Ballabgarh Contact Mixing Dataset
Description
This data set contains contact diary data with 24 hour recall period from respondents from villages of Ballabgarh collected in three waves from October 2015 to October 2016. Respondents completed a structured questionnaire regarding their contacts in the past 24 hours during a face-to-face interview. A caregiver responded on behalf of children under 6 years of age, while children aged 6-10 responded in the presence of a caregiver. An interviewer of the same gender interviewed all respondents aged 11-18 years. A contact was defined as a face-to-face conversation within a distance of three feet, which may or may not have involved a physical touch. Respondents were asked about the age and sex of their contacts, the place of contact (home, school, work, transport, or other), and whether the contacts were conversational or physical. In addition, the respondents were provided the option for reporting simultaneous encounters with more than one individual encounters with multiple individuals as "group" contacts, including the group size, duration of encounter, and the age range of individuals in the group. Details such as the age and gender of each individual in the group were not recorded. Respondents were interviewed three times in a period of 13 months. The data were gathered over three phases: October 2015 - February 2016 (winter); March - June 2016 (summer); and July - October 2016 (monsoon) and henceforth we refer to the three phases as seasons. Data dictionary, all contacts data and aggregated data are provided as .csv files
Files
Steps to reproduce
Comparison of contacts within and across the seasons: We performed two types of analyses of the data, the first being an analysis within each season, where we identified the differences in contact patterns of age groups for each season individually. For this study, statistical methods for independent samples were used for comparison . Number of contacts: For each respondent, we report the number of individual contacts over the previous day, the number of people contacted in a group setting and the total number of contacts by adding the individual and group contacts. Duration of contacts: The reported durations of contacts were recorded on a categorical scale: “<5 minutes”, “5-14 minutes”, “15-59 minutes”, “1-4 hours” and “>4 hours”. These were converted to numerical values by computing the mean duration for each interval. For instance, for the 15-59 minutes interval, a contact duration of 37.5 minutes was used. For the >4h category, we set the upper limit as 8 hours, the usual maximum for a working day. Age and gender-stratified contacts: We calculated contact numbers and durations stratified by the respondent age-groups (0-9, 10-19, 20-29, 30-39, 40-49, 50-59, 60-69 and 70+). For analyses comparing respondents within a season, we used the Kruskal-Wallis nonparametric H test to identify differences in the contact rates, followed by Dunn’s test to examine pairwise differences between age-categories, with the Bonferroni correction to account for multiple comparisons. Social networks to visualise contact patterns: We created social networks to visualise the contact mixing patterns at Ballabgarh. Since our survey was conducted on a sample of the Ballabgarh population, and all reported contacts were not respondents in the survey, we represented contacts using a directed graph. Contact setting: To understand if the contact patterns differed inside and outside home, boxplots for individual, group, and total number and duration of contacts were plotted for each season. Wilcoxon signed rank test was used to check if the difference between inside and outside home contacts was significant. The percentage of individual contacts involving a physical touch were calculated at each location (home, school, work, transport, other) for every respondent. Group contact purposes: Every respondent specified the purpose of each of their group contacts in a free-write format. These strings were inconsistent and often had spelling mistakes. To understand the difference in group contacts across the three seasons, we created 13 contact reason categories such as school, politics, weddings, work, and worship. A set of keywords was defined for each category (Keyword matching was performed to map the contact reason strings to these categories. Age-assortative mixing matrices: We calculated age-assortative mixing matrices for both the number and duration of contacts. These matrices represent the average number or duration of contacts between each pair of age categories.