Social Media and Electoral Predictions: A meta-analytic review

Marko M. Skoric; Jing Liu; Kokil Jaidka

doi:10.18690/978-961-286-280-0

Social Media and Electoral Predictions: A meta-analytic review

Marko M. Skoric^*, Jing Liu, Kokil Jaidka

^*Corresponding author for this work

Department of Accountancy, Economics and Finance

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

Abstract

Can social media data be used to make reasonably accurate estimates of electoral outcomes? We conducted a meta-analytic review to examine the predictive performance of different features of social media posts and different methods in predicting political elections: (1) content features; and (2) structural features. Across 45 published studies, we find significant variance in the quality of predictions, which on average still lag behind those in traditional survey research. More specifically, our findings that machine learning-based approaches generally outperform lexiconbased analyses, while combining structural and content features yields most accurate predictions.

Original language	English
Title of host publication	Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society
Editors	Andreja Pucihar, Mirjana Kljajič Borštnar, Roger Bons, Juergen Seitz, Helen Cripps , Doroteja Vidmar
Publisher	University of Maribor Press
Pages	765-781
Number of pages	17
Edition	1st
ISBN (Electronic)	9789612862800
DOIs	https://doi.org/10.18690/978-961-286-280-0
Publication status	Published - 12 Jun 2019
Event	32nd Bled eConference – Humanizing Technology for a Sustainable Society - Bled, Slovenia Duration: 16 Jun 2019 → 19 Jun 2019 http://arhiv.fov.um.si/ebled_2019/ https://press.um.si/index.php/ump/catalog/book/418

Publication series

Name	Bled eConference
Volume	BeC 32

Competition

Competition	32nd Bled eConference – Humanizing Technology for a Sustainable Society
Country/Territory	Slovenia
City	Bled
Period	16/06/19 → 19/06/19
Internet address	http://arhiv.fov.um.si/ebled_2019/ https://press.um.si/index.php/ump/catalog/book/418

User-Defined Keywords

Social Media
Election Prediction
Network Feature
Content Feature
Meta-Analytic Review

Access to Document

10.18690/978-961-286-280-0

Cite this

Skoric, M. M., Liu, J., & Jaidka, K. (2019). Social Media and Electoral Predictions: A meta-analytic review. In A. Pucihar, M. K. Borštnar, R. Bons, J. Seitz, H. Cripps , & D. Vidmar (Eds.), Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society (1st ed., pp. 765-781). (Bled eConference ; Vol. BeC 32). University of Maribor Press. https://doi.org/10.18690/978-961-286-280-0

Skoric, Marko M. ; Liu, Jing ; Jaidka, Kokil. / Social Media and Electoral Predictions: A meta-analytic review. Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society. editor / Andreja Pucihar ; Mirjana Kljajič Borštnar ; Roger Bons ; Juergen Seitz ; Helen Cripps ; Doroteja Vidmar. 1st. ed. University of Maribor Press, 2019. pp. 765-781 (Bled eConference ).

@inproceedings{3668278028fb40c0a4a0b69213d5f5d3,

title = "Social Media and Electoral Predictions: A meta-analytic review",

abstract = "Can social media data be used to make reasonably accurate estimates of electoral outcomes? We conducted a meta-analytic review to examine the predictive performance of different features of social media posts and different methods in predicting political elections: (1) content features; and (2) structural features. Across 45 published studies, we find significant variance in the quality of predictions, which on average still lag behind those in traditional survey research. More specifically, our findings that machine learning-based approaches generally outperform lexiconbased analyses, while combining structural and content features yields most accurate predictions.",

keywords = "Social Media, Election Prediction, Network Feature, Content Feature, Meta-Analytic Review",

author = "Skoric, {Marko M.} and Jing Liu and Kokil Jaidka",

year = "2019",

month = jun,

day = "12",

doi = "10.18690/978-961-286-280-0",

language = "English",

series = "Bled eConference ",

publisher = "University of Maribor Press",

pages = "765--781",

editor = "Andreja Pucihar and Bor{\v s}tnar, {Mirjana Kljaji{\v c}} and Roger Bons and Juergen Seitz and {Cripps }, Helen and Doroteja Vidmar",

booktitle = "Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society",

address = "Slovenia",

edition = "1st",

note = "32nd Bled eConference – Humanizing Technology for a Sustainable Society ; Conference date: 16-06-2019 Through 19-06-2019",

url = "http://arhiv.fov.um.si/ebled_2019/ , https://press.um.si/index.php/ump/catalog/book/418",

}

Skoric, MM, Liu, J & Jaidka, K 2019, Social Media and Electoral Predictions: A meta-analytic review. in A Pucihar, MK Borštnar, R Bons, J Seitz, H Cripps & D Vidmar (eds), Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society. 1st edn, Bled eConference , vol. BeC 32, University of Maribor Press, pp. 765-781, 32nd Bled eConference – Humanizing Technology for a Sustainable Society, Bled, Slovenia, 16/06/19. https://doi.org/10.18690/978-961-286-280-0

Social Media and Electoral Predictions: A meta-analytic review. / Skoric, Marko M.; Liu, Jing; Jaidka, Kokil.
Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society. ed. / Andreja Pucihar; Mirjana Kljajič Borštnar; Roger Bons; Juergen Seitz; Helen Cripps ; Doroteja Vidmar. 1st. ed. University of Maribor Press, 2019. p. 765-781 (Bled eConference ; Vol. BeC 32).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Social Media and Electoral Predictions: A meta-analytic review

AU - Skoric, Marko M.

AU - Liu, Jing

AU - Jaidka, Kokil

PY - 2019/6/12

Y1 - 2019/6/12

N2 - Can social media data be used to make reasonably accurate estimates of electoral outcomes? We conducted a meta-analytic review to examine the predictive performance of different features of social media posts and different methods in predicting political elections: (1) content features; and (2) structural features. Across 45 published studies, we find significant variance in the quality of predictions, which on average still lag behind those in traditional survey research. More specifically, our findings that machine learning-based approaches generally outperform lexiconbased analyses, while combining structural and content features yields most accurate predictions.

AB - Can social media data be used to make reasonably accurate estimates of electoral outcomes? We conducted a meta-analytic review to examine the predictive performance of different features of social media posts and different methods in predicting political elections: (1) content features; and (2) structural features. Across 45 published studies, we find significant variance in the quality of predictions, which on average still lag behind those in traditional survey research. More specifically, our findings that machine learning-based approaches generally outperform lexiconbased analyses, while combining structural and content features yields most accurate predictions.

KW - Social Media

KW - Election Prediction

KW - Network Feature

KW - Content Feature

KW - Meta-Analytic Review

U2 - 10.18690/978-961-286-280-0

DO - 10.18690/978-961-286-280-0

M3 - Conference proceeding

T3 - Bled eConference

SP - 765

EP - 781

BT - Proceeding of 32nd Bled eConference - Humanizing Technology for a Sustainable Society

A2 - Pucihar, Andreja

A2 - Borštnar, Mirjana Kljajič

A2 - Bons, Roger

A2 - Seitz, Juergen

A2 - Cripps , Helen

A2 - Vidmar, Doroteja

PB - University of Maribor Press

T2 - 32nd Bled eConference – Humanizing Technology for a Sustainable Society

Y2 - 16 June 2019 through 19 June 2019

ER -