**2.6 Inclusion and exclusion criteria**

For our inclusion criteria the following protocols were observed:


This means that the respondents who did not meet the above criteria were not allowed to participate in the study. Again, those who were unwilling to participate in the study were allowed to withdraw at any time.

#### **2.7 Data collection instrument**

Both structured and semi-structured instruments were used to collect data from the respondents in February 2021. The structured questionnaire was self-administered to the respondents. For the semi-structured interview, the researchers conducted the interview, which lasted for 20–30 minutes at each interview session. The questions were aligned with the objectives and the theories that were adopted for the study. Some of the questions that were asked included the types of precarious work activities engaged by the migrants; how the precarious work that was undertaken affected their health outcomes; and the attempts made by the government to address precarious work conditions among the migrants.

#### **2.8 Pretesting**

Pretesting of quantitative data was done at Madina with 40 questionnaires. The use of a 40-sample size for the pretest was due to the total migrant population available in the community at the time of pretesting of the data. This was because the upsurge of COVID-19 made most of the migrants go back to their places of origin. The aim of the pretesting was to allow for the restructuring of the questionnaire and the time needed to conduct each interview. Madina is a suburb of Accra in the La-Nkwantanang Madina municipality. The municipality has attracted and continues to attract, migrants from all over the country due to the major commercial activities that take place there. As a result, Madina has also become a home for a growing number of economic migrants [59].

#### **2.9 Data analysis**

The quantitative data gleaned from the study was coded and then analyzed with the use of SPSS software version 21. Chi-square tests were conducted to test whether there

was a statistically significant relationship between the expected frequencies and observed frequencies in one or more categories of the study at the 0.05 level. Logistic regression was also performed on whether or not an association existed between work undertaken by the migrant and whether it resulted in a poor health outcome. Part of this report refers to findings from the qualitative results. For the qualitative analysis, data were transcribed and translated for uploading to NVIVO software. The data were done by identifying themes and sub-themes, coding those themes, and then interpreting the structure and content of the themes. Through this method, a codebook was developed, discussed, and accepted. Nodes were then created within NVivo using the codebook. Data triangulation were used to ensure that the findings were validated across different participants. The data were analyzed to understand the feelings and experiences of the migrants.
