3. RESEARCH DESIGN AND METHODOLOGY
3.5. Methodology
There is a debate regarding qualitative or quantitative methods and the impact of various methodologies on the reliability and validity of the research results. Those in favour of quantitative methods, such as Mintzberg (1973) and Hodgson, et al. (1965) base their arguments on the objectivity and internal validity of results obtained. They consider bias on the part of the researcher as an inescapable part of the qualitative methodology. The validity of the results may, therefore, be questioned, and it would be difficult to compare the results of studies (Gill & Johnson, 1997). However, supporters of the qualitative method, such as Neustadt (1960) and Burgess (1993), base their criticism of quantitative methods on whether quantification is possible under all circumstances and the possibility of uncontrolled bias.
Scholars like Cormack (1991) as a solution suggests multiple methods are vital to reduce the qualitative and quantitative method’s drawbacks through triangulation, multiple research methods that can be used to gain a total picture of some phenomena. Saunders et al (1997) identify two major advantages of multiple research methods. First, different methods may be used for different purposes. Second, different data collection methods may be used to provide convergent evidence (a process referred to as triangulation). Based on this, this study, therefore, used a triangulation methodology, which comprised of both qualitative and quantitative methods for all specific objectives.
3.5.1 Objectives 1: Assessment of Household’s Behaviours and Perception of
Resource Consumption and Green Environmental Tradeoffs
This study was conducted in Kombolecha city, which is among industrial city in Amhara Regional State, Ethiopia. Hence, it mainly targeted ‘Kebele’ four administrations (‘Kebele’ here refers the lowest administration unit in Ethiopia) and the industrial Zone where people and factories are densely populated. The population frame was 3252 households, who composed of households that consumed water resources and recycled wastes differently. Nonetheless, the target population was sample households, who provided information in the period of primary data collection in the study area.
Based on household’s complex socio-demographic characters and consumption patterns, the target population was divided into mutually exclusive groups and classified into four major categories: namely, factory employees (1537), consumers excluding factory employee (1265), suppliers (450) and service providers (125) such as hotel, garage, café services etc. To get accurate information from each category, stratified random sampling techniques were applied to sample households. Household’s category served as a stratum. Out of each stratum, individual households were selected randomly to give each household an equal chance of being selected.
Out of four categories of households, the total sample households from all stratawas ni = n1+n2+n3+n4.
Accordingly, sample households were factory employee (n1=154), consumers (n2=126), suppliers (n3= 55),
and service provider (n4 = 4) from each stratum. Hence, 338 sample households from Kombolcha industrial
zone were selected to gather data using semi structured questionnaires and interview (for further detals use Annex 2).
The questionnaire and interview schedules consist of both open and close ended questions.
In the questionnaire, the household’s consumption behaviour (qualitative) was measured based on respondent’s consciousness to adopt green thinking (green mind), product consumption, technology and green job use; ability and willingness to buy green inputs; product (whether green or grey); consumption strategy; water quantity; awareness about green environment and management. As a guideline and standards, the study used Amhara Regional State; Kombolecha Municipality Clean and Beautification Office environment management manuals; and Water Supply and Sewerage enterprise Office manuals (2014-2017). Whereas, household’s qualitative perceptions were measured in terms of respondent’s emotionality and sensitivity to adopt a green mind, consumption, marketing, technology and job searches that balance the water consumption and recycles regards to the social, environmental, economic wellbeing and understandings. These qualitative measurement and characters were explained in nominal five-point Likert scale categories (from strongly agree up to strongly disagree). Besides, neutrality categorical scales such as don’t perceive, don’t behave, not at all, etc. were included in questionnaires.
Before conducting a questionnaire survey, the validity of constructs was checked. Cronbach’s alpha was computed to measure reliability and internal consistency of the measurement of qualitative characters and
scales. Face validity of qualitative measurement scales were checked by researcher, experts and pertinent with literature reviews. In order to test the dimensionality of qualitative scales measurement and constructs, an explorative factor analysis was applied following the procedures recommended by Kuznets (1955) and Churchil (1979). Moreover, the household’s awareness, behaviour, perception towards practicing a green mind, product consumption, marketing, and technology use inequality was computed by following the environment Kuznet’s curve model (Kuznets, 1955).
