iPURSE 2019

Permanent URI for this collection

https://ir.lib.pdn.ac.lk/handle/20.500.14444/37

Browse

Browsing iPURSE 2019 by Subject "Agriculture"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Internet based apps and social media in supply of agricultural information: a case study
    (University of Peradeniya, 2019-09-12) Liyanage, P. B.; Prasada, D. V. P.
    Agriculture needs current and relevant information such as availability of inputs, services, market access, market opportunities, and new technologies. Agricultural information exchanged over social media is an emerging communication pathway in Sri Lanka. It is important to find out the demand for agricultural information shared on the social media sites before evaluating the impact of online information on agricultural productivity. To assess the demand for online agricultural information, two samples of ‘Govipola’ Facebook page users (n=79) and ‘Govipola’ posts (n=149) were selected using simple random sampling. Two statistical models were estimated using the data collected. In Model 1 (i.e. identification of the preference and the satisfaction on the separate agriculture related post types on Facebook), data were collected by using structured questionnaire through an online survey. According to the findings, posts with images with short content are most effective. Previous knowledge on agriculture has the largest effect with a coefficient of 1.694 (p=0.05) for this preference. In Model 2 (i.e. the assessment of demand for agricultural information on social media sites) meta-analytics of the posts shared on Govipola Facebook page were used. The metrics such as lifetime ‘likes’, comments and the ‘shares’ proxy the demand for information on the Govipola Facebook page. ‘Boosting’ shown the highest impact for shares with a coefficient of 5.33 (p=0.05) and moderate impacts on ‘comments’ and ‘likes’ with coefficients of 2.299 (0.05) and 2.127 (0.05), respectively. Agronomic content shows the least impact with a coefficient of - 0.78 (p=0.05). For estimating the relationships, Logistic, Ordered logistic and Poisson regression models were used. Age of the post and request for response in the post displayed positive impacts on the penetration of information.