Advanced

Självkänsla analyserad med kvantitativ semantisk analys av statusuppdateringar på Facebook.

Wahlgren, Martina LU and Andersson, Amelie LU (2014) PSYK11 20141
Department of Psychology
Abstract (Swedish)
Studien hade två frågeställningar; att undersöka möjligheten att förutsäga en individs nivå på självkänsla genom en semantisk analys av Facebook-uppdateringar, samt att undersöka om det fanns speciella semantiska kluster av ord som var indikativa för självkänsla. Studien utfördes genom en webbaserad enkät som distribuerades via Mechanical Turk. Enkäten bestod av Rosenberg self-esteem scale och en uppmaning att deltagarna (N=155) skulle redovisa deras 15 senaste statusuppdateringar från Facebook. Genom kvantifiering av semantiska representationer, med hjälp av Latent Semantic
Analysis, på statusuppdateringarna visade studien att det inte gick att förutsäga en individs nivå på självkänsla, men det fanns en tendens. Den andra... (More)
Studien hade två frågeställningar; att undersöka möjligheten att förutsäga en individs nivå på självkänsla genom en semantisk analys av Facebook-uppdateringar, samt att undersöka om det fanns speciella semantiska kluster av ord som var indikativa för självkänsla. Studien utfördes genom en webbaserad enkät som distribuerades via Mechanical Turk. Enkäten bestod av Rosenberg self-esteem scale och en uppmaning att deltagarna (N=155) skulle redovisa deras 15 senaste statusuppdateringar från Facebook. Genom kvantifiering av semantiska representationer, med hjälp av Latent Semantic
Analysis, på statusuppdateringarna visade studien att det inte gick att förutsäga en individs nivå på självkänsla, men det fanns en tendens. Den andra frågeställningen besvarades genom att de semantiska representationerna klustrades till åtta kluster. Resultatet visade att kluster två och sju var signifikanta. Kluster två hade ett signifikant
samband med hög självkänsla där ord relaterat till resor, aktiviteter, produkter och mat förekom. Kluster sju hade ett signifikant samband med låg självkänsla där orden präglades av relationer och känslor. En slutsats som kan dras är att individer med hög självkänsla främst betonar vikten av nya erfarenheter och individer med låg självkänsla främst betonar sina relationer. (Less)
Abstract
The study had two purposes; to explore the possibility of predicting an individual's level of self-esteem through a semantic analysis of Facebook updates, and to investigate whether there were special semantic clustering of words that were indicative of selfesteem. The study was conducted through a web-based survey that was distributed via Mechanical Turk. The questionnaire consisted of the Rosenberg self-esteem scale and a request that participants (N=155) would report their 15 most recent status updates from Facebook. By quantification of semantic representations of status updates, through Latent Semantic Analysis, the study showed that it was not possible to predict an individual's level of self-esteem, but there was a tendency. The... (More)
The study had two purposes; to explore the possibility of predicting an individual's level of self-esteem through a semantic analysis of Facebook updates, and to investigate whether there were special semantic clustering of words that were indicative of selfesteem. The study was conducted through a web-based survey that was distributed via Mechanical Turk. The questionnaire consisted of the Rosenberg self-esteem scale and a request that participants (N=155) would report their 15 most recent status updates from Facebook. By quantification of semantic representations of status updates, through Latent Semantic Analysis, the study showed that it was not possible to predict an individual's level of self-esteem, but there was a tendency. The second purpose was answered through the semantic representations that were clustered into eight clusters, where cluster two and seven were significant. Cluster two had a significant association with high self-esteem where words related to travel, activities, products, and food appeared. Clusters seven had a significant association with low self-esteem in which words were characterized by relationships and emotions. One conclusion that can be drawn is that individuals with high self-esteem primarily emphasize the importance of new experiences and individuals with low self-esteem mainly emphasize their relationships. (Less)
Please use this url to cite or link to this publication:
author
Wahlgren, Martina LU and Andersson, Amelie LU
supervisor
organization
course
PSYK11 20141
year
type
M2 - Bachelor Degree
subject
keywords
Latent Semantic Analysis, Semantiska representationer, Självkänsla, Facebook, Statusuppdateringar, Prediktion, Själv-presentation, Self-esteem, semantic representation, Status updates, Prediction, Self-presentation
language
Swedish
id
4465079
date added to LUP
2014-06-16 08:53:51
date last changed
2014-06-16 08:53:51
@misc{4465079,
  abstract     = {The study had two purposes; to explore the possibility of predicting an individual's level of self-esteem through a semantic analysis of Facebook updates, and to investigate whether there were special semantic clustering of words that were indicative of selfesteem. The study was conducted through a web-based survey that was distributed via Mechanical Turk. The questionnaire consisted of the Rosenberg self-esteem scale and a request that participants (N=155) would report their 15 most recent status updates from Facebook. By quantification of semantic representations of status updates, through Latent Semantic Analysis, the study showed that it was not possible to predict an individual's level of self-esteem, but there was a tendency. The second purpose was answered through the semantic representations that were clustered into eight clusters, where cluster two and seven were significant. Cluster two had a significant association with high self-esteem where words related to travel, activities, products, and food appeared. Clusters seven had a significant association with low self-esteem in which words were characterized by relationships and emotions. One conclusion that can be drawn is that individuals with high self-esteem primarily emphasize the importance of new experiences and individuals with low self-esteem mainly emphasize their relationships.},
  author       = {Wahlgren, Martina and Andersson, Amelie},
  keyword      = {Latent Semantic Analysis,Semantiska representationer,Självkänsla,Facebook,Statusuppdateringar,Prediktion,Själv-presentation,Self-esteem,semantic representation,Status updates,Prediction,Self-presentation},
  language     = {swe},
  note         = {Student Paper},
  title        = {Självkänsla analyserad med kvantitativ semantisk analys av statusuppdateringar på Facebook.},
  year         = {2014},
}