After group key informant interviews, data collection was undertaking using questionnaire interviews. A researcher delivered data collection trainings for five data collectors. They were distributed the questionnaires and respondents were filled the questions. However, data collectors were read the questions for respondents and filled response for which they were not read and write on the questionnaires. Photograph was taken by experts using digital camera. This instrument was used to gather precise information and substantiate household perception and behaviour regarding to seek the green environment. The questionnaires were administered in several ways. But, for this study, a researcher read questions for respondents and writes their answer on questionnaire. It was interviewed, managed and collected by the researcher so as to probe respondent’s perception and behaviours. The collected data were analysed using the qualitative and quantitative techniques. The qualitative methods applied content methods, description and proportional techniques, and case analysis. The quantitative techniques were used both the inferential statistic and econometric regression and computation.
This study used econometric model to identify the correlation and to determine association between variables (perception, consumption behavior, resource consumption, level of green environment) and test the variable significance. Whereas, the descriptive inferential statistics were used to inference statistics and information from sample to a large population and help to evaluate the impact of variables. Gujirati and Maddala (1983&2004) and Greene (2004&2011) logistic regression model was applied to investigate the association between variables. Kuznet’s (1955) Environmental Kuznet Curve Model was also used to assess inequality between the household perception and behaviour along with their monthly income (economic instrument) and poverty status (social instrument) intertwined with their water consumption and recycling efficiency (environmental instrument).
This study, furthermore, used secondary data such as indexes, reports, manuals, national and international institution standard measurement scales and indexes that helped to substantiate and support the result of primary cross-sectional data. The sources of these secondary data were libraries, websites, publications, journals, and etc.
3.5.2 Objective 2: Determined Socio-eco efficiency Indicators on Resource Consumption and Recycling processes in Kombolecha
Like the previousobjective1, objective 2 was conducted in Kombolecha city. Relevant information concerning the household’s social (poverty status, behaviours and culture); economic (monthly income) and environmental aspects (water quantity and waste recycle) were keenly collected to integrate the three key indicators and determine the significant socio-eco efficiency indicators effect on water resource consumption and recycling processes. In pursuit of this, 338 sample households, who consumed water resources, were participated during data collection. Moreover, data were gathered purposively from 14 factories, which are consumed water resources (Kombolecha municipality, 2013). In this regard, factory’s production managers were purposively sampled respondents. Based on consumer’s (both household’s and factories) water consumption and types of production, the researcher classified them into six sectoral categories: cloth and garment produces, beer and soft drink, metal and steel, leather and related product, food and related processing, manufacturing and others sectors. Factory managers were presented as a sample and hence counted as 14 respondents. Based on Kombolecha municipality and investment profile document (2013), this study, thus, took all factories, such as two factories from cloth and garment producer company, one beer factor, two metal and steel producer, one leather and related, three food and related processing factories, two manufacturing and other three factories were sampled and taken to collect the primary data.
The data collection phase was undertaken from factories and other professionals using structured questionnaire, which consists of both open and close ended questions. Social, economic and environmental indicators on water resource consumption process were used as a guide line to prepare structured questionnaires. The various indicators of socio- eco efficiency framework in questionnaires were used as data survey instrument. Consistent with the proposed questionnaires, the descriptive survey methods were constructed and undertaken so that correlation levels or strength of relationships between
variables such as level of green environment and socio- eco efficiency indicators were assessed, characterized and quantified.
In doing so, this study generated a list of indicators in a questionnaire and respondents determined how each indicator criterion weighted on water and waste consumption process. Based on indicator criterion, the selection grid should have a scoring system for ranking the indicators. The weighted voting can be a simple Yes or No to a numerical rating system. Many numerical systems are possible such as (1-5) and (1- 10). The larger number or "YES" was represented a desirable rating. In some cases, large number may mean "less", for example cost of water or waste removals. In order to set scoring, the researcher asked every sample household to score each indicator against the criteria. Respondents completed the questionnaires to evaluate how well the economic, social and environmental indicators were pertinent to resolve the consumption and recycle inefficiency problems. The average score from each respondent were taken. Finally, total and average score were computed and summed based on the respondent’s scoring result.
Accordingly, this study used the highest ranked economic, social, and environmental indicators to gather information from the household and factory’s water consumption and waste recycles. For instance, monthly income, poverty status and culture and quantity of water were high ranked indicators among economic, social and environmental indicators. These indicators defined as the effect of household and factory’s water and waste consumption activities on the environment as well as the implication of those actions on other indicators integrity that described conditions during consumption process. Three major social, economic and environmental indicators and their integration were generated for respondent’s indicator voting and scoring purposes in the questionnaires.
