Analysis Of The Impact Of COVID-19 Science Popularization WeChat Official Accounts And Articles And Their Influencing Factors

Objective: To evaluate the impact of COVID-19 science popularization WeChat official accounts and articles,and explore their influencing factors to provide reference for improving the efficiency of health communication on the WeChat.Methods: The retrospective analysis was used to design the study.We searched for science popularization WeChat official accounts using "science popularization" as the keyword,and excluded those that had not published articles for 6 months and those that had been cancelled.Using the Python-based Scrap-Redis crawler framework to crawl a randomly selected 50% of the articles published by the public from December 31,2019 to July 31,2021.The COVID-19 science articles were filtered by keyword search and manual judgment in two steps.Using Qingbo data and Sougou WeChat search as data sources,we obtained information related to the influencing factors incorporated into WeChat official accounts through manual retrieval,and used Epidata 3.1 for double-entry.The article reads,likes,publish time,title and content HTML formatted text are automatically obtained by the crawler program and saved to the local MySQL database,and the information is automatically extracted by regular expressions matching.Given the accessibility of the data,the factors related to the impact of WeChat official accounts included in this study include only: original certification,video function,publish habits,number of monthly posts,number of original articles,median number of posted articles read and likes.The factors related to the impact of articles only include: word count,headline punctuation,multimedia,text style adjustment,paragraph style adjustment and special element.The WeChat Communication Index was used to evaluate the impact of WeChat official accounts,and the number of reads and likes were used to evaluate the impact of articles.The median and interquartile spacing were used to describe the distribution characteristics of the data.MannWhitney U test was used to compare the differences in WeChat Communication Indices of different types of accounts.Spearman’s correlation coefficient was used to measure the magnitude of correlation between numerical variables with a test level of α=0.05.Using the 75 th percentile of the overall distributions of WeChat Communication Indices,reads and likes as the boundary,the included official accounts and articles were classified into two categories of high and low impact,and binary logistic regression models of the impact of official accounts and articles were constructed to evaluate the role of the selected influencing factors on the impact of official accounts and articles,respectively.The independent variables were screened using the likelihood ratio test（forward）with maximum skewed likelihood estimation,а_enter=0.05,а_remove=0.10.Results:（1）A total of 201 WeChat official accounts and 22,173 COVID-19 science articles were included.The articles were published from December 31,2019 to July 31,2021.（2）The impact of both science popularization WeChat official accounts and COVID-19 science articles showed a right skewed distribution,with the vast majority having a low impact.The median and interquartile range of WeChat Communication Index were 239.4 and 215.5,respectively,and the median number of articles reads and likes were 101 and 1,with an interquartile range of 408 and 3,respectively.（3）The 75 th percentile of the overall distributions of WeChat Communication Indices,reads and likes were 353.9,434 and 3,respectively.The results of the binary logistic regression of official accounts’ impact showed that,for the influencing factors included in this study,only the number of original articles was related to the impact of official accounts,and the impact of official accounts with more original articles（≥92）was higher compared to those with fewer original articles（92）（OR=9.002,95% CI: 4.181-19.380）.The results of the binary logistic regression of article impact showed that for the influencing factors included in this study,higher reads were associated with indenting the ends of article paragraphs（OR=9.102,95%CI: 1.231-2.659）,inserting tables（OR=2.149,95%CI: 1.231-2.659）,adjusting paragraph word spacing（OR=1.823,95%CI: 1.547-2.148）,using exclamation points in headings（OR=2.068,95%CI: 1.942-2.202）,italicizing text（OR=1.247,95%CI: 1.040-1.494）,adjusting paragraph line spacing（OR=1.362,95%CI: 1.209-1.534）,changing text background color（OR=1.260,95%CI: 1.088-1.459）,adjusting（OR=1.278,95%CI: 1.169-1.397）,adjusting the distance before the paragraph（OR=1.207,95%CI: 1.095-1.330）,bold text（OR=1.165,95%CI:1.084-1.252）,and inserting a separator（OR=1.124,95%CI: 1.047-1.206）.All of the above factors,except for text boldness,are also influential factors for higher likes.Lower reads were associated with indenting the first line of the article paragraph（OR=0.633,95%CI: 0.594-0.676）,centering the paragraph（OR=0.739,95%CI: 0.631-0.866）,right justifying the paragraph（OR=0.507,95%CI: 0.472-0.545）,using images（OR=0.527,95%CI: 0.443-0.627）,underlining the text（OR=0.265,95%CI: 0.212-0.332）,changing the color of the text（OR=0.204,95%CI: 0.164-0.254）,and using music（OR=0.438,95%CI: 0.265-0.724）.All of the above factors were also influential factors for lower likes except for centering of article paragraphs and use of music,and use of question marks in article titles was associated with lower likes（OR=0.869,95%CI: 0.807-0.935）,but not with reading level.Conclusions: The impact of the science popularization WeChat official accounts and COVID-19 science articles included in this study varied greatly,and the number of high-impact official accounts and articles was extremely small.In terms of the influencing factors involved in this study,high-impact official accounts are mainly associated with publishing more original articles;high-impact articles are related to the use of exclamation marks in headlines,adjusting text（bold,italic,change text background color）,adjusting paragraphs（indenting at both ends,adjusting pre-and post-paragraph spacing,adjusting line spacing and word spacing） and the use of special elements（tables,dividers）.WeChat science article editors can improve the readability of the article by some text typography methods such as adjusting the two ends of the paragraph indentation,to improve the impact of the article appropriately.