**5. Materials and methods**

Survey questionnaires, interviews, focus group discussion, desktop survey and observation were used in this study. We collected data during two fieldwork periods in five Ethiopian cites: Addis Ababa, Mekelle, Bahir Dar, Hawas and Adama, between June 2014 and March 2017.

The researchers used a pre-tested structured questionnaire to gather information from 40 companies. Half of the 40 registered private companies were taken from Addis Ababa CAA, and the other three CADs were from the municipalities of the remaining four cities. The questions in the questionnaire were measured on a fivepoint Likert scale with categories ranging from 'Strongly disagree', 'Disagree', 'Cannot tell', 'Agree' and 'Strongly agree', 'Very poor = 1,"Poor = 2', 'Fair = 3', 'Good = 4 'Very good = 5'. Also, Binary Scale (Yes/No) questions were included [37]. Analyses were conducted on secondary sources comprising available archive records and document from private companies' monthly reports, contract agreement documents and CAA and CADs official waste-related documents.

Our interviewees were government experts from Ethiopian Environmental Protection Agency Ministry of Environment and Forest (EPA), Addis Ababa CAA and EPESA. We also visited company managers who gave us detailed information regarding their organisation's SWC practice. It is believed that managers initiate, devise and oversee the operations of their company. The interview guide for government officials was open-ended questions and Yes/No questions.

In this study, Focused Group Discussions (FGDs) were conducted the five cities. The responses remained open-ended, and this enabled respondents to produce as much information as they wish to provide. This approach allows respondents to describe and analyse their experiences or feelings in their own words without being constrained by any form of framework.

Two FGD groups were formed. The first group had a total of six individuals with a composition of participants of whom two urban health extension workers, two health development representatives army – one male and one female. Community representatives were one male and one female. This group represented service users. The other FGD group consisted of six participants – two municipal officials, two private companies and two MSE members. This group represented service providers. Generally, the two groups were formed so as to make them free to respond genuinely. Members of the first group were taken through the communication with kebelle officials since they work with the kebelle representing residents. Both the two FGDs lasted

80 minutes each in average. Moreover, for the purpose of triangulation 25 households, five in each city were interviewed. These households were selected after proving that they had lived in each city more than 5 years. The justification was based on the belief the more they live in each city, the more they know about the SWC practice.

The observation was made by visiting throughout the city both in planned and unplanned areas and by taking a view of the whole process of SWC and general cleanliness. This gave us an opportunity to observe the practices within households and private providers in the institutions, transfer stations and disposal sites.

A combination of both qualitative and quantitative methods provided the most accurate data for this study. Data were categorised and cross-tabulated. To get accurate description of SWC practices, official documents and interview results were analysed qualitatively using narration. The methodological part of this paper did not use mathematical models for analysing the data which could have strengthened the result of the study, this may be stare as a curb for this study, but this can be improved by the next research studies on the topic.