Based on these classifications of indicators, a pilot study was undertaken by distributing 20% of questionnaires to samples (people and factories) so as to check validity of content and constructs. In order to check correlation between variables and quantitative measurement scales, Pearson chi square value was calculated to measure and test internal strength and relationship between variables or indicators and level of green environment. To test the dimensionality of measurements scales and construct variables or indicators (socio-eco efficiency indicators), descriptive factor analysis was done following WBCSD (1996) and ESCAP (2009) indicator principles and criterion.
Meanwhile, the researcher computed households and factory’s intensity water consumption and waste recycle efficiency relative to green environment impacts. In this study context, waste is defined as an end product, which consists of both solid and liquid waste, having negative economic value on environment. In Kombolecha industrial Zone, household and factory’s water consumption and waste recycle intensity or productivity were measured using the formula:
Water consumption intensity: cubic metre of water per households and factory’s product Liquid waste recycle intensity: cubic metre of liquid waste per consumer’s product
In this case, environmental items were measured using physical units, such as cubic metre (m3) of water
and liquid waste consumption, tons (t) of solid waste. Whereas, water consumption and product value adds were measured using in monetary terms. For this study, Ethiopia currency called Birr were used to measure monetary value of resources such as water, wastes, costs, value adds and etc. To measure efficiency of indicators, it was computed the ratio of water consumption and recycle efficiency relative to households and factory’s value adds of product with respect to social, economic and environmental values. These ratios were measured environment burden of water and waste consumption per unit of economic and social values in Birr. For example, M3 of water consumption per value added of respondent’s products
were computed in Ethiopia birr.
On the other side, in this study, indicators were categorised in to eco - efficiency, socio - efficiency and socio- eco efficiency. Eco efficiency was computed economic value of products relative to environmental quality in physical and monetary terms. Whereas, socio- efficiency was measured social value adds of water and waste like health effect with respect to environmental quality in monetary terms. Socio - eco efficiency was calculated physical items of environmental quality (water and waste per units) relative to economic and social value add combination or summations. Similar conceptual formula was used to measure the indicator efficiency.
Eco efficiency of water = water consumption / M3 (environmental quality)
Economic value adds of water on products/Birr
Socio efficiency of water = cubic metre of water consumption (environmental quality) Social valued adds like health/cost in Birr
Socio- eco efficiency of water = Eco- efficiency of water + Socio- efficiency of water After computing and measuring these indicators, the content validity of variables (indicators) will also be checked by WBCSD, (1996), BAZF and ESCAP (2009) indicator criterion and principles; SO standards 14040 and latest criterion; UNEP (2009) and UNIDO (2011) environment and industry strategy manuals, FDRE Environmental Protection Agency Manual (FDRE, 2010); FDRE Industry Development Strategies (2010); and FDRE Product Quality Assurance and Measurement Agency manuals (2010) literatures and experts. Using environmental item in physical or financial terms relative to economic and social value adds, determinant indicators were identified on the water consumption and waste recycling process in the Kombolecha Industrial Zone.
It was, therefore, both qualitative and quantitative descriptive data analysis techniques were used to probe the data and interpret the result. The qualitative techniques were factor grounding theory and descriptive factor analysis. Whereas, the quantitative techniques were applied econometric models, descriptive statistical inferences and central tendencies such as percentage, mean ratio, average and etc. Importantly, econometric models were used to identify and determine association of indicators and their correlation. Hence, binary Logistic Regression Model (BLRM), Instrumental Variable model (IVM) and Two Stage Least Square estimation (TSLM) were used to measure association and correlation between variables. Model goodness of fit and correlation status of variables were measured and checked by Pearson chi square along with the guideline set by Guajarati (1983 & 2004) and Greene (2011). This study model fitness was computed 74 percent, which indicates this model sufficient prediction capacity between explained and explanatory factors.
The validity of statistics and econometric models were checked and accredited by Gujarati (2004) and Greene (2011) and Wooldridge (2012) criterions along with each model proposed purpose and importance to analyze the data for objective two. Secondary type of data such as WBCSD (1996), BASF and ESCAP
(2009) social, economic, and environmental indicator measurement, scales, indexes, ISO standards, reports and statistics were used to support and strengthen the primary data. The data sources were libraries, internet or website, journals and publications, factory profile and annual reports, Ethiopia environment protection agency, Amhara regional state, and Kombolecha municipal office unpublished documents. Data inserting, coding, editing and interpreting procedures were done using the latest SPSS24 and STATA 15 Software programs. An alpha value of 0.05was used as the level for determining the factor significance.
