Degree

Doctor of Philosophy (PhD)

Department

Geography & Anthropology

Document Type

Dissertation

Abstract

The use of Twitter as a channel for weather information inspires a deeper analysis of key information nodes during episodes of high impact weather, especially local meteorologists. To optimize communication on the channel, it is important to understand what kinds of messages produce exposure and attention among users—which translates to knowledge that could improve the reach of a warning. Literature identifies two key models that well describe the cognitive processing of tweets and warnings. The Protective Action Decision Model (PADM) describes risk perception and the factors that enable or disable one from acting on a warning. Particularly through environmental and social cues, the first steps of the PADM could be aided or impeded by a tweet. The Extended Parallel Process Model (EEPM) describes the components of an effective warning message. Even in a tweet, ignoring one or both of the two critical components of a warning—threat and efficacy—could inhibit a user from taking the correct protective action, if any at all. Through two case studies of tweets during high impact weather events in southeast Louisiana, messages containing photos and videos are most likely to appear in Twitter timelines and therefore generate the greatest exposure. Similarly, followers of a local meteorologist Twitter account will be most likely to retweet and therefore pay attention to messages containing photos and videos. The case studies also revealed that, particularly with warnings, tweets containing equal levels of threat and efficacy, as well as some personalizing factor such as a map or geographic indicator generate more retweets and therefore attention. In a subsequent survey, case study results were not duplicated via self-reported interests from respondents. An example photo was less popular and an example warning with minimal actionable information was most popular. The survey also revealed that Louisianans prefer websites and Facebook to receive weather information, while mobile phone apps and Twitter scored lower preferences.

Date

11-2-2017

Committee Chair

Keim, Barry

DOI

10.31390/gradschool_dissertations.4132

Share

COinS