Study Design and Sampling: A cross-sectional study, with a non-probability convenience sampling technique, was conducted during April 2020. A total of 402 responses were collected online from the metropolitan city of Karachi, Pakistan. The sample size of 384 was calculated using openepi.com with a double-sided confidence interval of 95% and a 5% error margin. Citizens of age 18-60 partook in the study, and respondents tested positive for Covid-19 were excluded from the study as it could lead to bias. The initial section of the questionnaire consisted of a portion of consent, which informed the participants about the study’s objective, their voluntary participation, and the maintenance of privacy of their data.
Data Collection and Measures: An online survey was created on Google Forms in the English language, and participants were invited to complete and submit the form. The survey questionnaire was designed carefully after an extensive literature review and was based on the course material of COVID-19 provided by WHO.[18] After designing a rough draft of the questionnaire, it was modified in two steps. Firstly, the study tool was sent to researchers and professionals from the medical background to analyze its simplicity and reliability. Secondly, a pilot study was conducted by selecting a small sample of the general population (n = 20) to test its practicality and affectivity. At the end of the survey, the partakers gave their feedback and amendments, which were considered, and the questionnaire was modified accordingly while ensuring its consistency with the published literature. After a thorough discussion, the questionnaire was finalized and approved by all authors and subsequently distributed. The data of the pilot study was not used for the final analysis. The questionnaire comprised of three sections. The first section entailed questions about socio-demographics, including age, gender, education, and occupation, followed by a question inquiring about their primary source of information regarding Covid-19, and lastly, whether or not they fact-checked the obtained information from an authentic source. A question was included asking whether the respondent have had been tested positive for Covid-19. The second and third sections dealt with information concerning practices and knowledge regarding COVID-19, respectively. A total of 12 statements on practices and 14 statements on knowledge were included, with each section consisting of an equal number of facts and myths.[19,20] For each of these statements, two response options were available upon which a score was calculated. In the practice section, these options included, ‘I have heard/ read this statement’ and ‘I have implemented this measure to prevent/treat COVID-19’, the latter being replaced by ‘I think this statement is accurate’ in the knowledge section. A score of +1 and -1 was awarded if a participant had selected both options for a true and false statement, respectively. Similarly, a score of +0.5 (for a fact) or -0.5 (for false statement) was awarded if the respondent had read/heard about it but did not believe it be true/did not implement it. No points were awarded if the first option, i.e., ‘I have heard/ read this statement,’ was not selected. Hence, a cumulative score ranging from negative six to positive six and negative seven to positive seven could be obtained for practice and knowledge sections. Subsequently, the respondents were categorized into those who used social media as their primary source of information and those who did not (social media v non-social media respondents) for ease of interpretation, analysis, and displaying of data. Lastly, at the end of sections 2 and 3, the respondents were asked how often they shared the information regarding practices and knowledge associated with COVID-19 with their friends and family.
Statistical Analysis: A spreadsheet was produced on Microsoft Excel 2019 for data input, which was later imported to Minitab v19.2 for all statistical analyses. For all quantitative data, mean and standard deviations were derived, whereas qualitative data were expressed in percentage and frequency. Inferential statistics were applied depending upon the nature of data and variables. Two sample t-tests were applied to each statement from the practices and knowledge section separately to determine the difference between respondents’ mean performances using social media as their primary source of information and those who do not. Two sample t-test was also applied to the same classification of respondents (social media and non-social media) to assess the difference by socio-demographic characteristics. A Pearson-rank correlation test was performed to find any correlation between all the respondents’ practices and knowledge. A p-value of less than 0.05 (p<0.05, CI 95%) was considered significant in all tests.