3.5.3 Objective 3: Evaluated the Extent to Which Indicators on Consumption of Water and Waste Recycles Would Impact the Environment
For this study, purposive sampling techniques were used to collect data from consumers (126), service providers (4) and factories employees (14) to evaluate indicators impact on water consumption and recycling processes in Kombolecha Industrial Zone. A simple random sampling technique was employed to collect data from mentioned categories. The study used structured questionnaires and field survey research method to collect information from respondents. In the questionnaire, the researcher categorised indicators into three dimensions. First, it applied indicators classification such as social, economic and environment indicators in socio- eco efficiency framework; second, the extent at which indicators impact on the environment in physical items is called characterization and third, calculation of indicator’s impact on environment relative to value adds in monetary terms, Ethiopia Birr is known as quantification. These indicators categorization was appropriate for propensity score matching estimation that gauge indicators impact on the green environment.
Product life cycle assessment following by WBCSD and ESCAP (2009), social life cycle assessment by BASF (2009) chemical company group and environmental life cycle and quality standards by ISO (2012&13), UNEP (2011) and Ethiopia Environment Protection Agency quality criterion (EPA, 2010) were used as a guide line instrument to design a questionnaire and an interview. Hence, interviews using prepared questionnaires were used for key professionals or experts who concern water and waste management and recycling process in the factory. Besides, field observations were undertaken. And hence photograph image were taken during interview and field observation. This was helped the study to consist precise information about waste recycles, wastes removal systems, operation management and etc. Thus,
primary cross sectional data, which used to evaluate indicators impact on water consumption and recycles, gathered from both households and factories.
Meanwhile, the researcher prepared objective scoring system for social, economic and environmental indicators so as to evaluate their potential impact on water and waste recycling process in a questionnaire. The study generated a list of possible indicators in questionnaire and hence each sample household and factory’s key professionals were allotted number of votes to select what indicator is their priority to evaluate water and waste recycling. Among many criterions, this study generated indicators in the questionnaire and took indicator’s concerns and understandability by respondents; flexibility, measurability, comparability to previous findings, long term reliability, temporal scope and measure scientifically to evaluate potential impacts on environment. In order to determine how each criterion was weighted and the selection grid would have a scoring system for ranking the indicators, the weighted voting was a simple ‘Yes’ or ‘No’ to a numerical rating system.
Many numerical systems are possible, but for this proposal 1-5 up to 10 were taken in a questionnaire. The larger number (or "YES") must always represent a desirable rating. In some cases, this may indicate "less", for example water and waste recycling rate. Each respondent was, therefore, total their score for each indicator. The score from each indicator was then averaged. The researcher summed total and average the score. Based on this ranked Indicators, water and waste recycle data were gathered from people and factories. Meanwhile, the researcher computed households and factory’s consumption intensity of water and waste recycling to keep environment.
In this context, waste is defined as an end product, which consists of both solid and liquid, having negative economic value on green environment. In Kombolecha industrial Zone, the household and factory’s water consumption and waste recycling intensity measured and evaluated as:
Water recycling intensity: cubic metre (m3)/household and factory’s product
Waste recycling intensity: cubic metre (m3)of liquid waste/household and factory’s product
Like objective 2 in this study, environmental items were measured using physical units such as meter cubic (m3) of water, liquid waste recycles. Whereas, the water consumption and waste recycle value add per
quantity of product were measured in monetary terms (Birr). Based on this, water and waste recycle (environmental items) relative to value adds of recycling on social and economic values were computed. Value add in this context is calculated as product sale minus cost of input like water. These ratios were evaluated quantity of water and waste recycles environmental burden relative to economic and social value in birr.
As a result, eco efficiency, which considers only economic and environmental aspects, of water recycle was calculated using the formula: Environmental item divide by economic value of water intensity or M3/ Birr.
Whereas, waste recycle was measured as: M3 of waste divide by waste intensity or M3/ Birr. Environmental
item (water and waste per m3) divided by social value in monetary terms (birr) measured socio-efficiency,
which considers social and environmental aspects. Hence, conceptual socio-eco efficiency of water and