BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:75a5856578d7bcadda1c591b48fafa63
DTSTAMP:20240429T111628Z
SUMMARY:Dr. Wendy Doe\, University of Maryland
DESCRIPTION: \n\nTOPIC: Social Media Brand Tracking: Measuring Brand Favora
 bility using\nLarge Scale Social Media Data\n\nWe contribute to the social
  media listening research and propose a\nbrand favorability measure\, base
 d on large-scale social media data\,\nthat can be used for benchmarking ag
 ainst other brands. We account for\ndirectional biases exhibited by social
  media posters (i.e.\, some\nposters are generally more positive while oth
 ers are generally more\nnegative)\, consider how this bias affects social 
 media metrics of\nsentiment\, and develop a method designed for large scal
 e social media\ndata that provides an adjusted brand favorability measure 
 that is\ncorrelated with traditional survey-based measures used by brands.
  We\ncollect and examine Facebook data for more than 3000 brands and 170\n
 million unique users. We find that for many brands\, average sentiment\nme
 trics deviate from the brand’s underlying favorability. We explore\nthes
 e deviations and examine how observable factors related to the\nbrand comm
 unity\, brand traits\, and brand activity can affect the\nreliability of a
 verage sentiment as a brand favorability measure\,\nallowing brand manager
 s to identify situations where their social\nmedia sentiment metrics are l
 ikely to deviate from and misstate\nunderlying consumer opinions. We find 
 that smaller brand communities\nwith limited opinion variance are associat
 ed with average sentiment\nmetrics that overstate brand favorability.\n 
DTSTART:20171013T150000Z
DTEND:20171013T162000Z
LOCATION:Gerri C. LeBow Hall\, 3220 Market Street\, TBD\, Philadelphia\, PA
  19104
END:VEVENT
END:VCALENDAR