{R^2}}$       (4)

$\mathrm{GoF}=\sqrt{\begin{array}{l}(0.547+0.629+ 0.522+0.839+0.647+0.654+0.735+0.878+0.884) 9\end{array} \times \frac{(0.382+0.584)}{2}}$         (5)

GoF $=\sqrt{0.704 \times 0.483}$      (6)

$G o F=0.583$      (7)

The GoF value for the behavior of storing WEEE is 0.583. GoF values are said to be small if = 0.1, moderate if = 0.25, and large ≥ 0.38 [37]. From testing R², Q², and GoF, it can be seen that the model form for the behavior of storing WEEE is robust. So that hypothesis testing can be done.

4.4 Testing the hypothesis

Hypothesis testing can be seen from the value of t-statistics and probability. The significance of the Outer Model parameter can be evaluated through the bootstrapping procedure. In PLS analysis with the path scheme, the number of replications used is 500 resamplings, and the output results, such as Table 10, obtained for the behavior of storing WEEE:

Table 10. Results of the WEEE storing model bootstrapping procedure

Furthermore, the probability value obtained is compared with the critical value to reject/accept the hypothesis. To reject/accept the hypothesis using probability, H1 is accepted if the value of p <0.05. The following Table 11 shows the results of evaluating probability values for the behavior model of storing WEEE.

Table 11. Evaluation of the probability value of the WEEE storing behavior model

The results of this study support the Theory of Planned Behavior (TPB) by showing that consumers' decision to keep used or damaged electronic equipment (WEEE) is driven by a belief in its potential future value, such as repair or use as a backup. This attitude is reflected in 32.5% of respondents keeping household appliances for later repair and 38.7% keeping personal appliances as a backup. In addition, uncertainty regarding the appropriate disposal site (24.6% of respondents) suggests that subjective norms and perceived constraints also play an important role. These findings underscore the need for education and adequate disposal facilities to reduce the tendency of WEEE storage.

By integrating these elements, existing literature, thoroughly discuss the implications of our results, we aim to provide a comprehensive understanding of the study's significance and its potential impact on the broader discourse in WEEE management to know how behaviour from user.