東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

What's the Secret Recipe for Persuas...

Masick, Kara,

Linked to FindBook

Google Book

Amazon

博客來

What's the Secret Recipe for Persuasive Message Content Success? /

Record Type:	Electronic resources : Monograph/item
Title/Author:	What's the Secret Recipe for Persuasive Message Content Success? // Kara Masick.
Author:	Masick, Kara,
Description:	1 electronic resource (124 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-02, Section: B.
Contained By:	Dissertations Abstracts International86-02B.
Subject:	Psychology. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31240777
ISBN:	9798383628157

What's the Secret Recipe for Persuasive Message Content Success? /
Masick, Kara,

What's the Secret Recipe for Persuasive Message Content Success? /Kara Masick. - 1 electronic resource (124 pages)

Source: Dissertations Abstracts International, Volume: 86-02, Section: B.

The science of persuasion remains an area of inquiry largely driven by siloed research programs that rarely get tested together. Many theories do not tell us what works best. We all believe more persuasive messages ought to result in more behavioral engagement; such a relationship, however, might be tempered by a host of other factors. To address the question "what works?," I conducted a three-staged study. First, I accumulated 556 message content qualities that experts have linked to persuasive success. Second, I collected message and persuasion data from social media to serve as my data source. Through the course of six weeks, I captured 1410 unique text-only messages that attempted to persuade the audience to take action that could be tracked (e.g., like, reply, retweet, and/or quote). Both the message content (predictors) and the behavioral data (outcomes) served as the basis for the next stage. Finally, I used a large language model (LLM) to "score" the messages according to the qualities identified in the first stage. The LLM was prompted to score every message on an 11-point (0-10) scale for each quality multiple times. In total, 1196 messages were scored ten times on 556 qualities. Once scored, I reduced the message quality ratings from 556 to 15 latent variables (via exploratory factor analysis). These latent variable scores then were used as predictors of the outcomes. The result from all three stages was an observational, longitudinal study that helped identify three consistent factors that predicted persuasive behaviors. My aim here was to create a recipe for best persuasion practices. Behavioral outcomes from persuasive social media messaging occur routinely (48.6% of my sample achieved a desired response) and are dependable enough to be predictable by message content qualities (pseudo-R2 = 0.25). With relatively high accuracy (74.4%), a logistic regression model predicted whether any message would persuade at least one user to engage. Despite those results, a simple, generalizable recipe does not exist, but there are grains of truth in almost every persuasion model available today - yet, most are wrong. Theoretical advances aside, the main contribution from this work is the application of LLMs for scoring observational data quickly and inexpensively. The implications of these methodological approaches may offer far greater heuristic value to the scientific community beyond the theoretical contributions.

English

ISBN: 9798383628157Subjects--Topical Terms:

519075
Psychology.
Subjects--Index Terms:

Large language models

What's the Secret Recipe for Persuasive Message Content Success? /
LDR:03858nmm a22004213i 4500 001 2435589
005 20251009122838.5
006 m o d
007 cr|nu||||||||
008 261011s2024 miu||||||m |||||||eng d
020 $a 9798383628157
035 $a (MiAaPQD)AAI31240777
035 $a AAI31240777
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Masick, Kara, $e author. $0 (orcid)0009-0000-0104-8688 $3 3825934
245 1 0 $a What's the Secret Recipe for Persuasive Message Content Success? / $c Kara Masick.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (124 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-02, Section: B.
500 $a Advisors: McKnight, Patrick Committee members: Shaw, Tyler H.; Stuewig, Jeffrey B.
502 $b Ph.D. $c George Mason University $d 2024.
520 $a The science of persuasion remains an area of inquiry largely driven by siloed research programs that rarely get tested together. Many theories do not tell us what works best. We all believe more persuasive messages ought to result in more behavioral engagement; such a relationship, however, might be tempered by a host of other factors. To address the question "what works?," I conducted a three-staged study. First, I accumulated 556 message content qualities that experts have linked to persuasive success. Second, I collected message and persuasion data from social media to serve as my data source. Through the course of six weeks, I captured 1410 unique text-only messages that attempted to persuade the audience to take action that could be tracked (e.g., like, reply, retweet, and/or quote). Both the message content (predictors) and the behavioral data (outcomes) served as the basis for the next stage. Finally, I used a large language model (LLM) to "score" the messages according to the qualities identified in the first stage. The LLM was prompted to score every message on an 11-point (0-10) scale for each quality multiple times. In total, 1196 messages were scored ten times on 556 qualities. Once scored, I reduced the message quality ratings from 556 to 15 latent variables (via exploratory factor analysis). These latent variable scores then were used as predictors of the outcomes. The result from all three stages was an observational, longitudinal study that helped identify three consistent factors that predicted persuasive behaviors. My aim here was to create a recipe for best persuasion practices. Behavioral outcomes from persuasive social media messaging occur routinely (48.6% of my sample achieved a desired response) and are dependable enough to be predictable by message content qualities (pseudo-R2 = 0.25). With relatively high accuracy (74.4%), a logistic regression model predicted whether any message would persuade at least one user to engage. Despite those results, a simple, generalizable recipe does not exist, but there are grains of truth in almost every persuasion model available today - yet, most are wrong. Theoretical advances aside, the main contribution from this work is the application of LLMs for scoring observational data quickly and inexpensively. The implications of these methodological approaches may offer far greater heuristic value to the scientific community beyond the theoretical contributions.
546 $a English
590 $a School code: 0883
650 4 $a Psychology. $3 519075
650 4 $a Communication. $3 524709
650 4 $a Rhetoric. $3 516647
653 $a Large language models
653 $a Message content qualities
653 $a Persuasion data
653 $a Text analysis
690 $a 0621
690 $a 0459
690 $a 0681
710 2 $a George Mason University. $b Psychology, Human Factors/Applied Cognition Concentration. $e degree granting institution. $3 3825935
720 1 $a McKnight, Patrick $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-02B.
790 $a 0883
791 $a Ph.D.
792 $a 2024
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31240777